*** lekma_ has quit IRC
*** SachaZzz has quit IRC
*** reinhard has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
<johannesV> good morning
<reinhard> hi johannesV
<dimas> good morning
<johannesV> db-drivers are getting smaller and smaller :)
<johannesV> great work !
*** btami has joined #gnuenterprise
<btami> good morning
<reinhard> good morning
*** holycow has quit IRC
*** kilo has joined #gnuenterprise
<johannesV> btami, do you know of a builtin-concept of "oid" or "rowid" in firebird ?
<kilo> good morning
<btami> as i know there is no rowid/oid in firebird
<johannesV> ok, this is what i've found from docs too
*** mnemoc has quit IRC
<johannesV> btami, do you know the names of the udf's for upper and lower ? or even better, can you give me the sql-code to incorporate these functions on database creation ?
<johannesV> i'd like to update common's interbase-driver (schema-creation) so it starts off with all functions needed by appserver/common
<johannesV> i have found the following declaration:
<johannesV> DECLARE EXTERNAL FUNCTION name [datatype | CSTRING (int)
<johannesV> [, datatype | CSTRING (int) …]]
<johannesV>    RETURNS {datatype [BY VALUE] | CSTRING (int) | PARAMETER n}  [FREE_IT]
<johannesV>    ENTRY_POINT <External_Function_Name>
<johannesV>    MODULE NAME <Library_Name>;
<johannesV> ah, firebird has an UPPER function built in, so only lower is missing
<btami>  *****************************************/
<btami> DECLARE EXTERNAL FUNCTION lower
<btami> 	CSTRING(255)
<btami> 	RETURNS CSTRING(255) FREE_IT
<btami> 	ENTRY_POINT 'IB_UDF_lower' MODULE_NAME 'ib_udf';
<reinhard> what? firebird has UPPER built in but not LOWER? I wonder what *that* person did smoke to get to such a decision :)
<btami> lol
<kilo> lol
<btami> i think it was inherited from interbase (Borland)
<kilo> and a corp can decide ANYTHING...
<reinhard> yeah, borland people always had some strange crack...
<kilo> i knew one personnally
<kilo> he was banned from the US...
<kilo> his visa expired and he forgot that...
<johannesV> ok, your definition is working, but it returns columns of a fixed size: 255 characters
<btami> it was ripped from ib_udf.sql
<btami> shipped by Firebird team
<btami> with comments:
<btami>  *
<btami>  *	l o w e r
<btami>  *
<btami>  *****************************************
<btami>  *
<btami>  * Functional description:
<btami>  *	Returns the input string into lower
<btami>  *	case characters.  Note: This function
<btami>  *	will not work with international and
<btami>  *	non-ascii characters.
<btami>  *	Note: This function is NOT limited to
<btami>  *	receiving and returning only 255 characters,
<btami>  *	rather, it can use as long as 32767
<btami>  * 	characters which is the limit on an
<btami>  *	INTERBASE character string.
<btami>  *
<btami>  *****************************************/
<johannesV> hmm, so that's quite bad ...
<johannesV> ok, createDatabase () will add the udf lower now
<johannesV> but what if one calls this on unicode values ?
<johannesV> ok, that lower() function lowers all ascii-characters, passing through everything else ... at least it does not throw an exception
<johannesV> reinhard, no row-id/oid concept for mysql (including innodb) and firebird/interbase - both are using the concept of a generator [which has to be added by the user, so no automatism behind]
<johannesV> shall i set up MaxDB on my database server again in order to check things like this ?
<btami> may i suggest an optional config option (in connection.conf)
<btami> rowid=tablename_id
<btami> assuming tablename_id is a generator
<kilo> underline is sacred cow
<johannesV> :)
<btami> s/is a/autofilled by a
<btami> then tablename-id :)
<kilo> lol
<reinhard> btami: you can achieve that with primarykey="columnname" in the datasource tag
<kilo> that's great.
<kilo> then why do we need this rowid thinggy exactly?
<reinhard> for cases where we have no primary key
<kilo> ah
<btami> ok
<btami> :))))
*** dcmwai has quit IRC
<btami> bbl
*** btami has quit IRC
<reinhard> bbl
*** reinhard has quit IRC
*** tiredbones has joined #gnuenterprise
*** tiredbones has quit IRC
*** tiredbones has joined #gnuenterprise
*** jamest has joined #gnuenterprise
<johannesV> bbl
*** johannesV has quit IRC
<dimas> <greg2> DB000:     primarykeyFields = dataObject._connection._primarykeyFields
<dimas> <greg2> DB000:  AttributeError: Connection instance has no attribute '_primarykeyFields'
<dimas> <greg2> make: *** [init] ïÛÉÂËÁ 1
*** kilo has quit IRC
<dimas> what could be cause of this?
<dimas> this is when creating backend and loading gnue.gsd
<jamest> this svn?
<dimas> may be 2 days old
<jamest> i was thinking that the datasources are undergoing some major work
<dimas> will try to update
<jamest> i would try that 1st
<jamest> as there were 8 commits involving them in the last 24 hrs :)
<chillywilly> cisco blows ppl
<dimas> ?
<chillywilly> I should make a router w/ strictly linux firmware to compete with cisco and kick them straight in the nads
<chillywilly> the linux kernel is better firmware than anything cisco has ever made
<jamest> why not use a bsd chillywilly
<jamest> like netbsd
<jamest> last i knew they were smaller footprint and typically had better networking code
*** jcater has joined #gnuenterprise
<chillywilly> you got anything to back that up?
* jcater flexes his biceps
*** reinhard has joined #gnuenterprise
<chillywilly> not you
<dimas> gnue-schema: Cannot load plugin 'psycopg'
<jamest> chillywilly: just years of reading articles
<reinhard> dimas: python 2.3?
<dimas> yes
<jamest> i don't have specifics
<dimas> the same with postgresql.psycopg
<reinhard> probably with every driver
<chillywilly> jamest: riiight ;)
<jamest> i thought it was pretty well known
<chillywilly> hand waving will get you nowhere sir ;)
<chillywilly> yes, netbsd has a smaller footprint for sure but I have no facts to backup your statement about networking performance
<dimas> reinhard: there is no problem on my machine but with the same distro at the friend's
<reinhard> erm
<reinhard> cvs version or packages?
<reinhard> s/cvs/svn/
<dimas> cvs
<chillywilly> http://www.nwfusion.com/columnists/2000/1120works.html
<chillywilly> there ya go jamest ;)
<dimas> updated on both sides up to now
<reinhard> dimas: I guess you already checked if python-psycopg is installed
<dimas> rpm is installed
<reinhard> dimas: and it works on your machine, with current svn?
<dimas> how to better check if it is loaded for interactive python?
<jamest> chillywilly: :)
*** mnemoc has joined #gnuenterprise
<dimas> psycopg is loading from python
<reinhard> dimas: it works on your machine with current svn?
<reinhard> (trying to find out if svn is broken or it's a problem of that machine)
<dimas> had no problem on mine
<dimas> and that another host is not accessible now for me
<dimas> until sunday
<reinhard> that one with the error message is not accessible for you?
<dimas> yes
<dimas> we will try again at sunday, i'll let you know
<dimas> and thanks anyway
*** SachaS has joined #gnuenterprise
<SachaS> hi everyone
<dimas> hi SachaS
<dimas> if form's data is changed and then appserver is switched off i can't close the form
<reinhard> right
<reinhard> you can't save, you can't rollback, and you can't close with unsaved data :)
<reinhard> but I agree we would want to change that )
<dimas> hey, it has closed with segmentation fault finally
<dimas> not that bad :)
*** sjc has joined #gnuenterprise
<reinhard> hey
<reinhard> hoary released today
<reinhard> and www.ubuntu.com already kinda DOS'ed :)
<dimas> does authentication working now?
<dimas> with authentication = True in gnue.conf
<mnemoc> how does GNUe authenticate? there is a well-known table to create?
*** wendall911 has joined #gnuenterprise
<dimas> there is useraccess class that could be loaded
<mnemoc> i assume i have the .gcd installed somewhere :)
<reinhard> dimas: it should, yes
<reinhard> mnemoc: there should be gnue_useraccess.gcd
<reinhard> in the samples dir
<reinhard> but the current authentication system will be replaced by a well-designed and more flexible one in one of the next milestone releases
<mnemoc> reinhard: month?
<reinhard> June or July
<reinhard> but please don't ask for the year
<reinhard> j/k
<reinhard> I have no idea :)
<mnemoc> :)
<mnemoc> do you plan migration procedures?
<reinhard> dunno
<reinhard> bbl
*** tiredbones has quit IRC
*** wendall911 has quit IRC
*** tiredbones has joined #gnuenterprise
*** bigbrother has joined #gnuenterprise
*** bigbrother has joined #gnuenterprise
*** jcater has quit IRC
*** kilo has joined #gnuenterprise
*** wendall911 has joined #gnuenterprise
<reinhard> jamest: you here?
<jamest> yes
<reinhard> you still sure you are ok with breaking SELECT * support?
<jamest> when fields=()
<jamest> ?
<jamest> on the datasource
<reinhard> jcater agreed yesterday basically but thought you would need it maybe
<reinhard> yes
<jamest> well, its broken anyway for commits
<jamest> so I've started updating my apps anyway
<reinhard> it is?
<jamest> yeah
<reinhard> i did break it?
<jamest> something did recently
<jamest> but i'm already in the process of updating my apps to new common
<jamest> so i'm ok with the breakage
<reinhard> ok
<reinhard> please tell me if you notice some breakage
<jamest> force me to be less lazy :)
<reinhard> don't want to fsck up your applications
<jamest> anymore than i've already done :)
<jamest> don't worry about my apps, it's not a major change to fix them on my end
<reinhard> seriously
<reinhard> i know that i'm basically butchering something that you and jcater wrote
<jamest> i've got a wrapper around the gnue wrapper
<reinhard> and "just worked"(tm) for you so far for what you needed
<jamest> so for me its just adding a field=() to all my calls
<reinhard> and I hope you don't get me too wrong as to coming here and breaking everything
<jamest> so far i'm ok with stuff
<jamest> are you expecting data loss issues?
<reinhard> I am not sure what we can expect
<jamest> as long as readonly works
<jamest> i should be fine in most cases
<reinhard> but let me put it this way
<jamest> as I can deal with bad single commits
<reinhard> I would not use gnue svn currently without daily backups :)
<jamest> i can't deal with delete all rows in the table errors
<jamest> daily wouldn't quite cut it
* jamest checks the stats
<jamest> so far this month my gnue based web stuff has handled 162 new sales
<jamest> and generated close to 300 invoices
<reinhard> wow
<jamest> and it's been a slow month
<reinhard> that's about what I have in half a year
<reinhard> :)
<reinhard> well, so far my changes seem to have caused low breakage
<jamest> yes
<jamest> my fears are where a delete is requested
<reinhard> but still i'm a rather cautious person
<jamest> and the "where" part of the statement gets dropped
<jamest> :)
<reinhard> well, the reorg of that part of the code is acutally through
<jamest> but I can create some more rules on my tables
<jamest> as i track every change to a table and store the previous values
<jamest> so we know who to yell at when someone does something wrong
<reinhard> ah cool
<reinhard> so you actually have a backup of every transaction
<jamest> yes
<jamest> i have a backup of every row in a table
<reinhard> ok
<reinhard> still I think my next commit is quite experimental
<reinhard> and I would recommend to not update production systems for a few days until it shows to be stable
<jamest> i will not then
<jamest> as i just updated this afternoon
<jamest> and it's been fine
<reinhard> cool
<jamest> i need to run, have to go get my kids tonight
<reinhard> next commit includes a heavy restructuring of the DBSIG2 driver
<reinhard> ok
<reinhard> have fun with the kids
<jamest> i look forward to reading the diffs :)
<reinhard> and say hello from me ;-)
<jamest> ok
<jamest> l8r
*** jamest has left #gnuenterprise
*** wendall911 has quit IRC
*** tiredbones has quit IRC
*** kilo has quit IRC
*** tiredbones has joined #gnuenterprise
*** sjc has quit IRC
<reinhard> good night all
*** reinhard has quit IRC
*** jcater has joined #gnuenterprise
*** chillywilly has quit IRC
*** chillywilly has joined #gnuenterprise
<chillywilly> bah
<jcater> bah
<jcater> black sheep
*** Garrido has joined #gnuenterprise
<chillywilly> new OOo build is up at: http://mirrors.isc.org/pub/openoffice/developer/680_m91/
<chillywilly> gotta be getting close to releasing that bugger
<mnemoc> i'm still fighting to get m86 built
<chillywilly> compiling it?
<chillywilly> I just convert the debs w/ alien
<chillywilly> the rpms that is
*** holycow has joined #gnuenterprise
<mnemoc> i like to build the stuff i use :)
*** Garrido has quit IRC