*** dimas has joined #gnuenterprise
*** holycow has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
*** sjc has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
<johannesV> good morning
<dimas> good morning
*** dimas has quit IRC
*** dimas has joined #gnuenterprise
<reinhard> good morning all
<dimas> hello reinhard
<johannesV> morning reinhard
*** dcmwai has quit IRC
*** kilo_ has joined #gnuenterprise
<SachaS> good morning
<kilo_> hi SachaS
*** btami has joined #gnuenterprise
*** mnemoc has joined #gnuenterprise
*** mnemoc_ has quit IRC
<btami> bbl
*** btami has quit IRC
*** dcmwai has joined #gnuenterprise
<tiredbones> Good Morning! For those of you that use Designer do you use the trigger facilities?
<tiredbones> Or do you do it with a editor?
<tiredbones> That is, do you use an editor to code your trigger or use the built in stuff that Designer offers?
<kilo_> outer editor
*** dimas has quit IRC
*** dcmwai has quit IRC
*** sjc has quit IRC
*** kilo_ has quit IRC
*** jamest has joined #gnuenterprise
*** holycow has quit IRC
*** johannesV has quit IRC
*** johannesV has joined #gnuenterprise
*** jcater has joined #gnuenterprise
<reinhard> I have a question on dictionary conditions
<reinhard> currently you have to tell buildConditionFromDict whether you want "LIKE" or "=" conditions
<reinhard> what would you think about autodeciding that depending on whether the string contains a "%" or "_"?
<jcater> when you mean autodecide
<jcater> do you mean, as a whole and not have the parameter that specifies it
<jcater> or keep the parameter, but check each field and if 'LIKE' is to be used, only use LIKE if % or _ is included,  but = otherwise?
<reinhard> that's what I would want to find out here?
<reinhard> s/?/:)/
*** titopbs has joined #gnuenterprise
*** dimas has joined #gnuenterprise
*** titopbs has quit IRC
<reinhard> jcater: would any of these 2 alternatives make sense to you?
<jcater> I like the latter
<jcater> i.e., you specify that you are after "LIKE"s
<jcater> but include intelligence on whether like or = is actually used
<reinhard> ok
<reinhard> and you would keep the default of GCeq ?
<jcater> but some tools may need all strings to be '='
<jcater> I'm indifferent
<jcater> though I don't know what all code is out there assuming that default already
<reinhard> I think you're right
*** dcmwai has joined #gnuenterprise
<jcater> reinhard, is object._parent support completely gone now?
<jcater> or are there deprecation hooks to keep old code working?
<jcater> I just need to know before I svn update
<reinhard> you have non-gnue code that uses _parent?
<jcater> yes
<reinhard> it is gone
<jcater> ok
<reinhard> you should now use setParent and getParent
<jcater> is _children still there?
<reinhard> yes
<jcater> ok
<jcater> I use GObject trees for several internal projects
<reinhard> ok
<reinhard> you use _parent a lot there?
*** titopbs has joined #gnuenterprise
*** jcater_ has quit IRC
<jcater> not a lot
<jcater> but inside my GObject subclasses
<chillywilly> hi
<jcater> howdy
<SachaS> aren't blocks independent?
<SachaS> i have two blocks in a form
<SachaS> but they relate to each other.
<SachaS> they both use the same datasource ...
<reinhard> then one block will follow what the other does
<SachaS> and with different datasource?
<reinhard> they will be independent
<SachaS> ok. thanks.
<SachaS> will try.
<SachaS> odd behaviour: if I use style="dropdown" of the entry for the only filed  of one block I get a traceback
<SachaS> if I do not use a style attribute it is working as text box
<reinhard> you may not use a block for the lookup datasource
<SachaS> what do you mean ?
<SachaS> in the block I do reference a datasource
<reinhard> you have 2 datasources, right?
<reinhard> one with your main data
<reinhard> and one where the dropdown would get it's contents
<SachaS> one block, which uses several datasources: (project_todolist, project_state, project_priority, project_project)
<SachaS> second block only one datasource: project_project
<SachaS> with the first block the application is working
<SachaS> eg without any second block
<SachaS> then i have to go into query mode and select a project from the project pulldown.
<reinhard> sorry I don't understand
<SachaS> and then hit query
<reinhard> SachaS: one block, which uses several datasources: (project_todolist, project_state, project_priority, project_project)
<reinhard> how can you use several datasources in one block?
<SachaS> sorry only one
<SachaS> dah
<SachaS> u are right
<SachaS> the block uses only one datasource
<reinhard> and the second datasource is to fill the project pulldown?
<reinhard> then that block has to go away
<SachaS> with the second block
<SachaS> i want to have a project selection
<SachaS> which creates a condition and sets the condition for the datasource of the first block
<reinhard> then you should be just able to remove that block
<reinhard> because the dropdown is bound to the datasource, not to the block
*** titopbs has quit IRC
<reinhard> jcater and/or jamest: I'm trying to understand the basical semantic difference between datasources and blocks
<reinhard> to me it seems like they are always connected 1:1, aren't they?
<reinhard> what is the "job" of a block as opposed to the "job" of a datasource?
*** sjc has joined #gnuenterprise
*** dcmwai has quit IRC
<jcater> a block contains forms-only attributes
<jcater> a datasource is universal across all tools
<jcater> other than that, it is 1:!
<jcater> 1:1
<jcater> but we wanted datasources that are valid in any one tool
<jcater> to be valid in all tools
<jcater> (at least, read-only ones)
<reinhard> ah
<reinhard> so the idea was like to have a xml datasource definition
<reinhard> to be included in a gfd and a grd
<reinhard> and be valid in both?
<jamest> more or less
<jamest> that's how it works now
<reinhard> ok
<reinhard> gotcha
<jamest> iirc forms and report both just import the parser definitions from the common datasource
<reinhard> yep
<reinhard> and I got it right that datasource:dataobject is also 1:1 and was only necessary because dataobject had some backend specific stuff while datasource is backend-independent?
<jamest> that i can't answer and I don't recall there being a dataobject :(
*** wendall911 has joined #gnuenterprise
<jamest> let me look
<jamest> reinhard: i think that's right
<jamest> i believe this was also to allow us to have more than one type of backend
<tiredbones> Is it oalso true that forms are onlu used with UI?
<jamest> you could have a sql type, or an object type
<reinhard> jamest: yes
<reinhard> however, the current DataObject/ResultSet code is able to handle both
<jamest> tiredbones: yes
<reinhard> jamest: and contains no more backend specific functions
<jamest> reinhard: i think the object type was never actually used only a placeholder
<reinhard> jamest: it was used
<jamest> but damn man, that's been more than a week ago :)
<reinhard> jamest: there was a DataObject_object and a DataObject_sql
<reinhard> class
<jamest> where week = year(s)
<jamest> DataObject_object was used?
<reinhard> yes
<jamest> ah, maybe the csv drivers?
<reinhard> no
<reinhard> all backends
<reinhard> the difference was
<reinhard> DataObject_object was when you had <field>s
<reinhard> DataObject_sql was when you had a <sql>
<jamest> ah, ok
<reinhard> but as it is now, would you (unless jcater speaks up) agree that DataObject could be phased out?
<tiredbones> When do you use blocks?
<reinhard> I remember jcater saying something that he never liked the distinction anyway
<jamest> reinhard: i really don't remember enough to be qualified to say
<jamest> i myself would not object to the change if it would help simplify the code base
<tiredbones> I under stand that block are use with forms
<jamest> tiredbones:
<jamest> er, hang on
<reinhard> jamest: I think that's very much the bad spell on gnue-common
<tiredbones> but I don't understand there use
<tiredbones> within forms.
<reinhard> nobody knows/remembers enough about how all of it works to dare to change something deeply :-)
<jamest> form
<jamest>   datasource
<jamest>   block
<jamest>   layout
<jamest> datasource links to data
<jamest> blocks contain fields that map to fields in the datasoure
<jamest> block tracks the current record being edited, the fields in the block enforce things like force uppercase, input length restrictions, etc
<jamest> layout contains entries which tie to fields in the block
<jamest> you can have more than one entry tied to a blocks field (usefull if you want the same entry fields on multiple pages of the same form)
<tiredbones> jamest: Is this M/D relationship?
<jamest> M/D used to be enforced at the block level years ago, this is now handled at the datasource level (reports needed it IIRC)
<jamest> reinhard: if somethings different now please tell tiredbones I'm full of it :)
*** wendall911 has quit IRC
<reinhard> jamest is correct :)
<jamest> with all the datasource commits i wasn't so sure :)
<tiredbones> jamest, reinhard: is the purpose of blocks to keep track of the virtual form?
<jamest> yes!
<jamest> it also allows you to place triggers at the block level to do certain processing
<tiredbones> Something else I'm confuse on, actually alot of thing. When I design a screen I
<tiredbones> usually use  the add screen as the edit screen.
<tiredbones> But it sound like I have to know before hand
<tiredbones>  what the user what before sending a screen up.
<tiredbones>  a screen.
<tiredbones> Is this correct?
<jamest> i didn't understand the question
<tiredbones> Ok, let try again.
<jamest> sorry
<tiredbones> If a person what to do an edit they have to click on the icon
<tiredbones> the says they what to do an edit.
<tiredbones> Is that right?
<tiredbones> jamest: I know I'm not using the tech ligo here.
<jamest> you mean the switch to edit mode icon?
*** wendall911 has joined #gnuenterprise
<tiredbones> jamest: yes
<jamest> i've never used that icon
<jamest> jcater added it iirc, but I'm not sure it's function
*** btami has joined #gnuenterprise
<tiredbones> jamest: Is that something that's going to stay?
<jamest> i think this may have been a way to tag certain forms as uneditible by default to prevent accidental changes
<jamest> which is probably a nice feature, but I'm not sure it was completed
<jamest> jcater?
<tiredbones> jamest: Could'n that be done with a trigger?
<jamest> probably, but one goal of forms was to make the standard stuff as easy as possible
<jamest> so if that feature would be needed for lots of people then we'd make a quick way to do it forms
<jamest> if it's what I think it is then maybe it shouldn't display that icon as active if the form doesn't require it
<tiredbones> Going back to the datasource thing for a minute.
<tiredbones> If I have in my gfd file a line like
<tiredbones> <datasource name="dtsrcPOInfo" cache="5"/>
<tiredbones>  what the difference in the this dataaource line?
<tiredbones> <block name="blckPO" datasource="dtscpoinfo">
<tiredbones> Do bothe datasource do the same thing?
<johannesV> could anyone post me to some 'good' docs about wxPython ?
<nickr> the wxWidgets api docs don't count as 'good'?
<johannesV> everything documented on wxPython.org is all about C++ and it looks like i can't find anything about the python implementation of wx
<johannesV> is that a doc for python or C++ ?
<jamest> johannesV: both
<jamest> let me get you url
<jamest> http://www.wxpython.org/onlinedocs.php
<nickr> the C++ and python apis are the same
<nickr> with noted differences
<jamest> under topic overviews you'll find a wxPython link
<jamest> the wx docs usually document the difference from the python version at the end of the C++ doc
<jamest> however last I used it you have to still read the wxPython source at times as some things are not documented at all
<johannesV> ok, thanks ... at least it looks like a good starting point
<johannesV> maybe i can get along with that wx-stuff till tomorrow ...
<tiredbones> Isn't Robin writting a book on wxPython?
<tiredbones> It might be online.
*** btami has quit IRC
*** jamest has quit IRC
*** jamest has joined #gnuenterprise
*** SachaS has quit IRC
*** SachaS has joined #gnuenterprise
*** johannesV has quit IRC
*** SachaS has quit IRC
*** SachaS has joined #gnuenterprise
<reinhard> jamest, jcater: I wonder if anybody calls GDataSource.addDetailDataObject or DataObject.addDetailDataObject manually
<reinhard> in case there is any chance you could find out whether you have any reference to "addDetailDataObject" in your non-gnue code it would be very helpful for me
<reinhard> thanks
<reinhard> bbl
<jamest> reinhard: i never call it
<jcater> I don't call it outside of forms/reports
<jcater> my dream was to make that GEvent based
<jcater> and have a DataSourceController (part of GConnections?)
<jcater> that so anyone could listen in to detail record changes
<jcater> just by hooking into the event
<jcater> I also dream of a day I don't have to work any more
<jcater> so don't give too much thought to them :)
<tiredbones> This datasource variable is confusing.
<tiredbones> In some the the material I'm reading it is related to
<tiredbones> persistence data.
<tiredbones> In another is links a block to a datasource.
<tiredbones> Can it do both?
<tiredbones> In another part of the document, datasource is associated with the block variable
<tiredbones> in the <entry> tag.
<tiredbones> Sorry for the confusing first sentence.
<tiredbones> My brain goes fast than what I can type.
<tiredbones> s/fast/faster/
*** SachaS has quit IRC
<reinhard> jamest, jcater: thanks for your reply, this makes life a lot easier for me
<reinhard> jcater: if you are still here, is there any special reason why GRDataMapper calls
<reinhard>       controlSection.resultset = self.sources.getDataSource(source)\
<reinhard>              .getDataObject().createResultSet(readOnly=1)
<reinhard> instead of just self.sources.getDataSource(source).createResultSet?
<reinhard> this is the only place where DataObject.createResultSet is called instead of GDataSource.createResultSet
*** jcater has quit IRC
*** jamest has left #gnuenterprise
<tiredbones> FYI,  this address returns Not Found",http://www.gnuenterprise.org/tools/common/docs/technotes/00
<tiredbones> Is Peter Sullivan still involed in the prodject?
<tiredbones> s/involed/involved/
<reinhard> night all
*** reinhard has quit IRC
*** titopbs has joined #gnuenterprise
*** sjc has quit IRC
*** wendall911 has left #gnuenterprise
*** wendall911 has joined #gnuenterprise
*** wendall911 has left #gnuenterprise
*** dimas has quit IRC
*** holycow has joined #gnuenterprise
<tiredbones> I was looking at some old archives and jcater metion he was going to look at Nola
<tiredbones> accounting package. Did anything ever develope?
*** jcater has joined #gnuenterprise
*** jcater has quit IRC
*** dcmwai has joined #gnuenterprise
*** titopbs has quit IRC
*** tiredbones has quit IRC
*** titopbs has joined #gnuenterprise