*** bigbrother has joined #gnuenterprise
*** bigbrother has joined #gnuenterprise
<reinhard> the functions appserver exports through xmlrpc are documented in doc/api
<insanekane> reinhard, can i reuse GNUe Common for that ? does it provide the necessary modules ?
<Vee2d2> @version
<bigbrother> Vee2d2: The current (running) version of this Supybot is 0.79.9.  The newest version available online is 0.79.9.
<reinhard> yes
<reinhard> you could use the datasources functions
<reinhard> however that would probably mean to build your application on the general gnue-common structure
<reinhard> to some extent at least
<reinhard> lekma: i think that was gnue-gsdgen which did that job
<insanekane> reinhard, you mean there is no generic way to use only the middleware part of Commons ?
<insanekane> reinhard, middleware + data abstraction ...
<reinhard> i have never tried
<reinhard> i have never used common outside gnue
<lekma> reinhard: so how do you generate gcd files??
<insanekane> reinhard, hmm ok ...
<reinhard> jcater and jamest have, so they might be able to tell you more
<reinhard> lekma: gcd is a file format intended for manual writing
<insanekane> reinhard, ok i will ask when they come to the channel ...
<lekma> manual??
<reinhard> our common experience up to now is that writing a gcd with your favourite editor is about 5-10 times faster than using gnue_classrepository.gfd
<reinhard> but i think we still support the other way around, too
<reinhard> just have to check, give me a second
*** jcater has joined #gnuenterprise
<reinhard> speak of the devil :)
<Vee2d2> are there any gnue relevent rss feeds out there.. ?
<reinhard> insanekane: voila :)
<Vee2d2> are the cousins gone forever, I didnt even find reference to them when searching the site
<Vee2d2> heya devil^H^H^H^H^Hkcater
<Vee2d2> jcater, too. :)
<insanekane> jcater, hello :)
<insanekane> reinhard, thanks :)
<jcater> morning
<insanekane> jcater, hi, i am trying to figure out how to use GNUe AppServer and COmmons indenpendently from the Forms and Reports framework ... can you give me some information on this ?
<reinhard> lekma: i think i have a brain blockade now
<reinhard> it looks like there is currently no way to get a gcd or gsd from the stuff you entered in gnue_classrepository.gfd
<reinhard> we completely gave up on that path
<reinhard> because people really had problems with that form
<jcater> insanekane: http://www.gnuenterprise.org/tools/common/docs/
<bigbrother> Title: GNUe Common Library Documentation - GNU Enterprise (at www.gnuenterprise.org)
<jcater> contains a good developer's guide
<jcater> also
<jcater> http://www.gnuenterprise.org/~jcater/examples/
<bigbrother> Title: Index of /~jcater/examples (at www.gnuenterprise.org)
<jcater> contians a few examples from my store
<jcater> bigbrother: wtf is your problem today?
<lekma> reinhard: ok
<insanekane> ok thanks
<reinhard> lekma: i hope very much you didn't spend too much time entering this stuff into the form
<lekma> reinhard: can you take a moment to explain me the new way??
<reinhard> lekma: sure
<lekma> i understand the manual creation of the gcd file
<lekma> but then what??
<reinhard> lekma: please take a look at sample.gcd in the samples dir
<reinhard> that's more or less everything appserver needs
<reinhard> a module self-contained in a single file
<reinhard> then you do gnue-gcd2sql foo.gcd
<reinhard> and appserver does the rest
<reinhard> creates all the classrepository entries
<lekma> ie what happens if i modify an existing class
<reinhard> creates the tables
<reinhard> if you add new properties in a gcd file
<reinhard> just redo gnue-gcd2sql and appserver will automatically update what's needed
<insanekane> jcater, so that means I can switch my existing wxPython+pickled-data application to wxPython+GNUeAppserver keeping the same UI ?
<reinhard> except that appserver won't touch existing columns in the database
<reinhard> i.e. not do any DROP COLUMN or stuff like that
<jcater> yes
<jcater> insanekane: you might also want to check out our subversion tree
<jcater> and look at GNUe Point of Sale
<jcater> it was a custom app I did for my business
<jcater> but is a standalone wxPython application
<jcater> that uses gnue-common's datasources
<lekma> reinhard: ok i'll fiddle a bit with this
<insanekane> jcater, oho ... i will check that ...
<insanekane> jcater, you mean the POS is only in the svn tree ?
<reinhard> lekma: i'm open for any question, thanks for playing :)
<jcater> insanekane: yes
<jcater> I didn't use appserver as my backend though
<jcater> but with the design of common
<jcater> appserver should work the same as using a database directly
<jcater> as far as getting appserver up and running, though
<jamest> bigbrother was upgraded
<jcater> I don't have any experience
<jamest> and installed moments ago
<jamest> any issues blame Vee2d2
<jamest> er, i mean derek
<jamest> sorry Vee2d2
<Vee2d2> :)
<jcater> bigbrother: @weather 38654
<insanekane> jcater, i am interested in using AppServer because  I wish to add some features like using an IM for doing almost-Natural-language queries, or for recieving notifications ...
<jcater> bigbrother: weather 38654
<Vee2d2> @weather 76248
<bigbrother> jcater: The current temperature in Olive Branch, Mississippi is 84°F (10:20 AM CDT on August 27, 2004). Heat Index: 92°F. Conditions: Clear. Humidity: 74%. Dew Point: 75°F. Wind: SSW at 9 mph (15 km/h). Pressure: 30.10 in (1019 hPa). Visibility: 10.0 miles (16.1 kilometers). UV: 7 out of 12
<bigbrother> Vee2d2: The current temperature in Keller, Texas is 83°F (9:53 AM CDT on August 27, 2004). Heat Index: 88°F. Conditions: Scattered Clouds. Humidity: 69%. Dew Point: 72°F. Pressure: 29.94 in (1014 hPa). Visibility: 7.0 miles (11.3 kilometers). UV: 4 out of 12
<jcater> bleh
<jcater> what a sucky day
<Vee2d2> :)
<Vee2d2> good one for the indoors for sure.
<lekma> reinhard: it seems that restarting gnue-appserver does also the trick if you still use classrepository.gfd
<reinhard> lekma: really
<reinhard> wow
<reinhard> this appserver thingy is better than i thought
<reinhard> must be cool people those developers
<lekma> :-)
<reinhard> lekma: however, the disadvantage remains
<reinhard> the stuff you entered in the form is not as easily transportable as a file
<reinhard> like
<reinhard> how would you make a .deb from it ...
<lekma> yep
<reinhard> you'd probably have to use gnue-gsdgen to create a dump of the gnue_* classes
<reinhard> then import it again in the other database with gnue-schema
<reinhard> then restart appserver on the target system
<insanekane> what is this package www in the svn tree ?
<reinhard> insanekane: that's our web page
<insanekane> oh its the website
<insanekane> reinhard, heh ...
<reinhard> i was faster :)))
<insanekane> this is checking that out too ...
<insanekane> reinhard, :) not from my side ...
<insanekane> reinhard, err ... no u indeed were faster ;)
<insanekane> hmm ... there was an exception when trying to view the svn through th web ...
<lekma> reinhard: can you run gnue-gcd2sql while appserver is running??
<insanekane> reinhard, is there any concept of like proxy objects which are instances of objects in the AppServer and abstract away the actual connections ? something like SQLObject ?
<reinhard> lekma: yes you can, and new modules, classes and properties should become active immediately
<reinhard> however appserver caches that info so for changes to become active, you might need to restart appserver
<lekma> reinhard: another question do you still need the gnue_id property in every class you create??
<reinhard> lekma: it is created silently and transparently
<reinhard> you don't need to mention it in the gcd
<lekma> reinhard: cool...
<reinhard> insanekane: yes, i think there is what you want
<reinhard> you can do things like
<reinhard> myCustomerList = find ('customer', {'group': 'abc'})
<reinhard> for c in myCustomerList:
<reinhard>   print c.name
<reinhard>   print c.street
<reinhard>   c.processed = True
<reinhard> ... this will iterate through all customers of group 'abc'
<lekma> reinhard: and the last question for today, i've seen (in packages folder) you can have localised version of modules and classes, how does it work?
<reinhard> print the database fields "name" and "street"
<reinhard> and set the boolean field "processed" to "True"
<reinhard> lekma: you mean the .hu.gcd?
<lekma> yep
<reinhard> that's easy
<reinhard> well, kinda :-)
<reinhard> if a module defines a class
<reinhard> another module can extend the class
<reinhard> so, for example, the module "foo" defines a class consequently named "foo_customer"
<reinhard> with name, street, city
<reinhard> the module "bar" could extend the class foo_customer to add the vatId
<reinhard> so if you only load foo.gcd, you have the table without vat id
<reinhard> if you also load bar.gcd, you have the same table, but with vat id
<reinhard> this is very good to add special fields needed in some countries
<lekma> reinhard: i think i'm starting to get it
<insanekane> reinhard, ok, so creating say a new customer would be like c = Customer(appserverConnection) ???
<reinhard> insanekane: not directly
<reinhard> it is c = new ('customer')
<insanekane> reinhard, hmm ok ...
<reinhard> in fact, there is only a single class
<insanekane> reinhard, aha ... and then setting c.name = "Some Name" would automatically update the db ?
<reinhard> exactly
<insanekane> hmm ok ... super :)
<insanekane> what abt transactions ?
<reinhard> transactions are currently handled by the database
<insanekane> begin() .... commit() ... rollback() i assume ? ;)
<insanekane> oho ...
<reinhard> and yes, you would have a "session" object with these functions
<lekma> reinhard: and old question (and this one will be the last for today, i really have to go) but since there seem to have been a lot of change maybe it's worth asking it...
<reinhard> but these functions don't do more than telling the db to actually do it
<insanekane> reinhard, wow excellent :)
<reinhard> insanekane: now that are the very basic functions of appserver
<insanekane> reinhard, how about implementing functionality to the AppServer ? like for eg, a Jabber server ?
<reinhard> we haven't even started to talk about the *really* fun stuff
<insanekane> reinhard, oh ... what kind of fun stuff ???
<reinhard> insanekane: yeah, that :)
<lekma> how would you implement "unique" attribute for a property and how would you create a "sequence" (pg sense of it)?
<reinhard> lekma: there are triggers
<insanekane> reinhard, what ?
<reinhard> insanekane: just a sec, please let me answer lekma's question first :)
<insanekane> reinhard, oh sure of course :)
<reinhard> lekma: you could use the onCommit trigger to create the sequence
<lekma> in the form??
<reinhard> it will only be generated just before saving
<lekma> or in the appserver??
<reinhard> lekma: no, just define a procedure with the name onCommit
<reinhard> it will be called before committing this object
<reinhard> automatically
<reinhard> no need to do anything in the form
<lekma> ok
<reinhard> lekma: sorry
<reinhard> s/onCommit/onValidate/
<reinhard> there is also an onInit that gets called on creation of a new record
<reinhard> but you can create a new record in a form and decide to not save it
<reinhard> then you lose a number in the sequence
<reinhard> insanekane: appserver is not targeted at internet services
<insanekane> reinhard, ok ???
<reinhard> i didn't quite understand what you meant mentioning a jabber server
<insanekane> well not a jabber server ... well sortof ...
<reinhard> well
<reinhard> just let me explain what you *can* do
<reinhard> :)
<insanekane> ok good idea :)
<reinhard> first to make the definitions clear
<insanekane> ok
<reinhard> to distinguish when we talk about database and when we talk about the appserver
<reinhard> we say "table" and "column" for databases
<insanekane> ok
<reinhard> and we say "class" and "property" for appserver
<insanekane> k
<reinhard> basically, as the first step, think of every class is a table in the db
<reinhard> and every property is a column
<reinhard> and then you can do more
<insanekane> ok
<lekma> reinhard: thx for the explanations, i'll be back maybe later or else tomorrow and i'll bug you a little bit more :)
<reinhard> you can, as explained above, access classes and properties as "native" python objects
<reinhard> also, you can define reference properties
<reinhard> so you can do
<reinhard> if i would be an invoice object
<reinhard> c = i.customer
<reinhard> and c then is a customer object
<reinhard> --
<insanekane> ok
<reinhard> next, you can have procedures
<reinhard> that is, you can do
<reinhard> c.doSomethingWeird()
<insanekane> yes i got that ..
<lekma> bye all
<reinhard> for your customer object
<insanekane> cya lekma
*** lekma has left #gnuenterprise
<insanekane> hmm ok
<reinhard> and behind that is a python code you wrote
<reinhard> even with parameters and a function result
<reinhard> --
<insanekane> u mean it resides on the server ...
<reinhard> yes, exactly
<insanekane> ok
<reinhard> next, you can have indirect properties
<reinhard> like in the above example, you can access
<insanekane> ok
<reinhard> i.customer.name
<reinhard> or i.customer.country.code
<reinhard> or whatever
<reinhard> --
<reinhard> next, you can have calculated fields
<reinhard> if you have a procedure
<reinhard> getopenAmount () for the customer object
<reinhard> you can access c.openAmount as read-only property
<insanekane> hmm ok
<reinhard> like it was a db field, but it is calcuated on runtime
<reinhard> and last, you can have triggers
<insanekane> ok
<reinhard> that are procedures with predefined names that automatically get called on certain events
<reinhard> onInit, onValidate, onChange
<reinhard> onDelete
<insanekane> hmm ok ...
<reinhard> code in this triggers can raise an exception
<reinhard> which means that the commit, change, delete etc. will not take place
<insanekane> ok
<reinhard> that's basically all
<reinhard> of course you can write any code you want into those procedures
<insanekane> ok ...
<insanekane> hmm i think thats more than enough for a variety of tasks
<insanekane> atleast for my lowly ideas ;)
<reinhard> now if you *would* use forms
<reinhard> you would get the ui for many of your classes "gratis" from appserver
<reinhard> because with current development version
<reinhard> appserver can generate an ui definition for gnue-forms on the fly
<reinhard> just from the class definition
<insanekane> oho
<insanekane> but thats a problem for me ...
<insanekane> so classes in AppServer can do one-to-many relationships ? does it do many-to-many ?
<insanekane> and all this is codified by writing an XML file describing the classes yes ?
<reinhard> yes to the XML file
<reinhard> no to many-to-many
<insanekane> hmm how come no many-to-many ?
<insanekane> is there are a better way acoording to GNUe ?
<reinhard> there is no many-to-many in a relational system
<reinhard> and appserver, while taking many advantages from the object approach
<reinhard> is in it's basic concept still a relational system
<insanekane> there is no many to many in a relational system ? no way ...
<reinhard> or i don't understand what you mean?
<insanekane> for eg. one book has many authors, an author may write many books ,,,
<insanekane> so u need 3 tables ... one for book, one for author and one to store the relationships between authors and books
<reinhard> yes
<reinhard> like in a relational db
<insanekane> yes ...
<insanekane> in OO ... we would have 2 classes, and in each class we would store a vector for the property on which the relationship exists ... in book, a vector for authors, and in author a vector for books_written
<insanekane> so is that possible in AppServer ?
<reinhard> there are no vectors
<reinhard> other than you create a detail table/class
<insanekane> hmm ok ...
<insanekane> so there is no concept of arrays ?
<insanekane> as a fundamental datatype ?
<insanekane> ok another question ... the GNUe is under GPL ... so any modules i create with it, are also GPL ?
<insanekane> hello reinhard are you there ?
<insanekane> reinhard, ill be back after dinner ... ciao :) and thanks for all the info :)
<insanekane> jcater, thanks for your help :) bye
*** insanekane has quit IRC
<SachaS> good there are any other appserver users ;)
* SachaS thinks jcater, jamest, derek are waiting till appserver version 1.0 i s out. in that case, go appserver go
<SachaS> going home
<SachaS> see ya later
*** SachaS has quit IRC
*** holycow has joined #gnuenterprise
*** johannesV_ has quit IRC
*** sjc has joined #gnuenterprise
*** wendall911 has joined #gnuenterprise
*** reinhard has quit IRC
*** kilo has joined #gnuenterprise
*** SachaS has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
<SachaS> hi reinhard
*** btami has joined #gnuenterprise
<reinhard> btami: good evening
<reinhard> i will be having a demo with gnue tomorrow
<reinhard> will show them windows, too
<reinhard> i installed newest exe files from download section
<reinhard> noticed that the buttons on the toolbar look quite strange
<reinhard> they all are in a black box
<reinhard> is that normal?
<btami> you have to install the latest common controls from ms download site
<reinhard> ah
<btami> anyhow, i'v tested them on XP only
<reinhard> you can't give me an url, can ou?
<reinhard> oh
<reinhard> win me here :(
<btami> and first got what you got
<reinhard> generally it seems to work
<reinhard> ah
<reinhard> ok
<btami> it needs 6.0 common controls
<reinhard> hrm
<reinhard> is that some dll or such
<reinhard> ?
<btami> yes
<reinhard> i'm not sure i will risk installing something on my running windows machine
<reinhard> and risk breaking something
<reinhard> thanks for your info
<reinhard> that's at least something i can tell them :)
<btami> np
<kilo> reinhard: common controls is official ms stuff
<kilo> you can install it safely
<reinhard> oh
<reinhard> it is safe to install because it is from ms?
<reinhard> that's like you say
<reinhard> it is true because i read it in the newspaper
<reinhard> ;-))
<btami> reinhard: maybe it will not help, i will try it on 98 later...
<btami> in few word the story is:
<btami> in the ToolBar.py i used a big .bmp befor
<btami> it contained all icons
<btami> then jason added 2 new icons to toolbar
<btami> and i don't want to re-create the big .bmp from individual .bmp files again
<btami> instead rewrote the code to use an empty ImageList first and
<btami> add individual .bmp icons in a for cycle as other uidrivers do
<btami> after this change i'v got that strange toolbar what you got too
<reinhard> ok
<reinhard> again thanks for the info
<btami> http://www.gnuenterprise.org/~btami/source-and-py2exe.JPG
<reinhard> off for today
<reinhard> night all
*** reinhard has quit IRC
<btami> bye
*** btami has quit IRC
*** sjc has quit IRC
<switch> :(
*** switch has left #gnuenterprise
*** tocer has joined #gnuenterprise
*** tocer has left #gnuenterprise
*** tocerdm has joined #gnuenterprise
*** tocerdm has left #gnuenterprise
*** jamest has left #gnuenterprise
*** kilo has quit IRC
*** jcater has quit IRC
*** Vee2d2 has quit IRC
*** Vee2d2 has joined #gnuenterprise
<Vee2d2> @seen
<bigbrother> Vee2d2: seen [<channel>] <nick>
*** dimas has quit IRC
*** jcater has joined #gnuenterprise
*** holycow has quit IRC
*** ajmitch_ has joined #gnuenterprise
<ajmitch_> morning
*** tocer has joined #gnuenterprise
*** tocer has left #gnuenterprise
<chillywilly> hi ajmitch_
*** wendall911 has quit IRC
<jcater> word up chillywilly
<jcater> and ajmitch
*** switch has joined #gnuenterprise
<ajmitch> just trying to learn how reports can be useful
*** switch has quit IRC
<chillywilly> whiirrred
<ajmitch> chillywilly
*** ncjp has quit IRC
<chillywilly> ajmitch
<ajmitch> hello
*** holycow has joined #gnuenterprise
*** jcater has quit IRC
*** sjc has joined #gnuenterprise
*** lekma has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
<SachaS> i am reading the log. reinhard, what do you mean we need more chicks? ;)
<lekma> hi all
<lekma> reinhard: can you specify foreign keys in a gcd file ?
<SachaS> lekma yes: as type you just use MODULENAME_CLASSNAME
<lekma> yeah i've seen this, but the foreign key point on gnue_id, what i want
<lekma> is to have it point on a specific field/property of another table/class
<lekma> is this possible
<lekma> ?
<SachaS> hmmm, if i understand right from your questions from yesterday you have another field as primary key in that other table/class.
<SachaS> eg a customer_id
<lekma> yes something like this
<lekma> which brings another question
<SachaS> you would have to ask reinhard, i think per default the primary key is gnue_id
<lekma> can you specify a primary key in a gcd?
<SachaS> the question is: can you define a primary key other than gnue_id
<lekma> yep, thx for clarifying it...
<SachaS> i have not seen that feature in a .gcd
<SachaS> again you have to check with reinhard. it could be a new feature.
<lekma> SachaS: can you take a moment to explain me what those new gld file...
<lekma> are
<lekma> ?
<SachaS> i have played with .gld files. but I still have not completely figured out how they work
<SachaS> i have a project
<SachaS> with a couple of .gcd files
<SachaS> so at the moment I want to concentrate to get the application right == at the moment I do not want to waste time on designing nice forms
<lekma> ok, thx
<SachaS> so i write a .gld file per class
<SachaS> and so I can check the application but dont waste time fiddling with creating nice forms.
<SachaS> does this answer your question?
<reinhard> lekma: currently there is no way to use a field as primary key other than gnue_id
<lekma> reinhard: ok
<reinhard> there *might* be this feature at some point but not in the near future
<lekma> SachaS: how do you use the gld file
<lekma> gnue-read-gld?
<reinhard> lekma: if you look at the gcd files, there's much information in there
<reinhard> and appserver is nearly able to build a form from only this info
<reinhard> the missing info is in the gld
<reinhard> which is the labels to use, the order, and some other details
<reinhard> gnue-readgld
<reinhard> *but* you need post 0.1
<reinhard> i.e. current svn version
<lekma> how do you call it? i mean the form resulting from a couple gcd/gld?
<reinhard> if your appserver connection in connections.conf is [appsconn]
<reinhard> and your class is foo_bar
<reinhard> then you call it with
<reinhard> gnue-forms appserver://appsconn/form/foo_bar
<lekma> wow
<lekma> i'm gonna test this right away
<lekma> reinhard: i still need to know if i can specify a foreign key other than one pointing on gnue_id, is this possible?
<SachaS> reinhard I tried to call a generated form. module x with class y. module z which extends class y of module x. Only the OnInit of modulex x with class y is called NOT all OnInit bound triggers.
<reinhard> lekma: as you can't specify a primary key other than gnue_id
<reinhard> there will be no foreign key other than gnue_id
<lekma> of course..
<reinhard> you can of course just have a normal field
<reinhard> but then you have no referential integrity checking from the db
<reinhard> you would have to do it yourself in triggers
<reinhard> SachaS: you told that to johannes, right?
<reinhard> I think he is looking at it
<SachaS> right. will check again whether all OnInit bound triggers are called of an extended class.
<reinhard> I think he might even already have fixed it
<SachaS> i was working on my project for 2 days. johannes was away for those 2 days. i had some issues.
<SachaS> one issue was: not all bound triggers were called.
<SachaS> another issues was: I had the impression that all extensions (extending properties) must be nullable. otherwise you get an error, obviously all previous records do NOT have the extended properties. in my view that was an limimtation. have to check again.
<SachaS> it might be importantant, when you use package X, which extends a class, that packages X needs/wants to have its properties non-nullable.
<SachaS> i think these were the two main issues I was seeing.
<SachaS> oh yeah, another issue was: to have default values set in a .gcd file.
<reinhard> SachaS: these latter 2 are closely connected
<SachaS> so a onInit fills a record with default values, or if a user hits "new record" the default values will be displayed...
<reinhard> if you don't have default values, extension properties *must* be nullable
<reinhard> or the extension can't be installed into an existing db
<SachaS> ok i see the connection.
<reinhard> ok, demo starting
<SachaS> with having default values, the values of the extended properties of previous records would be set to those default value.
<SachaS> good luck with the demol.
<rm-demo> SachaS exactly
<SachaS> s/demol./demo
<SachaS> maybe there needs a third type: n/a not availlable. because maybe its not a good idea to set the default value to previous records.
<SachaS> johannesV was trying to figure out how appserver could "remember" default values for properties. eg a new gnue_ table etc.
<SachaS> another issue was dependencies: but I think that could be solved like .deb packages. eg packages X depends on package Y which depends on package Z
<SachaS> i have an issue with .gld. for one property I do get a dropdown, for the other property I do not get a dropdown
<SachaS> both propertis are of a referenced data type
<SachaS> the one property where I get a drop down is defined in the same .gcd file
<SachaS> the one property where I do not get a drop down is (eg it references a class of a different module) defined in another .gcd file
<SachaS> re n/a the not available datatype or value. i remember a conversation with a guy of incomplete information. we somehow must give more information about the incomplete information. eg a default value for exmple, or a null value, a n/a value, or a not used value. maybe someone knows more about it.
* SachaS confirms, that two OnInit bound procedures are called. I used a gnue-designer(tm) made form against an extended class.
<SachaS> ok will test with .gld file again.
* SachaS confirms that all (2 in my example) OnInit bound procedures are called when using a .gld form.
<SachaS> when I look at the appserver log window, and when I can trust the order, then the OnInit of  the module which extends the class is called first, last is the original base class.
<SachaS> actually I was expecting the reverse. first the base class onInit, followed by all extended OnInit triggers.
<SachaS> in the current situation I imagine the extended onInit cannot access the base object, as it is not yet created.
<SachaS> will test.
<lekma> SachaS: how do you call an appserver procedure from a form?
<SachaS> the order the OnInit are called should follow the dependency, IMHO
<SachaS> eg the onInit bound trigger or procedure is called whenever you create a new record (or object)
<lekma> the sample.gcd/gfd show datasource.call("proc_name"), but it doesn't seem to work
<SachaS> other wise look in the samples dir, there is a form with a button when pressed calls a procedure on the appserver
<SachaS> ok let me check
<SachaS>   <trigger name="showRecord" type="NAMED"><![CDATA[
<SachaS>     retval = dtsPerson.call("address_show", {})
<SachaS>     if retval is not None:
<SachaS>       print "retval =", retval
<SachaS>   ]]></trigger>
<SachaS> did you use the second argument of the call procedure?
<lekma> ??
<lekma> this sample doesn't do anything here
<lekma> it is supposed to print address_name but it doesn't
<SachaS> i finish first my test, then i look at the sample, one minute.
<lekma> ok thx
<SachaS> reinhard: yes the order of the onInit should be reverse, i think. module x with class y. module z which extends class y. onInit of z_y is called first. there you cannot access x_y so thats a problem.
<SachaS>  RemoteError: Instance '48760175068174029133591956859853' of class 'project_project' not found
<SachaS> ok lekma testing the sample.
<SachaS> lekma, it works for me.
<SachaS> you must have a record in the form.
<SachaS> so search without any criterias
<ajmitch> hi
<SachaS> and you should get 4 records (spock, mccoy, chekov, and kirk)
<SachaS> then hit show record and appserver prints a statement
<SachaLunch> hi ajmitch.
<lekma> i got the records but it doesn't print anything
<lekma> is this supposed to work with 0.1.0 appserveror post 0.1.0?
<lekma> is this supposed to work with 0.1.0 appserveror or post 0.1.0?
<lekma> s/appserveror/appserver
<lekma> hi ajmitch
* holycow waves to all
<holycow> heya
<lekma> hi holycow
*** holycow has quit IRC
*** mixi^ has quit IRC
*** MiXi has joined #gnuenterprise
*** skiold has joined #gnuenterprise
<SachaLunch> lekma it should just work
<SachaLunch> where do you look for something to print?
<SachaS> its where you started the appserver.
<lekma> oohh, wait i'll check
<lekma> SachaS: ok i see it now
<lekma> SachaS: thx
*** Vee has quit IRC
*** Vee has joined #gnuenterprise
<SachaS> ok. lekma.
<SachaS> reinhard: seems like in an onInit I cannot access the objects properties. in samples.gcd onInit: print self.address_name I get an error:
<SachaS>  RemoteError: Instance '14117849080354379075377898981004' of class 'address_person' not found
<SachaS> how did the demo go?
<reinhard> SachaS: to get a dropdown, you have to mark the field you want to see in the dropdown as "search="100""
<reinhard> (in the referenced table)
*** ncjp has joined #gnuenterprise
<reinhard> SachaS: re order of OnInit calls - I agree that the OnInit for the module that defined the class should be first - can you please send a mail to gnue-dev?
*** Vee has quit IRC
<reinhard> lekma: please be aware that the print address_name runs server side!
<SachaS> reinhard: will send the email
<reinhard> SachaS: please also mention the error you get when accessing other properties
<reinhard> lekma: oh, just saw you already found out
<reinhard> demo went fine
<reinhard> they seem to be interested to develop something using appserver
*** ncjp has quit IRC
<lekma> reinhard: is the api.txt doc valid for 0.1.0?
<reinhard> yes
<reinhard> (at least it should)
*** ncjp has joined #gnuenterprise
<ajmitch> reinhard: jbailey got same problem with trying to build appserver deb, I've got to try & track it down now :(
<reinhard> hmmm
<lekma> reinhard: could you provide some example on how to use stringtable and stinglist types?
<reinhard> ajmitch: you have any error message?
<reinhard> lekma: not sure if api.txt is acutally what you want to read at all
<lekma> reinhard: :)
<reinhard> it's the api between appserver and gnue-forms / gnue-reports
<lekma> reinhard: ok, then how can i retrieve all the values from a given field/property?
<ajmitch> reinhard: just that it can't import gnue
<reinhard> ah
<lekma> reinhard: ie not in a form but in a gcd
<reinhard> ajmitch: that means probably that gnue.pth is missing
<reinhard> lekma: you mean in a procedure code?
<ajmitch> reinhard: that's the thing - it's not
<ajmitch> gnue.pth is present, has correct permissions, and is pointing to the right place
<lekma> yep
<reinhard> ajmitch: what happens when you simply do "import gnue" from the python prompt?
<ajmitch> reinhard: for me, it works fine
<ajmitch> on jbailey's box with my gnue-common 0.5.7, it didn't
<reinhard> lekma: can you please explain what you mean with "al the values from a given field/property"
<reinhard> can you give me an example?
<lekma> reinhard: another
<lekma> reinhard: let's say i have a module foo, a class bar an a property barId
<lekma> i want in an OnValidate() in the class foo_bar, to list all the values already in the db for property foo_barId
<lekma> is this possible
<lekma> reinhard: another question is where can i get the list of all predefine bound procedure (OnValidate, OnCommit,...)?
<reinhard> yes
<reinhard> list = find ('foo_bar', {}, 'foo_barId')
<reinhard> for i in list:
<reinhard>   print i.barId
<reinhard> --
<reinhard> you find the list in api.txt
<reinhard> brb
<lekma> reinhard: ok thanks
*** MiXi has left #gnuenterprise
*** MiXi has joined #gnuenterprise
*** Vee has joined #gnuenterprise
*** MiXi has joined #gnuenterprise
*** mixi^ has joined #gnuenterprise
*** MiXi has quit IRC
<lekma> reinhard: list = find ('foo_bar', {}, 'foo_barId') <-- is this post 0.1.0 material ??
*** Vee has quit IRC
*** Vee has joined #gnuenterprise
<reinhard> no, should work with 0.1
<lekma> if i use svn, when i change a gcd i should use gnue-readgcd instead of gnue-gcd2sql?
<reinhard> lekma: yeah, exactly
<reinhard> we renamed that because gnue-gcd2sql doesn't generate sql any more :)
<lekma> thx
<lekma>   File "/usr/lib/gnue/common/rpc/drivers/xmlrpc/pw_xmlrpc/ClientAdapter.py", line 96, in _runMethod
<lekma>     raise client.DistantError, e.faultString
<lekma> DistantError
<lekma> is what i got when using list = find ('foo_bar', {}, 'foo_barId') in Onvalidate
<lekma> in the form side, nothing unusual in server output
<lekma> with 0.1 and svn
<lekma> (the error output is slightly different between 0.1 and svn)
<reinhard> sorry mixed up parameter order
<reinhard> try
<reinhard> find ('foo_bar', properties = ['foo_barId'])
<lekma> ok thx
<reinhard> you can also add a sortorder = [fieldname, fieldname]
<reinhard> to define the sorting of the resulting list
<reinhard> and, in case you care, making the error output more usable is something that we are working on right now :)
<reinhard> just as a minor hint
<reinhard> in case you want to use this for a sequential id
<reinhard> i think it's a suboptimal way to do it
<reinhard> iterating through all records might be bad for performance if your table grows big)
<reinhard> better have a second class where you store the last barId you used
<SachaS> reinhard I was thinking about the bound triggers. the transaction has to be over all bound triggers, not each individual. not sure if that is currently the case or not.
<reinhard> the bound triggers will be part of the same transaction as the event that fires them
<reinhard> i.e.
<reinhard> if you store an object
<SachaS> ok. so thats fine.
<lekma> reinhard: no, i'm trying to test if an entry doesn't already exist... i'm not yet on sequences
<reinhard> all OnValidate triggers *and* the storing of the object itself are one single transaction
<reinhard> oh
<reinhard> lekma: then you'd probably want
<reinhard> if find ('foo_bar', conditions = {'foo_barId': valueToTest}):
<reinhard>   print "already existing!"
<reinhard> and leave the dirty job to the database
<reinhard> SachaS: you ask why we need more chicks
<reinhard> "there's nothing like a good meal, a good wine, and a bad girl"
<reinhard> ;-)
<SachaS> haha
<lekma> reinhard: i have the same error with find ('foo_bar', properties = ['foo_barId'])
<lekma> i'm using current gnue-forms and svn gnue-appserver
<reinhard> lekma: sorry i have to leave right now
<reinhard> :((
<lekma> ok, see you
<reinhard> the problem is
<reinhard> that you don't see the *real* exception in the error message
<reinhard> will be here again on monday
<reinhard> sorry to leave you without being able to help
<reinhard> will try to solve that on monday if you are here again
<reinhard> bye
*** reinhard has quit IRC
<lekma> thx and bye
*** jcater has joined #gnuenterprise
*** lxf has joined #gnuenterprise
*** tocer has joined #gnuenterprise
*** tocer has left #gnuenterprise
*** mixi^_ has joined #gnuenterprise
*** lekma has left #gnuenterprise
*** mixi^ has quit IRC
*** mixi^_ has quit IRC
*** mixi^_ has joined #gnuenterprise
*** lekma has joined #gnuenterprise
*** mixi^_ has quit IRC
*** mixi^_ has joined #gnuenterprise
*** derek has quit IRC
*** derek has joined #gnuenterprise
*** mixi^_ has quit IRC
*** mixi has joined #gnuenterprise
*** lekma has quit IRC
*** lxf has quit IRC
*** Vee has quit IRC
*** ajmitch_ has quit IRC
*** Vee has joined #gnuenterprise
*** ajmitch_ has joined #gnuenterprise
*** skiold has quit IRC
<chillywilly> anyone know of a utility to generate ann md5 password?
*** holycow has joined #gnuenterprise
*** jcater has quit IRC
*** jcater has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
*** cilkay has quit IRC
*** cilkay has joined #gnuenterprise
*** cilkay has joined #gnuenterprise
*** sjc has quit IRC
*** tocer has joined #gnuenterprise
*** tocer has left #gnuenterprise
*** jcater has quit IRC
*** jcater has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
*** insanekane has joined #gnuenterprise
*** sjc has joined #gnuenterprise
*** thomas has joined #gnuenterprise
<thomas> I am trying to find out the status of the financials.  Actually of all the packages.  The www.gnuenterprise.org status system doesnt work.
<thomas> Anybody out there know anything about the status of the packages?
<thomas> Is everything just in the planning stages?
*** sjc has quit IRC
*** thomas has quit IRC
*** insanekane has quit IRC
*** sjc has joined #gnuenterprise
*** sjc has quit IRC
*** sjc has joined #gnuenterprise
*** insanekane has joined #gnuenterprise
<insanekane> hi .. what version of wx is supported by gnuenterprise ? my version is generating a segfault when trying the classrepository.gfd file
*** insanekane has quit IRC
*** insanekane has joined #gnuenterprise
*** mixi^ has joined #gnuenterprise
*** mixi has quit IRC
*** mixi^ has quit IRC
*** mixi^ has joined #gnuenterprise
*** tocer has joined #gnuenterprise
*** tocer has left #gnuenterprise
*** insanekane has quit IRC
<SachaS> insanekane using the classrepository.gfd is outdated.
<SachaS> the latest way to develop an application is to write .gcd files
<SachaS> have a look at the sample.gcd file in the gnue-appserver/samples directory
*** holycow has quit IRC
*** insanekane has joined #gnuenterprise
*** skiold has joined #gnuenterprise
*** insanekane has quit IRC
*** skiold has quit IRC
*** skiold has joined #gnuenterprise
*** skiold has quit IRC
*** psu has joined #gnuenterprise
*** mixi^ has quit IRC
*** mixi has joined #gnuenterprise
*** mixi has quit IRC
*** mixi-two has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
*** reinhard has quit IRC
*** psu has quit IRC
*** skiold has joined #gnuenterprise
*** Ahewes has joined #gnuenterprise
<ajmitch_> hi
*** insanekane has joined #gnuenterprise
<insanekane> hi ... what version of wxPython is suppported for forms ??
*** holycow has joined #gnuenterprise
* holycow waves
*** insanekane has quit IRC
<holycow> hey all
<holycow> anyone around?
<Ahewes> I am
<Ahewes> not that I can help
<holycow> lol thats all right, just curious
<holycow> finally got a standalone machine setup with postgres, now to muck about and see if i can get gnue to connect even :)
<Ahewes> I
<Ahewes> 'm just reading the docs
<Ahewes> I have a lot of web apps that I would like to integrate.
<Ahewes> No idea yet how I'm going to get it all together.
<holycow> *nod* i don't see anything out there like gnue that is this versatile *nod*
*** skiold has quit IRC
<Ahewes> What the world (read: my world) needs is for some project to start hacking and chopping the major web apps to store their data in one place.
<holycow> unfortunately for that you need a framework
<holycow> agreed
<Ahewes> I use sql-ledger, request tracker, squirrelmail, interchange, and about four other apps on occasion.
<Ahewes> Individually, they all rock.
<holycow> lol i know, its a fucking mess
<Ahewes> I can use them on the commandline or at any pc.
*** havoc has quit IRC
<Ahewes> Or my accountant or customers or anyone else for that matter.
<holycow> *nod*
<Ahewes> But I have way, way too much data.
<holycow> every single framework i've seen is so very limited in scope
<holycow> or so far into theoretical data related stuff yeah
<holycow> storing all of those things in postgres, and having a framework to build any type user interface is really th eonly answer
<holycow> i would like to have same as y ou
<holycow> but also be able to build web interfaces as well as pda level interfaces
<holycow> sometimes you just need certain information on a pda for example
<Ahewes> well, I don't need a pda interface, but I do need to store all my customer data in one unified place.  I know there are haks for most of the apps I use to store *some* data in an LDAP store, but not all of them.
<holycow> well... you don't need it now may be a better way to say that, i'm betting that down the road ui display needs will change on a org by org basis
<holycow> :) just my personal guess tho
<holycow> as well as driving the data strage integration at the same time
*** havoc has joined #gnuenterprise
<Ahewes> is there a demo of gnue integrating with a web app somewhere?
<holycow> not to my knowledge
<Ahewes> bummer.  well, downloading it all again anyway.
<Ahewes> I periodically just download gnue, compile, and try to get an example working.
<holycow> i've only got it to work once long ago
<holycow> the documentation isn't newb friendly exactly
<holycow> i'm hoping to get enough questions answered that i can write the guide tho
<holycow> lol
* ajmitch returns
<holycow> hola aj
<Ahewes> does gnue have a wiki?
<holycow> yep
<holycow> main menu gnue.org
<Ahewes> thx
<Ahewes> trying the debian install again...there are a lot of recommended packages this time around.
<Ahewes> anyone know what this firebird c32 server is?
<Ahewes> hmmm...methinks there should be generic install notes instead of a different set for each platform.
<holycow> never heard of it
<holycow> apt-cache show package is usually helpfull
<Ahewes> well, the debian install notes are about three lines long, and the last line just gives this firebird database as an example.
<Ahewes> apt-get update
<Ahewes> apt-get install gnue-.*
<ajmitch> heh
<ajmitch> use postgresql or mysql, I think
<holycow> well firebird is a db like postgres, but i've never used it
<ajmitch> Ahewes: what debian branch are you using?
<holycow> ajmitch, how does the app server apply to work flow in general terms... for example:
<holycow> i have a workstation and db server
<ajmitch> holycow: don't ask me
<holycow> lol
<ajmitch> I don't know the appserver at all
<holycow> that sounds like you can create forms without app server?
<ajmitch> sure
<ajmitch> most use of gnue is without appserver
<holycow> really?
<holycow> oh neat
<ajmitch> Ahewes: I'm guessing you're using woody...
<holycow> okay so when i insert data source into a new app, i get a few options, none of which allow me to specify an ip, user, pass ... is there another place to specify a remote datasource?
<ajmitch>  /etc/gnue/connections.conf
<holycow> ahhhh!
<holycow> danke
<ajmitch> or wherever that file is on your box
<ajmitch> datasources are defined there :)
<holycow> *nod* sorry i expected that to be in properties, no that makes sense
<holycow> <-- using subversion grab, so in my home dir
<ajmitch> aha
<ajmitch> ok, I have to leave now, will be back in an hour or so :)
<holycow> cya, thx for answering my stupid questions
<holycow> :)
<holycow> oh no shit
<holycow> that worked
<holycow> heh
<holycow> something has to be done about these tiny little non resizeable popup windows tho
<holycow> heh
<holycow> i feel like i'm developing in lotus notes again
*** sjc has quit IRC
*** tocer has joined #gnuenterprise
*** tocer has left #gnuenterprise
<holycow> Ahewes, how far have you gotten with each install
<holycow> manage to connect to a db and make some fields yet?
<Ahewes> I got distracted by expert systems...
<Ahewes> I got the install of the packages done, and picked postgres, so
<Ahewes> I installed the psycopg package.
<Ahewes> I started postgres up but then got to thinking about other things I wanted to try with postgres, so I'm a bit off track.
<holycow> hehehe
<holycow> it's a terrific db eh?
<Ahewes> I think I should be ready to edit /etc/gnue/connections.conf
<Ahewes> but it's time to pick up a prescription from the pharmacy so I'll try it when I get back.
<Ahewes> postgres is very nice.
<holycow> i'm off for some dry ribs
<holycow> lol
<Ahewes> I think they hae their shit together.
<Ahewes> very good docs.
<Ahewes> ttys
<holycow> Ahewes, as a complete newb i found it frighteningly easy to setup
<Ahewes> I'll add to the wiki when I'm done.  I'm taking notes.
<holycow> trying to setup something like mysql, or god forbid mssql is a lil' more complicated
<Ahewes> back in an hour.
<holycow> cya
<ajmitch> Ahewes: what version of the gnue .debs are you using?
<holycow> ajmitch, what is the basic process of geting information out and into a table
<holycow> say first name last name, one form two text fields
<holycow> i'm not groking the concepts as i do little db dev type stuff
<ajmitch> hmm?
<ajmitch> you mean in a form?
<holycow> yep, just trying to understand the basics, simple form with two fields, for input
<ajmitch> what's not working?
<ajmitch> have you read the forms developers guide?
<holycow> ah, the one in the gnue-forms/doc directory?
<ajmitch> yeah
<holycow> oh okay, nope, didn't realize the help files were there ... makes sense
<holycow> X)
<holycow> two questions:
<holycow> my datasource has a name: that i gave it, source: tables (sounds right) but no connection: information
<holycow> where is that input, i presume in need to select the column/field/table somewhere?
<ajmitch> something like this
<ajmitch> <datasource name="dtsOrders" connection="ao" primarykey="id" table="orders"/>
<holycow> okay when i'm looking at the block editor it has 4 columns
<holycow> the datasource filed there is supposed to be what? the database name?  and the base table column is the table i'm working on?
<ajmitch> in designer?
<holycow> sorry, yes designer
<ajmitch> datasource is the datasource that you defined...
<ajmitch> and the table is the table you work on
<ajmitch> I see that the datasource editor isn't showing the connection, but the property editor does
<holycow> *nod* yep
<ajmitch> I don't know enough about designer to help much :)
<holycow> hehe no worries, i'm just hoping for luck
<holycow> :)
<holycow> ajmitch, who would be the designer expert in here?
<holycow> probably sacha
<ajmitch> jcater
<holycow> danke
*** cilkay has left #gnuenterprise
*** lekma has joined #gnuenterprise
*** jcater has quit IRC
*** reinhard has joined #gnuenterprise
<Ahewes> hmmm... someone working on the gnue site?: http://www.gnuenterprise.org/docs/
<bigbrother> Title: Index of /docs (at www.gnuenterprise.org)
<Ahewes> I get a directory listing.
*** SachaAway has quit IRC
*** johannesV has joined #gnuenterprise
<reinhard> Ahewes: the website is in a state of transition
<reinhard> for current docs, best way I think is to select the tool you want to look at
<reinhard> and then use the doc link from that page
<reinhard> i.e. if you look for a doc regarding gnue-forms
<reinhard> first go to the gnue-forms page
<reinhard> then there the docs are listed
*** SachaS has joined #gnuenterprise
<ajmitch> reinhard! :)
<SachaS> good morning
<ajmitch> hi sacha__
<reinhard> ajmitch!
<ajmitch> SachaS..
<SachaS> sacha__ is a simple irc client
<ajmitch> reinhard: /usr/lib/python2.3/site-packages/gnue.pth was getting 'usr/lib/gnue/python/'
<SachaS> in case bibrother has issues
<ajmitch> and I have no idea why it was doing that :)
* SachaS is trying johannesV new .gld patch
* SachaS will do so in a couple of minutes
<reinhard> ajmitch: isn't that correct?
<ajmitch> reinhard: no / at the beginning
<reinhard> oh
<ajmitch> so I've filed a grave bug against my own package until I know it's fixed :)
<ajmitch> I can't reproduce it, even in a chroot
<reinhard> I'd bet it depends on the version of python you use
<ajmitch> but I'm wondering if it's specific to the package versions in testing, which I haven't tried
<ajmitch> yeah..
<reinhard> distutils (who handles these paths) is part of the python package
<ajmitch> I know
<reinhard> well, maybe python-dev (not sure)
<ajmitch> and the call is distutils.sysconf.get_python_lib()
<ajmitch> iirc
<ajmitch> distutils.sysconfig.get_python_lib()
<reinhard> jbailey did try to build and failed, right?
<reinhard> what version of debian did he try to build on?
<ajmitch> it would have to have been sid or sarge
<reinhard> well
<ajmitch> but I'm not sure what he built with
<reinhard> that's actually the question :)
<reinhard> what did you try?
<ajmitch> sid
<reinhard> i always thought the buildd's run on sid
<ajmitch> I don't have a sarge chroot setup yet :)
<ajmitch> yes, they do
<ajmitch> but this is an arch: all package so the buildd's don't touch it
<reinhard> i can try to reproduce the error on sarge today
<reinhard> later today
<reinhard> will let you know
<reinhard> off to meeting
<ajmitch> at the moment I've got an os.path.abspath() call around the function call
<ajmitch> to see if it works :)
<lekma> ajmitch: i had the same problem when packaging common for gentoo (ie the usr without leading /)
*** dimas has joined #gnuenterprise
<ajmitch> lekma: great! how'd you fix it? ;)
<lekma> with a hammer:
<lekma> -      p = path [len(self.root):]
<lekma> +      p = "/" + path [len(self.root):]
<lekma> in setup.py
<ajmitch> I was worried about that...
<ajmitch> I might do the same
<lekma> :-)
<ajmitch> good to know that it wasn't just me seeing it :)
* ajmitch has to run away now, bbl :)
*** insanekane has joined #gnuenterprise
<insanekane> reinhard, hello are you there ?
*** holycow has quit IRC
<johannesV> SachaS: have updated the referenced-classes gld ?
<SachaS> no
<insanekane> what is the supported version of wxPython for forms component ?
<SachaS> johannesV: situation: module x with class y. module z which extends class y.
<johannesV> you have to add a search-attribute to the class beeing referenced in order to have it as dropdown
<SachaS> do you expect 2 gld's?
<johannesV> yes
<johannesV> one for module x and one for module y
<johannesV> err
<johannesV> z
<SachaS> ok.
<johannesV> same as gcd: one file per module
<SachaS> sorry but the info and search is confusing.
<SachaS> ok will try.
<SachaS> one sec.
<reinhard> lekma: what did you use for root= ?
<SachaS> johannesV: i have locale to: de_CH.UTF-8
<SachaS> it only takes de_CH thats OK.
<SachaS> so i call gfcvs -u gtk2 appserver://appserver/form/project_project
<SachaS> project_project => moduleX_class_Y
<reinhard> insanekane: 2.4 IIRC
<reinhard> insanekane: though you'd probably better use the gtk2 user interface
<reinhard> gnue-forms -u gtk2 ...
<reinhard> :q
<reinhard> errr
<insanekane> reinhard, ok ... i got a segfault with wxPython 2.5
*** sjc has joined #gnuenterprise
<reinhard> insanekane: that's possible
* SachaS has to think for a moment
<reinhard> wx is known for incompatible changes between versions :(
<lekma> reinhard: --prefix=/usr --install-lib=/usr/lib i think tht's all i did
<insanekane> reinhard, :)
<reinhard> lekma: you didn't use --root ?
<insanekane> reinhard, designer seems to require some attributes of the underlying python-C++ interface ..
<reinhard> for designer, you need 2.4
<reinhard> it won't work with anything except 2.4
<reinhard> lekma: the change you posted above would only have *any* effect if you use --root
<lekma> --root=/ i think, it's imported from the gentoo distutils classes and is a fake root up until the files are installed
<lekma> let me verify
<lekma> yep that's correct, in my case --root=/, that's standard installation procedure for gentoo
<reinhard> ah
<reinhard> there is a bug in setup.py
<reinhard> to build the path in gnue.pth, the value of --root is stripped from the directory
<reinhard> so if you have --root=/tmp/install and your dir ist /tmp/install/usr/lib
<reinhard> then everything is ok
<reinhard> but if --root=/
<reinhard> and lib = /usr/lib
<reinhard> then it strips the "/" at the beginning :(
<reinhard> must be the same problem with debian
<insanekane> reinhard, btw, when i tried to setup appserver from the cvs (ie, using the cvs install method), there were errors - t couldnt "find" the psycopg drivers ... but after installing using the normal method, it works ...
<lekma> reinhard: i'm playing right now with the rpc server, trying to test a xul client...
<reinhard> insanekane: that's strange
<johannesV> Sacha: is it now working ?
<insanekane> reinhard, indeed
<johannesV> you would have to add a gld for the moduleX
<johannesV> where classY get's a search-property or you have a gld for module-Z where classY has a search-property
<reinhard> lekma: did you find out anything wrt your problem with find()?
<SachaS> johannesV: i cant get the .gld right
<lekma> am i correct: Session.open ({'password': 'test', 'user': 'test'},) should open a session and return a session nr
<lekma> ?
<reinhard> yep
<lekma> reinhard: yep the find()is solved
<reinhard> what was it?
* reinhard is curious
<lekma> i was being stupid
<lekma> :-)
<reinhard> well
<insanekane> reinhard, which version of pyGTK should I install ?
<reinhard> that's a reliable source for problems :)
<lekma> :)
<lekma> btw is there a way to import procedures from another gcd?
<reinhard> lekma: you can write a .py file and import that from both procedures
<reinhard> if you want to share code
<lekma> you see, i'm being stupid again...
<lekma> :)
<SachaS> lekma: appserver has even a python shell running where you can connect to (just kidding)
<SachaS> ;)
<lekma> :)
<johannesV> SachaS: send me the gcd's and the gld's you have by now and i'll have a look at them
<lekma> is it forbidden to declare a dict like this: authentication = "{'password': 'test', 'user': 'test'}" (with the double quote)?
<lekma> i mean in python
<reinhard> with the double quote it will be a string
<reinhard> you have to leave away the double quote
<SachaS> busy by writing other email
<lekma> i'm not sure i can... it comes from a javascript, i'll test
<reinhard> lekma: I'm not sure how python translated dictionaries into xmlrpc
<reinhard> in any case
<reinhard> you have to write something in javascript that gets translated to the same xmlrpc construct
<insanekane> i think there is already an XPCOM component for XML RPC .... IIRC
<reinhard> lekma: looking at the webfrontend js code siesel once wrote it seems like js also knows the {} syntax
*** skiold has joined #gnuenterprise
<SachaS> johannesV: its working.
<johannesV> ok
<SachaS> johannesV: I saw you did the sequence of the OnInit, OnValidate etc
<johannesV> right
<johannesV> well it's not a real sequence
<johannesV> the only thing we can do is this way
<SachaS> i think they must follow the dependencies
<johannesV> which dependencies ?
<SachaS> module x depends on module y which depends on module z
<SachaS> so the OnInit goes: z, y, x
<johannesV> i can only say first to call the OnInit-trigger of the module which defines the class, and than calling OnInit of all modules extending the class (without a determined order)
<johannesV> how can a module depend on another one ?
<SachaS> hmmm
<SachaS> like module y introduced properties
<SachaS> module x uses thos properties
<SachaS> eg module y MUST be installed so that module x can be used.
<johannesV> well, but OnInit get's called for a given class !
<johannesV> so let's say the class 'address_country' has an OnInit-trigger
<johannesV> defined by the module 'address'
<johannesV> and there's a module 'foo' extending address_country by adding a new property and
<johannesV> another OnInit-Trigger (of the module foo)
<SachaS> have to run for lunch. will read log. have not thought it through but got the impression that dependencies must be checked. see you later
<johannesV> so the sequence will be: address_country.address_OnInit, address_country.foo_OnInit
<johannesV> i think you're mixing up something
<johannesV> the circumstance that a module extends a class does NOT introduce any dependency ...
*** insanekane has quit IRC
<lekma> reinhard: is there a way in geasSessionManager.py to know if the client is gnue-forms or some other?
<reinhard> nope
<SachaS> johannesV still.
<SachaS> i might mix something up.
<SachaS> lets say module X is base backage
<SachaS> module Y is finance package
<SachaS> module Z is statistics package
<SachaS> module U is production package
<SachaS> in module Z (or a class of module Z) we might need some case statements
<SachaS> eg module Z can produce a status of the organisation
<SachaS> so in module Z:
<SachaS> if module Y is installed to # crazy statistics stuff; end
<SachaS> if module Y is installed do
<SachaS>   crazy finance statistics
<SachaS> end
<SachaS> if module U is installed do
<SachaS>   crazy production statistics
<SachaS> end
* SachaS thinks we have to wait till i got a real use case
<SachaS> one could argue, that the Z statistics module extends certain classes with doCrazyStatics procedures
<SachaS> we have to wait till i get a real use case.
<SachaS> anyways ;) I got .gld working the way I need it.
<SachaS> and the order of OnInit is now with your patches as I was expecting it
<ajmitch> hi
<SachaS> hi ajmitch
* SachaS thinks that it cannot be that the sequence of the calls of the OnInit's can be arbitrarly
<SachaS> module U with class A, module V extends A, module W extends A
<SachaS> with W depends on V which depends on U
<SachaS> the OnInit must be:
<SachaS> U_A.A_onInit, V_A.V_onInit, W_A.W_onInit
<SachaS> because extensions of A in W might use properties of V_A
<SachaS> I think the same is true for onValidate
<SachaS> how about onDelete?
<SachaS> I guess there it takes the referse order:
<SachaS> W_A.W_onDelete, V_A.V_onDelete, U_A.A_onDelete
<SachaS> so how can one define the dependency?
<SachaS> if someone installs module U
<SachaS> followed by module W it will fail (if W depends on V)
<SachaS> does it fail?
<SachaS> actually I might not fail, actually it might even install
<SachaS> but if procedures in W_A use properties from V_A, then they will fail.
<SachaS> at run time
<johannesV> SachaS: module U defines the class; that's ok - so it will be the first OnInit and the last onValidate
<johannesV> but modules V and W are in a random order
<SachaS> ok. is there a way to define the order?
<johannesV> there can't be a dependancy between them, since they might not know each other
<SachaS> i think currently there is not.
<johannesV> no there is NO order
<SachaS> base module, finance module, add-on-finane module
<SachaS> I want to express, that add-on-finance module depends on finance module
<SachaS> as I said I do not have a use case in my application
<SachaS> but if many modules extend a class X I seriously wonder if the random order of the executions will cause problems.
<SachaS> s/the execution/the bound trigger execution/
<SachaS> anyways I shut up and do some gnue use cases :)
* SachaS started this morning with a weekly planning spreadsheet. now i think damn I should do this with gnue TOO. everything I do I think I should do it with gnue TOO :) Think thats a positive sign. OK I now realy shut upt.
<SachaS> BTW genereated forms kick ass for rapid prototyping!
<SachaS> somehow I would like to see hyperlinks in a generated form to open up a new form :)
<johannesV> SachaS: that would be gnue navigator then :)
<SachaS> yeah
<SachaS> gnue navigator does understand appserver://appserver/form/my_sexy_form
<SachaS> right?
<SachaS> :)
<johannesV> right
<johannesV> this is nothing specific to gnue-navigator but to gnue-forms
<johannesV> (reading appserver://... is built into GFInstance.py)
<SachaS> thanks.
<SachaS> its getting better and better.
<SachaS> whats on your (johannesV, reinhard, other appserver devs) guys radar?
* SachaS peeks into the roadmap
<johannesV> exception-handling, gnue_messages, company- & year-support, session-management
<ajmitch> johannesV: you don't need navigator to open another form.. :)
<ajmitch> that's what runform() is for :)
<SachaS> ajmitch still lives in 2-tier world
<SachaS> time to get into n-tier world ajmitch :)
<johannesV> ajmitch: right, but this way i would have to code the call to runForm () into a given form
<johannesV> but if a form is created dynamically it wouldn't have that code built in
* ajmitch is trying to fixup the gnue-common package
<reinhard> ajmitch: you read the logs?
<reinhard> what do you pass to --root ?
<ajmitch> reinhard: yeah, thanks for the top
<ajmitch> top
<ajmitch> bah
<ajmitch> tip ;)
<ajmitch> running build_scripts
<ajmitch> cd . && /usr/bin/python setup.py install --root=/home/ajmitch/debian/gnue/gnue-common-0.5.7/debian/gnue-common/ --no-compile -O0 --prefix=
<ajmitch>  /usr --install-lib=/usr/lib/gnue/python
<ajmitch> trying to create /home/ajmitch/debian/gnue/gnue-common-0.5.7/debian/gnue-common/usr/lib/python2.3/site-packages/gnue.pth
<ajmitch> self._quote is buggy, I presume :)
<reinhard> ah
<ajmitch> blame cdbs!
<reinhard> can you run it without the trailing / for --root?
<ajmitch> nope
<ajmitch> cdbs adds that in, iirc
<reinhard> damn
<reinhard> it didn't for earlier versions of cdbs, i guess
<reinhard> you are right, self._quote is buggy
<reinhard> in the sense that it should remove the trailing slash if there is one
<ajmitch> added that
<ajmitch> testing...
<ajmitch> appears to work
<ajmitch>     if self.root[-1] == "/":
<ajmitch>       self.root= self.root[:-1]
<ajmitch> in self._quote
*** lxf has joined #gnuenterprise
<reinhard> ajmitch: better would be
<reinhard> root = self.root
<ajmitch> true
<reinhard> if root [-1] == os.sep
<ajmitch> no point making it a global change
<reinhard>   root = root [:-1]
<ajmitch> reinhard: btw, when making tarball releases, is it possible to lowercase the directory name? :)
<ajmitch> change appears to work
<reinhard> is it a problem with the uppercase name?
<reinhard> we would have to change the name of the tarball, too
<reinhard> to gnue-common.0.5.6.tar.gz
<reinhard> instead of GNUe-Common.....
<ajmitch> not so much a problem as it gives me a warning when I build
<reinhard> ok
<ajmitch> currently I do the naughty step of repacking the tarball with a new dir name :)
<reinhard> i will talk with j* to rename the package names
<ajmitch> thanks
<reinhard> I have read several times that many people don't like tarballs that aren't fully lowercase
<ajmitch> it's probably in gnu coding standards as well
<reinhard> most likely
* SachaAway is away: away
*** mixi-two has quit IRC
*** dimas_ has joined #gnuenterprise
*** mixi has joined #gnuenterprise
<ajmitch> reinhard: ok, gnue-appserver should be buildable now :)
*** jamest has joined #gnuenterprise
<lekma> about the irc-logs, since 2004-08-24, they are all empty, are they still available somewhere else than http://www.gnuenterprise.org/irc-logs/ ?
<bigbrother> Title: Index of /irc-logs (at www.gnuenterprise.org)
<lekma> ?? bigbrother is a bot?
<ajmitch> jamest!!
<ajmitch> lekma: yes bigbrother is a bot
*** dimas_ has quit IRC
<jamest> that's me
<ajmitch> jamest: do you know of any tips/guides/howtos/anything on how to use & learn reports? ;0
<jamest> not really
<jamest> every time i use reports i use whatever samples are availabe
<jamest> and bug jcater
<jamest> :)
*** psu has joined #gnuenterprise
*** skiold has quit IRC
<reinhard> ajmitch: excellent
<reinhard> ajmitch: so i could expect upload of appserver, forms, and common soon?
<psu> plus docs ;-)
<ajmitch> reinhard: what forms version?
<ajmitch> 0.5.6?
<reinhard> 0.5.6
<reinhard> yes
<ajmitch> good thing you reminded me
<ajmitch> sponsor is busy at work, will try & get it uploaded soon
<reinhard> thanks
<ajmitch> if I see jbailey around I'll hunt him down also
<psu> hi ajmitch
<psu> up late, or up early?
<ajmitch> up late
<ajmitch> about 1:30 here
*** mixi has quit IRC
*** mixi has joined #gnuenterprise
*** mixi has quit IRC
*** mixi has joined #gnuenterprise
*** insanekane has joined #gnuenterprise
*** lekma has quit IRC
*** skiold has joined #gnuenterprise
<insanekane> reinhard, wxPython 2.4.2.4 is ok for designer ??
<reinhard> i *think*
<reinhard> not working with designer myself, I can't really tell
<reinhard> bbl
<johannesV> SachaAway: aren't you using gtk2 as your default-UI ? (cause you're adding a -u gtk2 argument to your gfcvs call)
*** lxf has quit IRC
*** lxf has joined #gnuenterprise
<SachaS> johannesV: ok changed to gtk2 as default in gnue.conf
<SachaS> thanks
*** skiold has quit IRC
*** lxf has quit IRC
*** johannesV has quit IRC
*** psu has left #gnuenterprise
*** johannesV has joined #gnuenterprise
*** Ahewes has quit IRC
*** dcmwai has joined #gnuenterprise
*** jcater has joined #gnuenterprise
*** psu has joined #gnuenterprise
*** dsan has joined #gnuenterprise
<dsan> hi there
<jamest> jo
<jamest> er
<jamest> hi
<dsan> pretty sure u already know it but err...
<psu> hi dsan
<dsan> viewcvs.cgi is broken on gnue website
<dsan> module "svn" not installed.
<dsan> hi psu
<dsan> erm... sorry ... s/cvs/svn ;)
<dsan> I was interested by your svn2cl script, but can't find it so far
<dsan> bbl
*** dsan has left #gnuenterprise
<SachaS> johannesV: i am again working with the .gld
*** dimas_ has joined #gnuenterprise
<SachaS> module x with class y and module z with class a
<SachaS> class y references class a
*** reinhard has quit IRC
<SachaS> i do not get it right so that i have a form of y with one extra tab for the display of class a
<SachaS> i do get an extra tab with one drop down of the one search property of the a.gld
<SachaS> pure z_a works fine
<SachaS> but not as detail of a masters x_y
*** insanekane has quit IRC
<psu> derek - r u about?
<psu> just wanted to say
<psu> I agree 1000% with you on web applications
<psu> My work laptop has been b0rked for the last 2 weeks
<psu> so I have had no option but to use webmail.
<psu> It's soooo slow
<psu> Feels like it's running on another computer, or something ;-)
<psu> The idea of using a web-based ERP app is now even less appealing than it was
<SachaS> johannesV: the label of an info tag of a .gld is not shown in the form.
*** psu has left #gnuenterprise
<johannesV> right
<johannesV> an info is used as a entry with style label
<johannesV> f.e. you want to have a dropdown with address_country.code and as an information address_country.name
<johannesV> so giving <property name="code" pos=".." search="100">
<johannesV> and <property name="name" pos=".." info="200">
<SachaS> oh ok that is the info for.
<SachaS> yeah thats what i thought of
<johannesV> info means just 'display as information'
<SachaS> ok is the space between a search and an info dependent on something? i get a long space between the two
<johannesV> no, there shouldn't be a space in between ...
<johannesV> do you have info before search or the other way round
<SachaS> check again
<SachaS> johannesV the master/detail thing in a .gld
<SachaS> can give me antoher hint about it?
<SachaS> it sort of works with the search and info
<SachaS> but it is intended to be used just for one line, right?
<johannesV> there's no master/detail in gld
<SachaS> like the country example given above
<johannesV> there is only lookups in gld
<johannesV> using our samples: country is the master, person the detail
<johannesV> so if editing a person, country is used for lookups
<johannesV> so in the country-class you would add search- and info-values
<SachaS> i dont understand.
<johannesV> because you want to say which properties are used when another class references a country
<SachaS> ok.
<SachaS> hmmm
<SachaS> person and country
<SachaS> two classes
<SachaS> if person references country
<SachaS> oh only looksup -> i cannot have a person displayed and edit its fileds and as well have its associated country (address in my example) in the same form.
<johannesV> of course you can
<johannesV> just look at the sample.gfd in appserver/samples
<johannesV> this is the way it works
<SachaS> damn
<SachaS> ok i look.
<johannesV> after looking at sample.gfd import the newly sample-*.gld in our appserver/samples and have a look at appserver://appserver/form/address_person
<johannesV> then you should get a similair form
<SachaS> ok the sample does not make sense as you do not want to change the country business class.
<johannesV> ?
<SachaS> changing like: de is no longer Germany but Deutschland
<SachaS> all you can do is assigning a new country to the person
<johannesV> *gotcha*
<SachaS> ok
<SachaS> ;)
<johannesV> as i have mentioned before
<johannesV> there is no master/detail in gld
<SachaS> arghhh
<SachaS> :)
<SachaS> cant you implement it?
<SachaS> ;)
<johannesV> what you are looking for is a relation like invoice and invoice-lines
<johannesV> no
<johannesV> we don't want to implmenet that (now)
<SachaS> can you change the roadmap?
<SachaS> ;)
<SachaS> ok gotcha
<SachaS> so i must write a custom form :(
<johannesV> having such relations in generated forms does not make much sense
<johannesV> well you can use the debug-file-option to store a generated form
<johannesV> and the modify it
<johannesV> :)
<johannesV> like appserver://appserver/form/masterclass?debug-file=myform.gfd
<SachaS> yes
<johannesV> this will store the gen.form code in the local file 'myform.gfd'
<SachaS> master/detail would be cool
<johannesV> if you do this for the detail-class as well you would have to merge both gfd' only
<johannesV> it cannot be handled in an automatic way so one could use the result
<SachaS> i just thought i can do the application without needing custom forms
<johannesV> it would be less effort to use something like designer to create them
<johannesV> btw. forms with such relations are usually subject to the customer
<SachaS> ok ok no master/detail now
<johannesV> so one would have to design them manually anyway
<johannesV> we have talked about this a few days ago (i thought)
<SachaS> we did
<SachaS> even we sometimes talk of the same thing, we might be speaking of it on different levels :)
<SachaS> well
<SachaS> no m/d :(
<SachaS> ok
<SachaS> no worrie
<SachaS> no worries
<SachaS> thanks.
<johannesV> np
<SachaS> johannesV: there is a capitalisation issue:
<SachaS> DB001: 0:02:30.526 [ServerAdapter:406] localhost - - [30/Aug/2004 19:41:22] "POST / HTTP/1.0" 200 -
<SachaS> Auflösen: Session.request (1, 'partner_customer', [], [], ['partner_address', 'partner_changedWhen', 'partner_ba', 'partner_bA', 'partner_shortName', 'partner_isActive', 'partner_name', 'partner_createdWhen', 'partner_changedBy', 'partner_createdBy'])
<SachaS> do you see the partner_ba and partner_bA ?
<SachaS> it should be a partner_bA
<SachaS> this happens when I go into search mode and then hit query a generated form.
<johannesV> hmm, i don't think mixing case this way is a good idea ... :)
<johannesV> i'm not sure if db-backend supports this
<johannesV> appserver's classrepository is case-sensitive, but the backend usually not
<SachaS> well I have to be case correct when i create the .gld file
<SachaS> otherwise appserver does not accept it
<SachaS> if I set it to lowercase in the .gcd and .gld it works
<SachaS> if i set the field to camelcase in the .gcd and .gld it does not work
<johannesV> gcd and gld must match
<SachaS> if they match in camelcase the generated form does not work
<johannesV> i think importing a 'non-existent' property from gld would raise an error
<SachaS> if they match in lower case the generated form  works
<johannesV> ?
<johannesV> i'll try that ...
<johannesV> just a sec
<SachaS> its a referenced field I am talking
<johannesV> camelcase does NOT work ?
<SachaS> nope
* johannesV has to reboot after distupgrade
*** johannesV has quit IRC
*** johannesV has joined #gnuenterprise
<johannesV> SachaS: using camel-case properties works well
<johannesV> but i've lost my syntax-highlighting during the last upgrade :(
<johannesV> have to tweak that vim again
<SachaS> johannesV I have cameCase issues
<SachaS> in my setup
<SachaS> the ones which are not working are:
<SachaS> references to other classes from other modules.
<SachaS> i can send you the files, if you want.
<SachaS> the project :)
<johannesV> yes please send them
<johannesV> (have tried it with ref's too which doesn't matter)
<SachaS> hmmm
<SachaS> ok send it to you
<johannesV> ah, got syntax-highlighting and filetype-detection back in vim :)
<johannesV> good, i'll look at it tomorrow morning
<SachaS> ok
<SachaS> thanks
<SachaS> i can do it all lowercase until tomorrow ;)
<johannesV> SachaS: but send the camelcase file anyway ... so we can catch a bug if there is one !
<johannesV> ok, i'm off for today
*** johannesV has quit IRC
*** jcater has quit IRC
*** SachaZzz has quit IRC
*** jcater has joined #gnuenterprise
*** dimas_ has quit IRC
*** jamest has left #gnuenterprise
*** sjc has quit IRC
*** holycow has joined #gnuenterprise
<holycow> hey all
*** jcater has quit IRC
*** jamest has joined #gnuenterprise
*** holycow has quit IRC
*** dcmwai has quit IRC
*** dcmwai has joined #gnuenterprise
*** jcater has joined #gnuenterprise
*** holycow has joined #gnuenterprise
*** jcater has quit IRC
*** jcater has joined #gnuenterprise
*** jamest has quit IRC
*** VeeOffice has joined #gnuenterprise
*** ajmitch_ has quit IRC
*** ajmitch has quit IRC
*** Vee2d2 has quit IRC
*** ajmitch has joined #gnuenterprise
*** ajmitch has quit IRC
*** ajmitch has joined #gnuenterprise
*** holycow has quit IRC
*** dcmwai has quit IRC
*** jcater has quit IRC
*** reinhard has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
*** dimas_ has joined #gnuenterprise
*** SachaS has joined #gnuenterprise
<SachaS> good morning
*** dimas_ has quit IRC
<SachaS> i guess there is no polymorphism in gnue business classes.
<SachaS> like there is a class "partner"
<SachaS> class "supplier" would be a subclass of "partner"
<SachaS> class "customer" would also be a subclass of "partner"
<SachaS> the class "contact" would have a reference to etiher a supplier or a customer. with polymorphism the contact could have a reference to a partner class, but when creating the contact instance it is assigned to eg a customer.
<johannesV> SachaS: no
<johannesV> there's no such a mechanism
<SachaS> ok. thanks.
<SachaS> did you have a look at the files I sent you?
<johannesV> yes i gave them a look, but it would take a bit more time to check them
<SachaS> guess I am at a stage where I could start to look at the code ;) the developer cannot keep up ;)
*** lekma has joined #gnuenterprise
<lekma> morning everyone
*** mixi^ has joined #gnuenterprise
*** mixi has quit IRC
* SachaS changes to lower case to temporarly avoid a problem
<johannesV> SachaS: your gcd's has way too long identifiers !
<johannesV> partner/forms/partner-de_CH.gld: PropertyNotFoundError: Die Klasse 'partner_contact' hat keine Eigenschaft 'ohterPhone1'
<SachaS> yeah i fixed that
<johannesV> timemgt/schemas/timemgt.gsd: UnicodeEncodeError: 'ascii' codec can't encode character u'\xfc' in position 20: ordinal not in range(128)
<SachaS> dah?
<johannesV> which class was not functioning well with camelcase ?
<ajmitch> hi
<SachaS> are you looking at the lastest one? i sent you two gzip files
* ajmitch tries to fix up his form so that it loads... :)
<SachaS> let me check
<ajmitch> ah, learning forms & reports in 24 hours - such fun :)
<SachaS> in partner/schemas/partner.gcd
<SachaS> all properties of type address_address
* ajmitch hopes people will be here for the next few hours to answer stupid questions ;)
<SachaS> if they are lowercase eg billingaddress it works
<SachaS> if they are camelcase billingAddress they fail
<johannesV> address_address works fine
<johannesV> it looks ugly (cause of missing gld i think) but it works
<SachaS> ok i got locale set to de_CH.UTF-8
<SachaS> so whats the name of the property in the partner.gcd file where the datatype is address_address?
<johannesV> ok, if i use ?language=de_CH it's using the proper gld-data
<johannesV> and it still works ?
<SachaS> i think you have the borked version
<SachaS> so in the partner.gcd file, do you have a customer with:
<SachaS> o billingaddress, or
<SachaS> o billingAddress ?
<SachaS> if it works, can you do a query & search
<johannesV> billingAddress
<johannesV> i've to leave for lunch now
<SachaS> ok so it fails here
<SachaS> ok
<SachaS> in my case the form comes up
<SachaS> but when i do a query & search it fails
<ajmitch> gah nasty
<ajmitch> DB000: _mysql_exceptions.ProgrammingError: (1064, "You have an error in your SQL syntax.  Check the manual that corresponds to your MySQL server version for the right syntax to use near 'contactID,deadline,date,id FROM orders WHERE ((1 = 0))' at line")
<ajmitch> mysql...
<SachaS> lunch time
* SachaLunch makes happy dance, he got another positive email
* ajmitch does the sad dance of pain :P
<ajmitch> much better...
<ajmitch> DB000:   File "/usr/lib/gnue/python/gnue/forms/uidrivers/_base/UIdriver.py", line 335, in updateEntry
<ajmitch> DB000:     handler = entry._displayHandler
<ajmitch> DB000: AttributeError: GFEntry instance has no attribute '_displayHandler'
<ajmitch> johannesV: can I blame you for this one? ;)
<ajmitch> forms 0.5.6
*** reinhard has quit IRC
<johannesV> ajmitch: i'll have a look
<johannesV> ajmitch: how can i reproduce that ?
<ajmitch> not sure, it just happens..
<ajmitch> I'm just fighting to try & get some forms & reports done asap
<johannesV> ajmitch: can you try the flollowing: remove all *.pyc files and try again
<ajmitch> doesn't matter
<ajmitch> my <field> entry had typecast=""
<johannesV> which kind of typecast ?
<ajmitch> forms & designer really can't handle forms with slight errors :)
<ajmitch> typecast was empty
<johannesV> ah, ok
<johannesV> i'll try that :)
<johannesV> i haven't ever tried with an empty typecast, caus i'm writing gfd's manually ... *lol*
<ajmitch> I can get designer to not start by missing something like 'masterlink' or 'detaillink' :)
<johannesV> ok, it's reproducable. if an empty typecast is given the error you've mentioned is raised
<ajmitch> yup
<johannesV> ok, found it
<ajmitch> fast work
<johannesV> but what would be the right thing to do
<johannesV> raise an exception like: sorry, but i cannot determine which displayhandler to use
<ajmitch> or use a default?
<johannesV> or just do nothing (very bad idea)
<johannesV> or ... right
<ajmitch> give a warning & use a default
<johannesV> use a default, but which one ?
<johannesV> maybe text
<ajmitch> string..
<ajmitch> ok, text
<johannesV> ajmitch: please svn-up :)
<ajmitch> I'm not using svn...
<ajmitch> especially as I'll be using this on a windows box :)
* ajmitch runs svn up anyway :)
<johannesV> so, how would you get the latest code then ?
<ajmitch> I report bugs, and use packages :)
<johannesV> hmm, ok, but i can treat this bug as fixed now
<ajmitch> ok
<ajmitch> johannesV: do you have any tips on how to use reports? :)
<johannesV> what kind of tips ?
<ajmitch> general howto..
<ajmitch> the current documentation is non-existent
<ajmitch> I'm wanting to do a basic printout from a forms trigger
<ajmitch> fill in some info on the form, hit a button & print out a report based on that
<johannesV> hmm, haven't tried something like that, but sounds like you would use parameters ?
*** Vee has quit IRC
*** Vee has joined #gnuenterprise
*** Vee has quit IRC
*** Vee has joined #gnuenterprise
<ajmitch> johannesV: that's nice to say that, but the lack of documentation gives me *nowhere* to start on that track :)
<johannesV> well, the only documentation available is the sourcecode; i haven't done much with reporting too
<johannesV> what kind of information would you like to change with the form ?
<johannesV> params for the query ?
<ajmitch> basically filling out an order form, so entering in a set of values - generating a printout to send on to the factory workers :)
<johannesV> and the data in the order form isn't stored in a db ?
<ajmitch> it is
<johannesV> hmm, so where's the problem ? calling the report from a trigger ?
<ajmitch> yes
<ajmitch> or the general overview of reports which is sadly lacking :P
<johannesV> i was looking at the sample-reports
<SachaS> johannesV: currently the appserver does not have a odbc interface, does it?
<johannesV> ?
<SachaS> so I could connect with a normal odbc client to it
<johannesV> don't mix rpc with odbc: you need an rpc client to connect to appserver, and you need an odbc-client to connect to a database via ODBC ...
*** mixi has joined #gnuenterprise
<SachaS> yeah i was asking if I could connect to the appserver via odbc
<SachaS> :)
<SachaS> guess not :)
<SachaS> not yet
<SachaS> i can access the db via odbc
<ajmitch> curses
* ajmitch kicks forms & designer
<ajmitch> they're both segfaulting now
<SachaS> anyways no problem
<SachaS> ajmitch i see you are about to create a report
<SachaS> :)
<SachaS> guess thats my next step too
<johannesV> SachaS: it does not make sense to connect to appserver via odbc ?!
<ajmitch> I'll *try* and make a report
<ajmitch> the documentation at the moment is appalling :)
*** mixi_ has joined #gnuenterprise
<SachaS> johannesV: reporting tools can connect to db's via odbc. so I was thinking why not access appserver via odbc for reporting
<SachaS> non gnue reporting tools I am talking about
<SachaS> eg i firedup openoffice and could do a report but against the backend database
<johannesV> so one would have to write an odbc-driver for appserver which is an interface between odbc and rpc
<ajmitch> that would be ugly
<SachaS> the odbc driver could use the rpc interface of appserver
<ajmitch> I'm not volunteering to write it! :)
<johannesV> nor do i
<SachaS> :)
<SachaS> haha
* ajmitch selects some nice loud music for his morning alarm & goes off to sleep
<SachaS> night ajmitch
<ajmitch> night
<johannesV> night ajmitch
<SachaS> johannesV: i think my application is basically working
<SachaS> all with generated forms
<SachaS> no master/detail but thats not necessary, you can just do that in a different form
<SachaS> so I can enter times I was working for a project.
<SachaS> next thing would be to generate a nice report
<johannesV> that's right
<SachaS> eg a list of work done in a time range (15 July til 15 August)
<johannesV> i think this is missing in the roadmap atm, but it was planned to build it once we have a bit more time
<SachaS> what is missing?
<johannesV> like appserver://appserver/report/module_class
<SachaS> auto generated reports?
<johannesV> which generates a basic report for that class (tabular or the like)
<SachaS> kick ass :)
<SachaS> how is your gnue time in the future?
<SachaS> can you continue to contribute as much as you could up til now?
<johannesV> i'm working fulltime gnue atm
<SachaS> it would be great that you would get some gnue paid work
<johannesV> this depends on how much money BYTEWISE Software likes to invest into gnue in the future
<SachaS> eg you could develop an application for a customer
<johannesV> at the moment i'm full paid
<johannesV> :)
<johannesV> well, that's another one, right. it would be much better to have a customer paying the bills :))
<SachaS> company's should get attracted by gnue.
*** mixi^ has quit IRC
<SachaS> maybe a company would be willing to do a prototype of some sort with gnue
<SachaS> that would be good, yes.
<SachaS> ok so I have to look at gnue reports again.
<SachaS> on my todays todo list I had an item: "think about migration". thats a hard one. eg if I start today to use my system to fill in real data.
<SachaS> eg make snapshot install from current cvs.
*** mixi has quit IRC
<SachaS> so I know that the application is still working tomorrow
<SachaS> and that I can keep developing on a new release of it
<SachaS> and one day ... migrate from that snapshot to the latest release.
<SachaS> same with packages: what if I change my packages in such a way that they are no longer compatible? => migration issues
* SachaS thinks maybe I skip the "how to migrate" todo item :)
<johannesV> well that are general design issues
<SachaS> concerning reports: its allways possible to get XML from gnue-appserver via gnue-report and THEN use a tool such as jasper (http://jasperreports.sourceforge.net/) reports which imports XML data.
<bigbrother> Title: JasperReports - Home (at jasperreports.sourceforge.net)
<SachaS> general design issues. I am sure you have gone through some in your professional career :)
<SachaS> eg migrating bytewise customers to new releases etc....
*** reinhard has joined #gnuenterprise
<johannesV> :)
<johannesV> reinhard: please give me a call if you've a bit time left
*** jamest has joined #gnuenterprise
* SachaS thinks auto generated reports would be cool.
<SachaS> new tools were release to the open source community
<SachaS> yesterday SuSE/Novell with their openxchange server (http://www.open-xchange.org)
<bigbrother> Title: OPEN-XCHANGE&#8482; : The Collaboration and Integration Server Environment (at www.open-xchange.org)
<SachaS> today Orbeon with a "integration suite", which includes an XML presentation server (with XForm (server side)) (http://www.orbeon.org)
<bigbrother> Title: Orbeon - Open Source Integration (at www.orbeon.org)
*** johannesV_ has joined #gnuenterprise
*** johannesV has quit IRC
*** btami has joined #gnuenterprise
<johannesV_> wohoo ...
<johannesV_> diggin into GBaseApp shows a lot dead and malfunctioning code ...
<wayneg> haha ... and that's a good thing how ? =)
<jamest> johannesV_: ?
<lekma> is it possible to use appserver over ssl?
*** jcater has joined #gnuenterprise
<jcater> w00t
<jcater> I was right on that website url problem yesterday
<lekma> hi jcater
<jcater> wrong channel
<jcater> howdy lekma
<johannesV_> jamest: you might have noticed the gnue.common.apps.errors file
<johannesV_> according to that new 'standard errors' i was trying to add a replacement of 'handleStartupError' to GBaseApp
<johannesV_> doing this with an sys.exceptionhook handling 'uncatched exceptions'
<johannesV_> we then can replace the handleStartupError calls with apropriate exceptions or (in most cases: we do not catch exceptions and call handleStartupError)
<johannesV_> that was why i was reading through all that code in GBaseApp and some other modules
<btami> hi all
<jamest> i didn't know we had dead code in baseapp
<btami> no logs from 2004-08-24
<johannesV_> getCommandLineParameters(self,paramList)
<johannesV_> doesn't get called (i think)
<johannesV_> or
<johannesV_> catching GConnection.InvalidFormatError is not necessary because it won't get raised
<johannesV_> (ConfigParser.ConfigParser instances won't raise DuplicateSectionError, or MissingSectionHeaderErrors at all)
<johannesV_> which has been filed as python-bug in Oct. 2003
<johannesV_> so instanciating GConnections.Connections () with an 'empty' connections.conf just works fine without an exception
<jcater> well
<johannesV_> or
<jcater> I never wanted handleStartupError to be replaced w/exceptions
<johannesV_> handling commandline-args like --debug-file= ...
<jcater> that seems backwards
<jcater> I like the idea of catching uncaught ones transparently
<jcater> but why would we eliminate the ability to call the handle* functions directly
<lekma> reinhard: u around??
<johannesV_> there's no need to do so. if it's enought to raise an exception
<johannesV_> we would need less code for catching exceptions and convert them into a handleStartupError
<johannesV_> which eliminates the risk of catching too much exceptions
<johannesV_> i think the point is in dividing errors into groups
<johannesV_> and adding an exception handler which is aware of that groups
<johannesV_> f.e. a UserError could be just a line saying something like 'hey, the file you requested isn't there'
<johannesV_> but if a SystemError occurs it will give a complete traceback, helping a dev. to find the bug
<lekma> reinhard: am i correct in saying, the _sessNo returned by a geasSessionManager.open can be used as the session_id for a geasSessionManager.call?
*** nickr has quit IRC
<lekma> reinhard: or should i call getSessionId() to know the session_id?
*** nickr has joined #gnuenterprise
<reinhard> lekma: the former
<lekma> reinhard: _sessNo = session_id, correct (sorry to bug)?
<lekma> reinhard: the code in geasSessionManager seems to say different
<lekma> reinhard: sorry, i was wrong
<lekma> forget it
<lekma> do i need svn version of forms to use appserver://... like url??
<lekma> is it possible to use appserver over ssl?
<lekma> a lot of question...
<SachaS> lekma, do you have the sample in gnue-appserver/samples working?
<lekma> yep samples work
<lekma> wait
<lekma> i just ran the setup-db.sh, this worked, do i need something else?
<SachaS> well then you have to start the appserver and then the client
<SachaS> see you later
*** SachaS has quit IRC
<lekma> [Errno url error] unknown url type: 'appserver'
<lekma> when trying gnue-forms appserver://gecocer/form/adress_country
<lekma> do i need svn version of forms to use appserver://... like url??
*** insanekane has joined #gnuenterprise
<insanekane> reinhard, are you busy atm ?
<insanekane> anyone here ?
<lekma> yep
<insanekane> lekma, im having problems runnning a form ... i designed the modules using classrepository.gfd ... and the form using gnuedesginer ... bu running the form, gives me an error that the relation is not on the database ...
<insanekane> but i have already made the classes using the classrepository form ...
<lekma> you should not use classrepository.gfd anymore
<insanekane> oh ...
<insanekane> then what should i use for making the classes ?
<lekma> create a gcd file (see in samples/)
<insanekane> oho ... ok
<insanekane> ok, after i create. then what should i do ?
<lekma> and then run gnue-readgcd --connection=gnue -u gnue -p gnue mymodule.gcd
<insanekane> ok ...
<lekma> change connection user and password according to your settings
<insanekane> well, i havent made any changes to settings ... so hacker/secret will work right ?
<lekma> and the tables, modules, classes... should be available in appserver
<lekma> no
<lekma> hacker/secret is for using with forms
<lekma> gnue/gnue is a db user (who should own the gnue db)
<insanekane> ok ...
*** bluesbaron_ has joined #gnuenterprise
<lekma> to test your appserver you can run setup-db.sh in samples
<insanekane> ok ... so after that i use the designer to make the form ... and in the data sources tree, i select appserver with hacker/secret to make the form ?
<insanekane> lekma, ok ... will do that
*** bluesbaron_ has left #gnuenterprise
<lekma> yep
<insanekane> lekma, are there any external resources like tutorials other than on gnue website, that i can refer ?
<lekma> insanekane: unfortunately no yhat i know
<lekma> but you always find some useful ressources in doc directories
<lekma> of each packages
<insanekane> hehe ... i checked them :)
<lekma> even if it's sometimes a bit outdated
*** btami has quit IRC
<insanekane> lekma, what is the difference between a bound procedure and a normal procedure ?
<lekma> good question, i think reinhard is the expert on this subject
<lekma> what i understand
<insanekane> lekma, ok
<lekma> is  bound procedures are a set of predefine procedures
<lekma> that are always called if they exist
<lekma> like OnInit, OnValidate....
<insanekane> aha
<lekma> but you should ask SachaS or reinhard, they know better
<insanekane> okie
<insanekane> lekma, thanks a lot for ur help :)
*** SachaS has joined #gnuenterprise
* SachaS is away: away
*** insanekane has quit IRC
<lekma> SachaS: after having setup appserver with samples, is there a form i can call with an appserver:// like url??
*** holycow has joined #gnuenterprise
*** jbailey has joined #gnuenterprise
<jbailey> Anyone seen the top article on /.? =)
<johannesV_> lekma: in principle you can call all classes using an appserver://appserver/form/classname style url
<johannesV_> but if you have provided a gld-file for some classes the generated form will look better
<johannesV_> btw: the appserver-url is constructed as follows "appserver://<connectionname>/form/<classname>[?debug-file=<local-file>[;formwidth=<x>][;formheight=<y>][;language=<lang>]]
<johannesV_> where <connectionname> is the name of the connection in connections.conf file defining appserver
<johannesV_> <classname> is the class of which you'd like to have a form for
<johannesV_> everything between [] is optional
<johannesV_> if a debug-file option is given the generated forms-code will be stored in the given 'local' file
<johannesV_> the other params determine maximum form with and height as well as the language to use like de_AT or en or en_US and so on
<johannesV_> ok, i'm off for today
*** johannesV_ has quit IRC
<reinhard> back
<reinhard> i think we have to remove gnue-classrepository.gfd
*** holycow has quit IRC
<dimas> reinhard: then what should we use in exchange?
<reinhard> gcd files
<dimas> ok
* reinhard beats reinhard
<reinhard> write docs! write docs!
<reinhard> sigh
<dimas> :)
<reinhard> it really would be about time
<reinhard> i think if i had written docs 2 weeks ago
<reinhard> i meanwhile would have saved so much time here
<reinhard> more than i would have needed to write the docs :)
<dimas> let's document it right now
<dimas> if i wrote gcd file then how would i transform it?
<dimas> into gsd?
<reinhard> gnue-readgcd -c <connectionname> <gcdfile>
*** gsoti_away has joined #gnuenterprise
<reinhard> that reads the gcd directly in your database
<dimas> and i can use gsscvs to extract gsd directly from database?
<dimas> err
<dimas> ggcvs
<reinhard> ggcvs or gnue-gsdgen, yes
<reinhard> but currently, gnue-gsdgen only exports data
<reinhard> not schema
<dimas> reinhard: thanks, so clear and simple :)
<reinhard> the gcd file is more or less the "source file" for a gnue package
<reinhard> night all
*** reinhard has quit IRC
*** gsoti_away has left #gnuenterprise
*** sjc has joined #gnuenterprise
*** insanekane has joined #gnuenterprise
<insanekane> hello all ....
<ajmitch> morning
<insanekane> im getting an error with using the setup.py script in the svn version ....
<insanekane> hello ajmitch
<insanekane> basically, the error is pygettext command not found ... but i am able to run pygettext command on the console without error
<insanekane> has anyone seen this problem before ?
<ajmitch> nope...
<insanekane> okie
<insanekane> the latest stable release doesnt have a gnue-readgcd (ie. 0.1.0) how come ?
<ajmitch> umm
<ajmitch> was it in 0.0.9?
<insanekane> no 0.1.0 ...
<insanekane> err ...
<insanekane> i dont know abt 0.0.9
<insanekane> is it new in the svn ?
<ajmitch> probably
<insanekane> hmm ok
<ajmitch> if it's not in the last stable release then it's most likely that it was added recently..
<insanekane> ajmitch, do you know how i could resolve this pygettext problem ?
<ajmitch> is it in your $PATH when you run setup.py?
<insanekane> ajmitch, this is the error i am getting when i issue ./setup.py build
<insanekane> running build
<insanekane> sh: line 1: pygettext: command not found
<insanekane> Could not find 'pygettext'. Strange.
<insanekane> It should be included in the Python distribution.
<insanekane> well ... i am able to issue a pygettext in the console and it doesnt give any error
<ajmitch> yes, and all that setup.py does is call pygettext on the system
<ajmitch> what does pygettext --version say?
<jcater> what does pygettext --version return?
* jcater goes back to his hole
<ajmitch> hi jcater
<ajmitch> just the man I wanted to hunt down
<ajmitch> you are the reports guru, I heard? :)
<insanekane> jcater, 1.4
<insanekane> pygettext.py (xgettext for Python) 1.4
<ajmitch> ok..
<jcater> ajmitch: yip
<jcater> so they say
<ajmitch> got any hints for someone wanting to do a simple printout of a report from a form?
<ajmitch> eg, user hits 'print this sheet' & it spits out something based on what they entered?
<jcater> lucky for you, I'm working in reports today
<ajmitch> yay
<insanekane> has gnue been used in production environments ? or is it still experimental ?
<ajmitch> it's been used in production
<insanekane> ajmitch, what kind of modules where used in it ?
<ajmitch> afaik forms & reports
<ajmitch> jcater can answer questions about that :)
<insanekane> no not gnue module ... i mean what kind of packages ? like inventory ? accounting etc ...
<ajmitch> I don't think any packages as such have been used in production
<ajmitch> just things done using the gnue tools
<ajmitch> jcater: quick tip on how to call reports from forms? :)
<jcater> give me a few minutes
<jcater> I'm adding a runReport atm
<ajmitch> ok
<ajmitch> thanks :)
<jcater> svn update and test out my runReport function
* jcater runs off to beat users
* ajmitch hands jcater a large bat
*** insanekane has quit IRC
*** holycow has joined #gnuenterprise
*** insanekane has joined #gnuenterprise
<insanekane> hello all
<insanekane> jcater, there is a problem running the sample.gpd file given with gnue-appserver ...
<jcater> I know nothing about gnue-appserver
<jcater> though I think gpd's are supposed to be run using gnue-navigator
<jcater> not gnue-appserver
<insanekane> jcater, no the problem is with navigator
<insanekane> the sample comes in gnue-appserver ...
<insanekane> anyway ... shall i paste the traceback ? it is wierd ...
<jcater> syre
<jcater> sure
<insanekane> DB000: Traceback (most recent call last):
<insanekane> DB000:   File "/usr/local/gnue/lib/python/gnue/navigator/UIwx.py", line 189, in OnTreeItemActivated
<insanekane> DB000:     object.run()
<insanekane> DB000:   File "/usr/local/gnue/lib/python/gnue/navigator/GNObjects.py", line 185, in run
<insanekane> DB000:     self._controller._clientHandlers[self.type](self, self._params)
<insanekane> DB000: TypeError: _runReport() takes exactly 2 arguments (3 given)
<insanekane> i tried printing self._params ... it is an empty dictionary ...
<insanekane> jcater, any ideas how to fix ?
<jcater> not without digging into it
<jcater> looks like it has to do with reports
<jcater> support in navigator
<insanekane> nope ...
<jcater> not appserver though
<insanekane> it fails before it reaches the code to do with reports
<jcater> it is calling _runReport() on the fail
<insanekane> well ... the call fails ...
<insanekane> and not because of any code in the _runReport()
<insanekane> but the exception is wierd ... since we are passing only 2 objects ... how does it think we are passing 3 ?
*** insanekane has left #gnuenterprise
*** insanekane has joined #gnuenterprise
<insanekane> erm ..
<jcater> insanekane: are you using svn to get our source?
<jcater> I fixed that one traceback
<insanekane> jcater, i already got the source
<insanekane> err ... im on dialup ... if the fix is minimal i will do the same changes here if you tell me how
<jcater> GNObjects.py:94
<jcater>  def _runReport(self, step, parameters={}):
<jcater> \
<jcater> (ignore that last \ line)
<insanekane> okie
<insanekane> so i have to add that attribute ?
<insanekane> i mean parameter ?
<insanekane> oh ok i get it ..
<insanekane> jcater, ok im getting one further error on params.update(self._params) ... like 130 of the same file
<insanekane> should i comment it out ?
<insanekane> or replace self._params with parameters ?
<jcater> I think so
<insanekane> ok i will try
<insanekane> jcater, ok i got the "merged" output on console but not in the GUI ... is that normal ?
<jcater> um, well, you are the first to get gnue-navigator working with reports
<jcater> I wouldn't say that's normal though
<jcater> but it hasn't been finished apparently
<insanekane> ok so noone else is trying to get it work ?
<insanekane> aha ok...
<jcater> well, I've been asked a few times to look at it
<jcater> but you are the first I didn't completely brush off
<insanekane> hehe i hope thats a good sign :)
<insanekane> jcater, what is the expected result of running a report in navigator ?
<insanekane> jcater, ok i got it output to an xml file ... so i guess it is upto the devloper to convert it into a format suitable for viewing correct ?
<jcater> sounds like you need to specify a filter
<jcater> I don't recall exactly how to do that with gnue-navigator
<jcater> but it might be something like "html"
<jcater> or "text"
<insanekane> aha ...
<insanekane> so the GUI for controlling all this is not yet made yes ?
<jcater> reports mostly has been used in "batch" mode
<jcater> so everything has been passed via command line
<jcater> stuff is in place for dialogs to prompt user for stuff like printers, etc
<jcater> just not finished
<jcater> I got hit with some major work projects
<jcater> so had to quit for a while
<insanekane> jcater, suppose i wanted to add a GUI for controlling this ... would I just modify UIwx.py or something else (for the wx backend ie) ?
<jcater> within the context of navigator-run reports, yes
<jcater> for command-line reports, we have a GRRunUI
<insanekane> jcater, ok got it
<lekma> does the master= masterlink= detaillink= syntax in datasources has changed???
<jcater> not in several years
<jcater> why?
<lekma> it doesn't work with an appserver connection...
<jcater> ah
<insanekane> jcater, btw, excellent :) it is working :)
<lekma> btw are import tags in forms repaired??
<insanekane> jcater, which is the preferred gui backend ? gtk2 or wx ?
<jcater> wx is the testbed/reference implementation
<jcater> and is what I use
<insanekane> ok good ;)
<lekma> jcater: how do you use fk_source and fk_key in a field tag??
<jcater> look at gnue-forms/samples/zipcode/zipcode.gfd for an example
<jcater>       <field name="state" field="state" fk_description="description"
<jcater>              fk_key="state" fk_source="validator">
<jcater> that says the current field (state) should be linked to the state field of the "validator" datasource
<jcater> but when displayed on screen (i.e., via a dropdown)
<jcater> to use "description" as the displayed text
<insanekane> jcater, thanks a lot for your help :) will bug you often :)
<insanekane> good night all
<holycow> dayumn, jc is here and i don't have the designer in front of me :)
<holycow> jcater, i was playing around with designer over the weekend, and its really a fantastic piece of work
<jcater> it has a long ways to go
<holycow> what i couldn't grok for example is how to create an input form with two text fields and input the data into a simple table, say first name last name
<holycow> is there anything in way of tutorial onthis?
<holycow> jcater, *nod* but i see where its going
<jcater> Form's Developer Guide
<holycow> its really going in the right direction
<holycow> oh okay lemme look at that
<holycow> oh is that the open office document you wrote?
<jcater> yeah
<holycow> that deals with an ntier setup with appserver tho... what about simple designer -> postgres?
<jcater> ?
<jcater> http://www.gnuenterprise.org/tools/forms/docs/
<bigbrother> Title: GNUe Forms Documentation - GNU Enterprise (at www.gnuenterprise.org)
<jcater> bigbrother: did I ask you anything?
<jcater> no, didn't think so
*** insanekane has quit IRC
*** jamest has left #gnuenterprise
<holycow> hehe
<holycow> k.
<holycow> yeah thats the document i read that
<holycow> one hell of a document, thats an amazing amount of work
<jcater> that shouldn't have anything to do with appserver
<holycow> as a newb i wasn't able to grok how to create a simple to input first and last name into a simple table however
<jcater> ah
<holycow> *nod* you are right, i'm mixing it up with a ton of other things i read, its kinda meleting into one at this stage :)
<holycow> thats not to say that the document should have any such newb related info
<holycow> i think the document is perfect for its intended use
<holycow> i know its gotta be super simple tho for example:
<holycow> i can successfully connect to postgres, which is on another machine
<holycow> i can create a db connection
<holycow> and a block
<holycow> then i create two fields, first name, last name
<holycow> and i think an enter key
<holycow> it's not clear how to link the fields to the table columns, and have the enter button update the db at first glance
<holycow> any tips on easy way to do it? i know i will haveto probably write a trigger of some sort
*** mpirun has joined #gnuenterprise
<jcater> ah
<jcater> easiest way currently is probably to have two blocks
<jcater> the first one only contain those two fields
<jcater> and it shouldn't be tied to a datasource
<jcater> the second should be tied to the datasource
<holycow> oh!
<jcater> then use an On-Enter trigger on the first block's last field
<jcater> to block2.initQuery()
<jcater> then set each of the corresponding 2 fields in the second block to the values in the first
<jcater> then block2.executeQuery()
* jcater is in a tad of a hurry
<jcater> so sorry for the crash coars4e
<mpirun> jeez...you guys really got as much done as your website suggests?
<holycow> mpirun, these guys are amazing
<holycow> when gnue starts to catch on a lot of companies are going to be pissed off
<holycow> jcater, no worries on the crash course!
<holycow> thats EXACTLY what i needed!
<holycow> awesome, awesome,thx, can't wait to get home.
<mpirun> before i get started, does this install fairly easily on rhel3, or should i start on something else?
<holycow> once i get my first simple form up i'm gonna make a mini tutorial
<holycow> weeeellll, not sure
<holycow> debian is by far the easiest to install anything on
<holycow> there are debian packages and you just do apt-get install gnue and your done
<holycow> however
<holycow> i have been using the latest svn downloads
<holycow> for that there you download into a dir from subversion repository
<holycow> and run some scripts to create symlinks and such... but i can't remember the steps for youright now
<holycow> third option is to compile/install on rhel, that shouyld be fairly standard as with any compiled/install proggy
<holycow> i think gnue ought to ship with a local db, particularly on windows, something like the newly open sources ibm db... cloudscape or whatever its called :)
<holycow> just for easy of 'getup and go'
*** jcater has quit IRC
*** jcater has joined #gnuenterprise
<ajmitch> holycow: of course debian is easy to install on - but you do still need to setup configuration in /etc/gnue
<ajmitch> I might get around to automating that with debconf when I use gnue a bit more
<ajmitch> I'd like to get into appserver
<holycow> oh that would be coo aj
<holycow> cool even
<holycow> i think the problem that you may run into is in that gnue is so flexible and powerfull
<holycow> you can have a local db, networked db
<holycow> you can have app server in both cases
<ajmitch> sure
<holycow> then you will have a newb situation ... i'm not sure whether that needs to be a consideration, but we will get questions from newbs about what all the steps mean
<holycow> most people don't know how to setup postgres, etc.
<ajmitch> but just something simple for reports filter setup, for example
<ajmitch> if they don't know how to setup postgres then they shouldn't be trying to use it
<holycow> lol :) thats what i mean
<ajmitch> anyway, I have to go
<ajmitch> bbl
<holycow> then the qeustion is, do you include something like cloudscape by default or .. ?
<holycow> k. later :)
*** gsoti_away has joined #gnuenterprise
*** lekma has left #gnuenterprise
*** jcater_ has joined #gnuenterprise
*** sjc has quit IRC
*** holycow has quit IRC
*** qman has joined #gnuenterprise
*** holycow has joined #gnuenterprise
*** holycow has quit IRC
*** holycow has joined #gnuenterprise
*** gsoti_away has quit IRC
*** jamest has joined #gnuenterprise
*** jamest has quit IRC
*** jamest has joined #gnuenterprise
*** egging12 has joined #gnuenterprise
*** chillywilly has quit IRC
*** egging12 has quit IRC
<qman> hrm..installed to /opt, and it made a /etc/opt/gnue...that correct?
*** jcater_ has quit IRC
*** jamest has quit IRC
*** jcater_ has joined #gnuenterprise
*** holycow has quit IRC
*** holycow has joined #gnuenterprise
*** SachaS has quit IRC
*** lekma has joined #gnuenterprise
*** holycow has quit IRC
*** lekma has left #gnuenterprise
*** chillywilly has joined #gnuenterprise
*** chillywilly has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
<dimas> hi all
<dimas> johannesV: is there a cvs equiv of the gnue-readgcd script?
*** vergil has joined #gnuenterprise
<johannesV> dimas: gcdcvs
<johannesV> and for gnue-readgld it's gldcvs
*** lekma has joined #gnuenterprise
*** SachaS has joined #gnuenterprise
*** SachaS has left #gnuenterprise
*** johannesV has quit IRC
*** SachaS has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
<SachaS> hi johannesV
*** johannesV_ has joined #gnuenterprise
<SachaS> hi johannesV_
<johannesV_> hi sacha__
<SachaS> ;)
<SachaS> johannesV_ do you think its easy to have condition of a datasource be changeable in a form?
<SachaS> i think at the moment once the form is loaded, the condition of a datasource is fix.
<SachaS> jamest once moved it into the trigger namespace
<SachaS> but after editing the condition in a trigger it did not affect the query.
<johannesV_> SachaS: i don't know
<SachaS> ok. thanks.
<johannesV_> i assume using createResultSet () or simpleQuery () is not what you want ?
<SachaS> i think even with simpleQuery it did not work.
<SachaS> if i remember right
<johannesV_> what problem would you like to solve ?
<SachaS> ok.
<SachaS> Arbeitsrapport
<SachaS> so i have a table with work items each with a start dateTime and with a stop dateTime
<johannesV_> ok
<SachaS> so i want to have a way to let the user chose: query all work items with start >= 15.08.2004 and stop <=31.08.2004
<SachaS> i was thinking this is typically a condition of the datasource
<johannesV_> and what does happen after selecting the date range ?
<SachaS> i mean i do not have the date range fix coded in the form
<johannesV_> sure, but will the result be displayed in the form or printed in a report ?
<johannesV_> or something else ?
<SachaS> ah
<SachaS> was thinking in the form
<SachaS> i am at the verge
<SachaS> where shall i design a nice form
<johannesV_> so i think this is not a big thing and i think it should work
<SachaS> where shall I design a nice report (html,xml etc)
<SachaS> ok.
<SachaS> thanks
<SachaS> have to play with it.
<SachaS> s/where/when
<johannesV_> i think using an unbound block for the selection fields and a button with a trigger for populating the data-block will do the job
<johannesV_> but i'll try something similair; just a sec
<SachaS> actually i am not sure when to build a form with explicit the main gnue form navigation bar
<SachaS> and when to build a form with extra buttons with extra query's and populating things etc.
<johannesV_> what's an explicit gnue form navigation bar ?
*** johannesV has quit IRC
<johannesV_> SachaS: using a simpleQuery works fine
<johannesV_> so your problem is solveable without any specials or tricks
<SachaS> can you past your simpleQuery call, plase?
<SachaS> ok no special tricks :)
<SachaS> what are you currently working on?
<johannesV_> exceptions and then gnue-messages
<johannesV_> well you have a block which has no datasource bound to it, say it's called "blkSelection"
<johannesV_> this block has two fields, fldStart and fldEnd
<johannesV_> then you have a block "blkItems" which is bound to a datasource "dtsItems"
<johannesV_> in a button's trigger-code you can say something like:
<johannesV_> dtsItems.simpleQuery ({'start': blkSelection.fldStart.get (), 'stop': blkSelection.fldStop.get ()})
<johannesV_> and that's it
<johannesV_> so enter both dates and press the button :)
<SachaS> content of {} is the condition, right?
<johannesV_> Note: you would have to check if there are other conditions on that datasource
<johannesV_> right
<SachaS> how about the >= ?
<johannesV_> each key is a db-field and each value the corresponding value; condition is 'EQ'
<johannesV_> multiple conditions are concatenated using AND
<johannesV_> oh, damn
<johannesV_> right
<johannesV_> how silly i am ...
<johannesV_> you'd like to have a range ...
<johannesV_> hmm
<johannesV_> so we would have to use the createREsultSet ()
<johannesV_> i'll check again
<SachaS> something like:   {'start': LQ blkSelection.fldStart.get(), ...} with LQ being like less or equal to?
<SachaS> jamest thought having condition in the trigger namespace would solve it:
<SachaS> eg dtsItems.condition.set( ... )
<SachaS> and then run the query on it
<SachaS> oh i think to remember when I was asking jamest about that it was rather the question to change the order properties of a datasource.
<johannesV_> wow, just found a bug
<johannesV_> in datasources.drivers.dbsig2.dataobject
<johannesV_> hmm
<johannesV_> SachaS: you would have to create a condition using something like GConditions.buildTreeFromPrefix ()
*** MiXi has joined #gnuenterprise
*** mixi_ has quit IRC
<johannesV_> ok, SachaS, now i got it
<johannesV_> you can do it like:
<johannesV_> from gnue.common.datasources.GConditions import buildTreeFromList
<johannesV_> cond = buildTreeFromList (['and', ['ge', ['field', 'start'], ['const', blkSelection.fldStart.get ()]],
<johannesV_> ['le', ['field', 'stop'], ['const', blkSelection.fldStop.get ()]]]]
<johannesV_> )
<johannesV_> dtsItems.createResultSet (cond)
<johannesV_> that's it
<johannesV_> looks a bit ugly here but it's ok
<johannesV_> :)
* SachaS shakes head
<SachaS> :)
<SachaS> johannesV_ why is setCondition not working?
<SachaS> ;)
<SachaS> in GDataSource.py
<SachaS> dont reply :)
<SachaS> ;)
<SachaS> the setCondition is in there but is not working
<SachaS> guess the idea was to have it working in the trigger namespace
<SachaS> just browing through GDataSource.py
<SachaS> seems like dtsItems is an instance of a GDataSource
<johannesV_> the problem is not setCondition
<johannesV_> the problem is to give a 'usable' condition instance to that object
<johannesV_> and this condition has to be created (i.e. from buildTreeFromList ())
<SachaS> hmmm
<SachaS> ok gotcha
<SachaS> so maybe setCondition(cond) would also work (cond from above)
<johannesV_> maybe
<SachaS> johannesV_ what is the difference between a trigger calling createResultSet(cond) and the user pressing search query ?
<SachaS> does gnue-forms when the user sets values while in search mode create the condition accordingly?
<SachaS> as you have shown above?
<SachaS> ok damn :) dtsItem.simpleQuery( X ) where X is the condition as a dictonary, where the name value are equal conditions.
<dimas> johannesV_: is that enough to load certain gcd's to initialize new appserver db?
<SachaS> simpleQuery( X ) sets the equal condition in line 149
<SachaS> ok
<SachaS> thats why a simpleQuery does not work
<SachaS> hi dimas
<dimas> hello SachaS
<SachaS> simpleQuery( X ) sets the equal condition (GConditions.GCLike) in line 149
<johannesV_> dimas: loading a gcd updates gnue_* tables AND creates the physical tables for the new classes
<johannesV_> so it's enought
<johannesV_> but you might have to restart appserver to know the new classes
<SachaS> dataObject is obscure :)
<johannesV_> SachaS: ?
<SachaS> johannesV_: dataObject = Connection.supportedDataObjects['object']
<SachaS> do not grasp what DataObjects are; not to worry
<SachaS> relates to the underlying db
<dimas> johannesV_: DB000: psycopg.ProgrammingError: ERROR:  Relation "gnue_module" does not exist
<dimas> like i messed something
<dimas> :)
<dimas> messed and missed
<dimas> johannesV_: loading a gcd updates gnue_* *updates*, so they should exist before?
<dimas> how to create gnue_* ?
<SachaS> dimas I think you need to load the base.gsd file (that in the gnue-appserver/samples
<SachaS> further you have to load the auth.gcd file
<SachaS> further you need a auth.gsd file which does not exist but was moved into the sample.gsd file
<SachaS> johannesV_: maybe the tabledata of table gnue_useraccess should go into a auth.gsd
<SachaS> dimas loading base.gsd, auth.gcd and your manually created auth.gsd (take info from sample.gsd gnue_useraccess part)
<SachaS> its puzzling why base.gsd and auth.gcd plus user access data is in the sample.gsd is actually in the "sample" directory as they are necessary for every application.
<dimas> yes
<dimas> base.gsd is loaded
<johannesV_> dimas: in appserver/samples run setup-db.sh or setup-cvs-db.sh
<johannesV_> SachaS: no auth.gsd this is included in auth.gcd !
<johannesV_> it is NOT in sample.gsd
<dimas> johannesV_: thanks
<SachaS> johannesV_: user name and password is in sample.gsd and not auth.gsd
<johannesV_> no, auth.gcd defines the class gnue_useraccess
<SachaS> yes
<johannesV_> which updates gnue_*
<SachaS> yes
<SachaS> why is there no auth.gsd ?
<johannesV_> the sample-account is of course introduced by sample.gsd
<SachaS> but that data is in samples.gsd ?
<johannesV_> because it is a *sample* account
<SachaS> :)
<SachaS> ok
<johannesV_> :)
<SachaS> but does not help much for a beginner
<johannesV_> well
<SachaS> guess its always the same
<SachaS> default users is bad :)
<SachaS> so maybe that why there is no auth.gsd :)
<johannesV_> no, auth.gcd defines the class gnue_useraccess
<johannesV_> which has nothing to do with it's data
<johannesV_> anyway, that authentication class is just a hack
<SachaS> but does not the appserver NEED gnue_useraccess ?
<johannesV_> it will be replaced by somewhat better mechanism
<johannesV_> right, atm it needs it, but it's too simple for production use (in real life)
* dimas will read the backlog later
<SachaS> openldap caugh caugh ;)
* dimas is off to client
<SachaS> gnue supports single sign on ;)
<SachaS> anyways
<SachaS> i will shut up again
<SachaS> :)
<johannesV_> np
<johannesV_> i've to leave for a short while anyway
<SachaS> cool.
<SachaS> thanks today I looked at gnue source code
<SachaS> guess I have to learn what {}, [] in python are :)
<SachaS> actually I think its crazy if you can import gnue stuff in a trigger
<SachaS> guess you can realy screw up the gnue-forms viewer :)
<SachaS> did I say I shut up ?
<VeeOffice> There's a new upgrade script on ash, ~vin/upgrade_bro, to upgrade brother with some bug fixes and fix the log rotation cron script (logs will re-appear at midnight, like a backassward cinderella story?). Just needs someone with root@ash to run it
*** jamest has joined #gnuenterprise
*** btami has joined #gnuenterprise
<btami> hi all
<btami> anyone can restart logs please ?
<VeeOffice> @misc last --with upgrade
<bigbrother> VeeOffice: <VeeOffice> There's a new upgrade script on ash, ~vin/upgrade_bro, to upgrade brother with some bug fixes and fix the log rotation cron script (logs will re-appear at midnight, like a backassward cinderella story?). Just needs someone with root@ash to run it
<VeeOffice> :)
*** insanekane has joined #gnuenterprise
<insanekane> lekma, hello are you there ?
<insanekane> anyone here ?
<btami> VeeOffice: thx
<btami> jamest/jcater ???
<btami> insanekane: hi
<insanekane> btami, hi :)
<insanekane> btami, have you used gnue-designer ?
<btami> a bit
<insanekane> btami, i made a form but not able to get it to work ...
<insanekane> btami, actually, it is an extremely simple form ...
<btami> what is the problem?
<insanekane> btami, i first made my module and class ... the class has 2 attributes: name, and sign ... so i made the .gcd file, and used gnue-readgcd to setup the db ... i checked it manually using psql, and it was made correctly ....
<insanekane> btami, then using gnue-designer, i started an empty form, and dragged the 2 properties onto the form and saved it ...
<insanekane> btami, but, when running the form ... it opens the form correctly, but i cant add any instances of my class ...
<btami> any error messages?
<insanekane> well it says "NoneType object doesnt have any attribute gnue-name"
<insanekane> btami, wait .. i will get u a traceback ..
<btami> wait
<btami> is your appserver set up correct?
<btami> you made it with setup-db.sh ?
<insanekane> btami, err um ... its working now .... :) ...
<insanekane> btami, yeah i used setup-db.sh
<btami> hmm
<insanekane> its working now ...
<insanekane> i wonder if it is because i restarted appserver ...
<btami> yes, IMO
<insanekane> btami, do you know if instances of a class can be given its own implementation of a procedure ? or are procedures shared for all instances of a class ?
<btami> i think the latter
<insanekane> btami, ok stupid question i guess ... heres a better one ... how do i make a one to many relationship between 2 class ?
<insanekane> in .gcd and in the form ....
<btami> insanekane: i'm not up to date on .gcd and appserver terminologi, but
<btami> you can try designer wizard first
<btami> to produce one2many forms
<insanekane> desginer has wizards ?
<btami> file-new-wizard
<btami> sorry, have to go now
<btami> bbl
<insanekane> btami-swpool, thanks :)
*** jamest has quit IRC
*** jamest has joined #gnuenterprise
*** bigbrother has joined #gnuenterprise
*** bigbrother has joined #gnuenterprise
*** insanekane has quit IRC
<SachaS> insanekane one to many relation:
<SachaS> an company has many projects
<SachaS> a project belongs to one company
<SachaS> <class name="company"> <property name="name" type="string(30) /> </class>
<SachaS> <class name="project"> <property name="name" type="string(30)"/> <property name="belongsToCompany" type="company"/></class>
<SachaS> i omitted the module name so you have to add that to the type of the project class eg if module where class company is defined is called module X then you have to set <property name="belongToCompany" type="X_copmany"/>
*** insanekane has joined #gnuenterprise
<SachaS> <SachaS> insanekane one to many relation:
<SachaS> <SachaS> an company has many projects
<SachaS> <SachaS> a project belongs to one company
<SachaS> <SachaS> <class name="company"> <property name="name" type="string(30) /> </class>
<SachaS> <SachaS> <class name="project"> <property name="name" type="string(30)"/> <property name="belongsToCompany" type="company"/></class>
<SachaS> <SachaS> i omitted the module name so you have to add that to the type of the project class eg if module where class company is defined is called module X then you have to set <property name="belongToCompany" type="X_copmany"/>
<insanekane> SachaS, oh great :) ... but what about navigating to the projects from the company ?
<insanekane> SachaS, btw, can I get the source for the appserver_howto file ?
<insanekane> SachaS, are you busy ?
<insanekane> ermm ... stupid question sorry
<SachaS> the appserver howto is outdated
<insanekane> SachaS, yes i know :) thats why it would be good to have it :)
<insanekane> SachaS, most of it is usable i think ...
<SachaS> the classrepository.gfd will even be removed, i think
* SachaS sent appserver howto to insanekane
<insanekane> SachaS, yeah i know ...
<insanekane> erm ... eyah i know you sent it :
<insanekane> SachaS, writing a little wxPython program to make a gcd file ...
<SachaS> u are ?
<insanekane> SachaS, yes ... just a little utility ...
<insanekane> or is gcd also going to the trash bin ?
<SachaS> haha not as far as I know, I hope not
<SachaS> i am building upon .gcd's
<insanekane> :)
<insanekane> SachaS, if you are free ... can you explain to me how I would build a form for the above company-project example, where i can see the company name, and under it a grid of all the projects it is involved in ?
<johannesV_> insanekane: what about writing a gtk2 program for creating gcd ?
<johannesV_> or what about vim ?
<insanekane> umm ... cuz i dont know gtk2, or vim ...
<johannesV_> am i right if you don't like pure xml ? :))
<insanekane> moreover, ive used wx a lot so i am very familiar with it and so it will be very easy to make the "utility" ...
<insanekane> johannesV_, whats not to like in XML ?
<insanekane> johannesV_, pure XML ? is there any other kind ?
<insanekane> ;)
<johannesV_> well if you create a wx program this one will create the gcd (which is xml)
<insanekane> johannesV_, yes, i am hoping to output the gcd as an xml ... and if possible to directly push it into the database, like readgcd does ...
<SachaS> gcd is xml
<insanekane> SachaS, yes i know its XML :)
<johannesV_> hmm, so it would be better if readgcd has an option for reading gcd from standard-in then
<johannesV_> to avoid duplicating that part
<insanekane> johannesV_, umm i guess so
<insanekane> johannesV_, have i seen u somewhere else ? like #scribus
<johannesV_> insanekane: don't know :)
<insanekane> johannesV_, do you come to #scribus ?
<johannesV_> no, i don't know scribus
<insanekane> johannesV_, ok just asking :)
<johannesV_> np
<insanekane> johannesV_, when i meant pushing to the database ... i was actually thinking of duplicating the readgcd ... but now i think u have given a better option :)
*** jamest has quit IRC
<johannesV_> :)
<insanekane> johannesV_, but that will require a change to the current readgcd as it is ? ie, does it support reading the gcd from the stdin currently ?
<johannesV_> no it doesn't by now, but i could change it
<johannesV_> :)
*** jamest has joined #gnuenterprise
<insanekane> johannesV_, oho ...
<insanekane> johannesV_, ok, i guess ur changes will happen way before i complete the utility :)
<insanekane> ok good night all ...
*** insanekane has quit IRC
<SachaS> bye
*** SachaS has quit IRC
*** sacha__ has quit IRC
*** jamest has quit IRC
*** jamest has joined #gnuenterprise
*** SachaS has joined #gnuenterprise
* SachaS is away: away
*** reinhard has joined #gnuenterprise
*** reinhard has quit IRC
*** jcater_ has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
*** jcater_ has quit IRC
<dimas> i'm getting:
<dimas> ------------------------------------------------------------
<dimas> Error: Unable to login to datasource.
<dimas>        <ProtocolError for localhost:8765/: 500 Internal error>
<dimas> ------------------------------------------------------------
<dimas> where to check what cause this?
<dimas> on the appserver side it's traced with:
<dimas> DB001: 0:00:56.696 [ServerAdapter:452] Unexpected Exception in XML-RPC code: exceptions.TypeError:isinstance() arg 2 must be a class, type, or tuple of classes and types
<dimas> DB001: 0:00:56.698 [ServerAdapter:406] exodus.vbc.elcom.ru - - [01/Sep/2004 22:32:20] "POST / HTTP/1.0" 500 -
<chillywilly> you broke it
<reinhard> local or remote appserver?
<dimas> local
<dimas> chillywilly: i'm innocent :)
<chillywilly> neigh!
<chillywilly> :P
<reinhard> which python version?
*** lekma has quit IRC
<reinhard> and do you use pw_xmlrpc or py_xmlrpc?
<dimas> 2.2
<dimas> and it set to plain xmlrpc by default
<dimas> checking for py_xmlrpc
*** sjc has joined #gnuenterprise
<reinhard> most probably you use xmlrpclib (that is pw_xmlrpc)
<reinhard> can you please try
<reinhard> python
<reinhard> >>> import xmlrpclib
<reinhard> >>> print repr (xmlrpclib.Fault)
<dimas> Python 2.2.2 (#1, Nov  6 2002, 18:18:06)
<dimas> [GCC 3.2.1 20020903 (ALT Linux, build 3.2.1-alt0.4)] on linux2
<dimas> Type "help", "copyright", "credits" or "license" for more information.
<dimas> >>> import xmlrpclib
<dimas> >>> print repr (xmlrpclib.Fault)
<dimas> <class xmlrpclib.Fault at 0x81b1d44>
<reinhard> are you using a release or svn version?
<dimas> svn
<reinhard> will you be able to be here tomorrow
<reinhard> I think johannesV_ will be able to help you
<reinhard> as he is fiddling around with exceptions currently
<reinhard> and I was quite offline for 2 days now and haven't followed too closely
<dimas> reinhard: sure, thanks
<dimas> it just worked with xmlrpc.py_xmlrpc
<mpirun> do any of the applications have releases yet for download?
<dimas> seems like client still talks via pwxmlrpc and throws exception on commit
<dimas> changed setting in connections.conf to py_xmlrpc also and it worked
<dimas> mpirun: not as complete business packages
<reinhard> so your server runs with py_xmlrpc
<reinhard> and your client did run with pw_xmlrpc
<reinhard> and they don't like each other
<dimas> yes
<dimas> and i set correct value in conn...conf already< thanks
*** bluesbaron_ has joined #gnuenterprise
*** johannesV__ has joined #gnuenterprise
*** bluesbaron_ has left #gnuenterprise
<mpirun> to start with, i'd like ot build a basic CRM for sales contacts..important, but not exactly critical...should i start with the releases, or pull current?
<dimas> i would suggest current
*** johannesV_ has quit IRC
<mpirun> cool..thanks.
<reinhard> mpirun: you might want to look at gnue-packages
<reinhard> contains some pre-alpha stuff
<reinhard> for packages
*** johannesV__ has quit IRC
*** btami-swpool has quit IRC
*** reinhard has quit IRC
*** sjc has quit IRC
*** jamest has quit IRC
*** jbailey has quit IRC
*** jamest has joined #gnuenterprise
*** Vee has quit IRC
*** Vee has joined #gnuenterprise
*** wt has joined #gnuenterprise
<wt> hola, gnue people
<wt> anyone here interested in helping?
<wt> I am looking for how to install the gnue packages
*** jcater_ has joined #gnuenterprise
<wt> hi
<wt> you look like you are from my neck of the woods
<wt> it is you
<wt> jcater, whuzzup?
*** jamest has quit IRC
*** jamest has joined #gnuenterprise
*** Amorphous has joined #gnuenterprise
<jcater_> wt: howdy
*** jcater has quit IRC
*** jcater_ has joined #gnuenterprise
*** MiXi has quit IRC
*** mixi has joined #gnuenterprise
<wt> jcater, hi'
<wt> hep me
<wt> guess I missed out
<wt> boo hoo :((
<jcater> wt: what's up?
*** jamest has quit IRC
<wt> jcater, ?
<wt> everything
<wt> everything is up
<wt> you never emailed me back about the job posting, btw
<wt> I wasn't sure how you wanted to proceed with that
<wt> does anyone know what the appserver is good for?
<wt> I am guessing nothing
<wt> hehe
<wt> I freaking love acl
<wt> s
<wt> jcater, what is adrius?
<jcater> a friend's game
<jcater> not very far
<wt> ah
<jcater> wt: ah
<wt> I am thinking of using gnue
<jcater> I did email you back
<wt> should I?
<jcater> but it got bounced
<jcater> and I never resent it
<wt> wt@midsouth.rr.com?
<jcater> yeah, but I sent from home
<jcater> and rr doens't like email servers from home
<wt> crazy
<wt> yeah
<wt> well...I have gnue's connections.conf working
<jcater> we've usually used college students to fill this position
<wt> you consider interns?
<jcater> I did contact those two professors last week
<jcater> no response
* jcater is a former student of theirs too, so kinda odd
<wt> yanu is a <nasty word>
<wt> ewelch will eventually get back to you
<jcater> he just takes some getting used to :)
<wt> I had 4 years
<wt> it must take longer
<wt> okay, but down to business
<wt> are there any erp packages for gnue?
<wt> even the Base?
<wt> and if not, do you do for hire work for gnue?
<jcater> we are mainly tools now
<jcater> though I have seen some effort going into the base
<jcater> I don't do for hire work
<wt> where would I get that schema file?
<jcater> I'm already too swamped
<jcater> (hence the need for help :)
<wt> figured
<wt> just wasn't sure
<jcater> um
<jcater> SachaS: were you working on the base?
<wt> I am gonna be starting a business real soon now
<jcater> it would be in our svn tree, I'm sure
<wt> I want to use gnue for the backend
<jcater> under gnue-packages/
<wt> is there a web interface?
<wt> and does schema calculate ddl diffs now?
<jcater> not sure about the diff support yet
* jcater has been so busy w/other stuff for the last 6 months
<jcater> I'm just now getting back up to speed with
<wt> ah
<jcater> what's been happening
<wt> I think the designer is pretty cool, btw
<wt> whoever's baby that is, it is looking good
<jcater> thanks
<jcater> plus, you should know my interest/effort doesn't extend beyond the 2-tier tools
<wt> Ideally, I like a zope forms client
<wt> but that is pie in the sky
<jcater> i.e., not into appserver oor the applications
<wt> 2 tier is fine for me
<jcater> I used zope for a long while, but am moving off of it now
<wt> what are you going to?
<wt> I hate PHP
*** Amorphous has quit IRC
<jcater> mod_python-type stuff
<jcater> my two bitches w/zope
<jcater> zodb and
<jcater> well
<jcater> zodb
<wt> what's wrong with it?
<wt> api?
*** Amorphous has joined #gnuenterprise
<jcater> can't use source repositories
<jcater> for one
<jcater> so it doesn't play nice with all the other stuff I have
<wt> like svn?
<jcater> and I have trouble w/zodb bloating up when hit hard on my commerce site
<jcater> yeah, svn
<jcater> so I have to clean it up each night
<wt> ah
<jcater> i.e., it will go from 20Mb to over 2Gb
<wt> new versions fixing that?
<jcater> because of crappy session handling
<wt> would that not be a zope problem instaed of zodb?
<jcater> dunno
<wt> anyhow
<wt> gnue
<jcater> I'm moving to pure-python stuff now
<wt> since that is this channel
<jcater> bah
<jcater> someone who wants to stay on topic
<jcater> that's not the #gnue way
<wt> why is the docs dir f'ed on the website?
<jcater> oops
<jcater> need to fix that
<jcater> that is old cruft
<jcater> docs are under each tool's doc/ structure now
<jcater> i.e.
<wt> other projects is also f'ed
<jcater> http://www.gnuenterprise.org/tools/forms/docs/
<bigbrother> Title: GNUe Forms Documentation - GNU Enterprise (at www.gnuenterprise.org)
<jcater> bigbrother: stfu
<jcater> I hope to finish the website in the next month or so
<wt> okay
<wt> so I kinda understand some of the concepts
<wt> that is an improvement
<wt> the websvn is f'ed
<wt> this is kinda disheartening
<jcater> sigh
<jcater> yeah
<jcater> our svn got 0wned not too long ago
<jcater> so we had to do an emergency source install
<wt> is the code safe?
<jcater> yeah
<jcater> just someone using us to relay into other machines
<jcater> we checked though
<jcater> but our main server is debian
<wt> bad ass
<jcater> and debian + source installs
<jcater> aren't nice together
<wt> nope
<jcater> I haven't fixed websvn as a result
<wt> that's why everything is packaged for debian
<jcater> yeah, but a few things don't play nice in debian as packages
<jcater> svn being one of them
<wt> is it not just a matter of installing a newer version?
*** Amorphous has quit IRC
<jcater> we almost lost our repository on numerous occasions, and moving from debs to source stopped that behavior altogehter
<jcater> wt: because we are a popular project, we got hit w/a 0-day exploit
<wt> I have never lost an svn repo
<jcater> so no debs available
<wt> ah
<wt> hmmm
<wt> wonder how I would load a package?
<wt> I am guessing these are forms and other stuff
<wt> anyone care to throw me a bone?
*** jcater has quit IRC
<wt> no one to talk to
<wt> sad is me
*** chillywilly has quit IRC
*** dimas has quit IRC
*** chillywilly has joined #gnuenterprise
*** dimas has joined #gnuenterprise