*** dcmwai has quit IRC
*** dimas has joined #gnuenterprise
*** lxf has joined #gnuenterprise
*** sjc has joined #gnuenterprise
*** lxf has quit IRC
*** kilo has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
<dimas> hi reinhard
<dimas> :)
<kilo> good morning
<dimas> hi kilo
<kilo> hi dimas
<kilo> anyone using mysql?
<dimas> not me
<reinhard> good morning all
<reinhard> kilo: I don't think anybody is using mysql in production
<reinhard> johannes has installed it and does some tests from time to time what's necessary to maintain the driver
<kilo> a chap wrote me in private, he is using mysql on winxp and asked for a decent connection.conf entry
<reinhard> should be the usual provider/dbname/host/port
<dimas> reinhard: my problem is disappeared :)
<kilo> hmm, thx reinhard, will tell him then
<dimas> if count my dirty hands as main cause, it still exist
<reinhard> dimas: and do you have any idea what made it go away?
<dimas> strange i can't explain at what point it could get wrong, when today i broke it completely (it did not work even without filters) i updated ServerAdapter and xmlrpclib.py
<dimas> thats all
<dimas> broken ones saved
<reinhard> hmm
<reinhard> *very* strang
<reinhard> strange
<dimas> yes, because xmlrpclib was fresh as i switched to python2.3
<reinhard> I'd again say that you had local modifications in the xmlrpc part of common, but then actually you tried with a completely fresh copy of appserver and common yesterday
<dimas> and i freshed ServerAdapter yesterday
<dimas> after some rest i will into them
<dimas> will look
<ajmitch> hi
<dimas> hello ajmitch
*** mnemoc has quit IRC
*** btami has joined #gnuenterprise
*** dimas has quit IRC
*** dimas has joined #gnuenterprise
*** dimas has quit IRC
*** dimas has joined #gnuenterprise
*** sjc has quit IRC
*** holycow has joined #gnuenterprise
*** mnemoc has joined #gnuenterprise
*** dere1 has quit IRC
*** btami has quit IRC
<ajmitch> reinhard: packages in ubuntu hoary are in sync with sid now
<reinhard> ajmitch: great!
<reinhard> ajmitch: isn't there some kind of freeze for hoary?
*** holycow has quit IRC
*** tiredbones has joined #gnuenterprise
*** kilo has quit IRC
<dimas> reinhard: you here?
<reinhard> yes
<dimas> can you tell me a little about languages in app server
<dimas> if i initialized with only *-C.gld
<dimas> what would i get?
<reinhard> then you will only see English no matter what language you use
<reinhard> for generated forms
<dimas> could it somehow affect all foms?
<dimas> now all new data is eaten
<dimas> old data displayed properly
*** jamest has joined #gnuenterprise
<dimas> old/new - before/after filter initialization (.gld loading)
<reinhard> not sure I understand you correctly
<reinhard> you have a class that previously had no filter?
<reinhard> and added a filter later?
<dimas> yes
<reinhard> hmmm
<reinhard> is that possible at all?
<dimas> i added class that required filter later
<reinhard> ah so you didn't add a filter to an existing class
<dimas> hm
<reinhard> but you added a new class that had a filter?
<dimas> yep
<reinhard> and you don't see data in that new class?
<dimas> other classes with data existed before
<reinhard> or you are missing data in the old class?
*** jamest has quit IRC
<dimas> no, old classes is good
<reinhard> ok
<reinhard> so new class doesn't display data you entered?
<reinhard> is the data in the db backend?
<reinhard> can you check whether the filter column is filled out correctly?
<dimas> if I insert new data in any classes it commits without error but do not displayed later
<dimas> ok
<dimas> data should go into postgres
<dimas> reinhard: i have gl-ru.gld, btw
<reinhard> so there must be a difference between the rows inserted before and those inserted afterwards
<reinhard> in the db
<dimas> looking into it
*** jamest has joined #gnuenterprise
<dimas> reinhard: rows were commited, when i query them i see proper number of rows queried but fields empty
<dimas> if dump to gsd to this rows only gnue_ stamps fields exist
<reinhard> empty fields in db?
<dimas> second time i inserted ascii data over previous russian
<dimas> in the db it is good
<dimas> could paste here
<reinhard> just a sec, brb
<dimas> now i loaded gl-ru.gld, restarted app server - all forms does not work
<dimas> DB009: 0:05:01.619 [ServerAdapter:464] Unexpected Exception in XML-RPC code: exceptions.UnicodeDecodeError:'ascii' codec can't decode byte 0xd0 in position 0: ordinal not in range(128)
<dimas> DB009: 0:05:01.631 [ServerAdapter:418] exodus.vbc.elcom.ru - - [11/Mar/2005 17:37:47] "POST / HTTP/1.0" 500 -
<dimas> 
*** jcater has joined #gnuenterprise
<reinhard> dimas: this seems to happen in ServerAdapter between line 441 and 458
<reinhard> you could remove the try - except block around that to get the complete traceback
<reinhard> but actually it looks as if you use an old version of that file
<dimas> reinhard: i would better not change ServerAdapter today
<dimas> joke
<dimas> old like yesterdays
<reinhard> because for me, this message is on debug level 1??
<reinhard> what does the copyright line of ServerAdapter say in your version?
<dimas> there is no version line there
<reinhard> copyright
<reinhard> which copyright year
<dimas> 1->9 is my changes, sorry
<reinhard> ah, ok
<dimas> Copyright 2001-2005
<reinhard> ok, then it's ok :)
<reinhard> just thought you have an old copy of something there
<dimas> ServerAdapter from rev 7007
<reinhard> is ok, yes
<dimas> seems just how languages behave
<dimas> i have initialized something with *-C.gld
<dimas> now i added some *-ru.gld
<dimas> what i have with languages now?
<reinhard> the gld's should only affect the look of the generated forms and the messages displayed
<reinhard> not the data itself
<reinhard> not in any way
*** dcmwai has joined #gnuenterprise
*** dcmwai has quit IRC
*** dcmwai has joined #gnuenterprise
<reinhard> jcater: if you have a minute, I have a problem understanding a feature of datasources
<reinhard> it's the _primaryIdField and _primaryIdFormat stuff
<reinhard> oracle seems to be the only driver using it
*** wendall911 has joined #gnuenterprise
*** btami has joined #gnuenterprise
<reinhard> jcater: ?
*** SachaS has joined #gnuenterprise
<reinhard> jcater, jamest, all: just committing a *quite* intrusive change in dbdrivers, please test before using it in production.
<reinhard> especially stuff dealing with primary key generation and handling
<jamest> test?
<reinhard> erm sorry
<reinhard> I mean "please test by using it in production"
<reinhard> ;-)
<jamest> ooooh
<jamest> ok, np
<reinhard> other question
<reinhard> we have a whole bunch of dbdrivers that have been unmaintained and not functional for years now
<reinhard> db2, gadfly, informix, ingres, ldap, odbc, sqlrelay, sybase
<reinhard> they haven't even followed the last few structural changes
<reinhard> I wonder if we should remove them from svn, move them to gnue-contrib, or whatever
<jamest> probably
<reinhard> they confuse me in my code cleanups as the seem to "use" stuff that's no longer supported or not even available any more
<reinhard> and when I grep for a function or a variable, I get buttloads of mathces from these drivers
<jamest> i'd say ditch them
<reinhard> what would you think? svn remove?
<reinhard> (can be recovered anyway through svn if somebody wants to revive them)
<jamest> i think that's fine
<reinhard> ok, thanks
*** btami has quit IRC
*** johannesV has joined #gnuenterprise
<johannesV> reinhard, glds also affect filters: they determine the labels an props to be queried in login dialog
<johannesV> btw i had serious troubles with GParser and friends regarding garbage collection today
<johannesV> this is because appserver processes xml files in the same process multiple times ... (gcd, gld)
<johannesV> so the leaking parser and obj.trees accumulated
*** btami has joined #gnuenterprise
<johannesV> bbl
*** mnemoc has quit IRC
*** tiredbones has quit IRC
<johannesV> back
<johannesV> reinhard, you still there ?
<reinhard> yes
<johannesV> i've added a method 'breakReferences ()' to ParserObj in GParserHelpers.py
<johannesV> this method clears the ref to the _parent and also clears the _children and _inits sequences
<reinhard> ok
<johannesV> doing this it allows xml-object trees to get freed properly
<johannesV> but i somehow don't like the name 'breakReferences' since it does a bit more
<reinhard> what about free () ?
<johannesV> anyway, it'd be better to use weakref.ref () for _parent, as well as all children
<johannesV> but this would imply a massive change to the code
<johannesV> and i really don't like to do that now
<reinhard> what's weakref?
<johannesV> yes, free () sounds good to me (as i've thought of this name too)
<reinhard> (kudos to good old delphi :))
<johannesV> that's a python module designed for creating chaches and mappings
<reinhard> well, I'd vote for no more massive changes
<reinhard> it'd be good if you could commit over the weekend
<reinhard> as on monday a lot of customer work will wait for you, so you might be away from gnue for a week or two
<johannesV> it creates a weak reference to an object, which does not interfere with the garbage colletion
<johannesV> ok, so i'll stick with my solution right now, as it seems to have no more leaks (atm)
<johannesV> and maybe i can commit it this evening or at least tomorrow ...
<reinhard> excellent
<johannesV> i was quite shocked after i've found out about those 6000+ leakin' objects after a sighup to appserver ...
<johannesV> and every additional sighup added another 6000+
<johannesV> (and these are numbers with a quite small repository ~600 props)
<reinhard> do you actually follow my changes in common?
<reinhard> quite some restructuring
<johannesV> yes, i'm reading all commit-mails ..
<johannesV> i felt good while reading it ... so just go on in that manner :)
<reinhard> :)
<reinhard> let's hope jamest and jcater feel the same :)
<jamest> lol
<johannesV> btw, i've added support of typedefs like "number(5,2)" to gnue-schema (in my working copy)
<johannesV> ok, i'm off for another hour
*** sjc has joined #gnuenterprise
*** Egyptian[Home] has joined #gnuenterprise
<johannesV> reinhard, after updating svn i get a traceback on starting appserver
<johannesV> shall i paste ?
<reinhard> yep please
<johannesV> DB000:     resultSet= self._dataObject.createResultSet(conditions,readOnly,sql=sql)
<johannesV> DB000:    File "/home/johannes/prj/gnue/.cvsdevelbase/gnue/common/datasources/drivers/Base/DataObject.py", line 90, in createResultSet
<johannesV> DB000:     readOnly=readOnly, masterRecordSet=masterRecordSet, sql=sql)
<johannesV> DB000:    File "/home/johannes/prj/gnue/.cvsdevelbase/gnue/common/datasources/drivers/DBSIG2/DataObject.py", line 190, in _createResultSet
<johannesV> DB000:     query = self._buildQuery(conditions, additionalSQL = sql)
<johannesV> DB000:    File "/home/johannes/prj/gnue/.cvsdevelbase/gnue/common/datasources/drivers/DBSIG2/DataObject.py", line 317, in _buildQuery
<johannesV> DB000:     whereClause = self._conditionToSQL(conditions)
<johannesV> DB000:    File "/home/johannes/prj/gnue/.cvsdevelbase/gnue/common/datasources/drivers/DBSIG2/DataObject.py", line 235, in _conditionToSQL
<johannesV> DB000:     where = " WHERE (%s)" % (self.__conditionToSQL (cond._children[0]))
<johannesV> DB000:    File "/home/johannes/prj/gnue/.cvsdevelbase/gnue/common/datasources/drivers/DBSIG2/DataObject.py", line 266, in __conditionToSQL
<johannesV> DB000:     result.append(self.__conditionToSQL(element._children[i]))
<johannesV> DB000:    File "/home/johannes/prj/gnue/.cvsdevelbase/gnue/common/datasources/drivers/DBSIG2/DataObject.py", line 278, in __conditionToSQL
<johannesV> DB000:     raise GConditions.ConditionError, tmsg
<johannesV> DB000:  ConditionError: Bedingungselement 'or' erwartet mindestens 2 Argumente, hat aber nur 0
<johannesV> (it's only the tail)
<reinhard> hmm... doens't look like something related to my changes at all
<johannesV> hmm, maybe it's not your fault ...
<johannesV> i'll try something
<reinhard> it's not my fault in any case - it's derek's fault
<reinhard> but that's a completely different topic
<reinhard> :)
<johannesV> ah, right, i forgot about derek ... *lol*
<johannesV> ok, got it ... sorry for blaming you
<reinhard> np
<johannesV> hmm, found another leak :/
<johannesV> this looks like a never-ending story ...
<johannesV> seems as if it's something with conditions again ...
<johannesV> but i need some further tests ...
<johannesV> damn, this new repository rocks ... starting appserver is a matter of a second :)
<johannesV> (including complete reload of gcd/gld)
<reinhard> johannesV: you tell this just to tease us, right?
<reinhard> so we wait eargerly for your commit ;-)
<johannesV> *lol* ... yeah, exactly ...
<johannesV> but i think i'll commit it, and fix memory leaks afterwards ...
<reinhard> I think you can do that
<johannesV> hm, no i'll change readgcd/gld a bit first ...
<johannesV> but this isn't too complicated
<johannesV> ok, the leak comes in on adding a *new* record using sample.gfd
<johannesV> it's 41 objects, all of them are dealing with GCondition instances ... so maybe there's still a ref-cycle in there ...
<johannesV> think i'll change the station now
<johannesV> bbl
*** johannesV has quit IRC
<btami> reinhard: the interbase dbdriver has a problem
<btami> c:\python23\scripts>C:\python23\python.exe C:\python23\scripts\gnue-appserver
<btami> Szolgáltatások exportja ezen keresztül: xmlrpc (port 8765)...
<btami> Osztßlyk÷nyvtßr ˙jrat÷ltÚse...
<btami> gnue-appserver: Nem sikerŘlt a firebird plugin bet÷ltÚse
<btami> Az alßbbi plugin-eknÚl t÷rtÚnt hiba:
<btami> * gnue.common.datasources.drivers.interbase.interbase: NameError: name 'kinterbasedb' is not defined
<reinhard> btami: let me fix
<reinhard> should be fixed
<reinhard> thanks for reporting
<btami> np
<btami> another one: after F8,F9, then pressing "Show record" button in appserver's sample.gfd
<btami>    File "c:\python23\Lib\site-packages\gnue\common\datasources\drivers\DBSIG2\Connection.py", line 269, in makecursor
<btami>     cursor.execute (s)
<btami>    File "C:\python23\Lib\site-packages\kinterbasdb\__init__.py", line 1582, in execute
<btami>     self.description = _k.execute(self._C_cursor, sql, params)
<btami>  ProgrammingError: (-804, 'isc_dsql_prepare: Dynamic SQL Error. SQL error code = -804. Function unknown. LOWER. ')
<btami> is LOWER in SQL standard ?
<reinhard> I thought so
<btami> hmm
<reinhard> I think this problem relates to case insensitive queries
<reinhard> debug level 3 should now give you exact SQL statement output
<reinhard> although I wonder what query is caused by the "show record" button
<btami> DB003: 0:00:42.831 [c:\python23\Lib\site-packages\gnue\common\datasources\drivers\DBSIG2\Connection:260] DBSIG2 Statemen
<btami> t: SELECT t0.gnue_text,t1.gnue_id,t0.gnue_id FROM gnue_message t0 LEFT JOIN gnue_module t1 ON t1.gnue_id = t0.gnue_modul
<btami> e WHERE (((LOWER(t1.gnue_name) = 'address') AND (LOWER(t0.gnue_name) = 'nameis') AND (t0.gnue_language = 'hu_HU')))
<btami> DB003: 0:00:42.841 [c:\python23\Lib\site-packages\gnue\common\datasources\drivers\DBSIG2\Connection:261] DBSIG2 Paramete
<btami> rs: None
<reinhard> ah
<reinhard> let me test something
*** siesel has joined #gnuenterprise
<siesel> hi
<reinhard> btami: yes, I get exactly the same statement, but it works for postgres
<reinhard> there is no LOWER in interbase?
<reinhard> hey siesel
<reinhard> long time no see
<reinhard> (like always) ;-)
*** johannesV has joined #gnuenterprise
<siesel> :)
<johannesV> back
<btami> hi siesel
<siesel> hi btami, reinhard, johannes
<johannesV> btami, lower is sql-standard
<btami> reinhard: there is lower in udf library, but i'v never used
<johannesV> i thought i've checked this for my firebird too
<johannesV> the lower is used for loading messages
<johannesV> i'll check my firebird again, just a sec
<johannesV> btw, hi siesel
<johannesV> :)
<johannesV> hm, right ... there's no function lower
<btami> it's in ib_udf
<johannesV> what's ib_udf ?
<btami> but i don't know how to use it
<btami> an external lib shipped with interbase/firebird
<johannesV> how can one use this lib ?
<btami> this is the question :)
<johannesV> ah, got it ... sing the isql utility, execute the ib_udf.sql script on your database:
<johannesV>    ./isql /path/to/yourdb/yourdb.gdb -i /opt/interbase/examples/ib_udf.sql
<johannesV> but, how to use this through kinterbasedb ...
<johannesV> hmm, no such file (or similair) in any deb-package ...
<johannesV> so it looks like we have no support for 'lower' and 'upper' in interbase/firebird atm
<btami> i think you can use any valid SQL statement with exec() in kinterbasdb
<johannesV> this means i need another change of appserver
<johannesV> btami, can you find out if it's possible to use some kind of 'lower' function with ib/firebird ?
<btami> trying...
<reinhard> bbl
*** reinhard has quit IRC
<SachaS> hi siesel
<btami> johannesV: just did an execute() with
<btami> DECLARE EXTERNAL FUNCTION lower
<btami> 	CSTRING(255)
<btami> 	RETURNS CSTRING(255) FREE_IT
<btami> 	ENTRY_POINT 'IB_UDF_lower' MODULE_NAME 'ib_udf';
<SachaS> in about 4 weeks i am leaving bonn
<btami> and it worked
<johannesV> btami, ok, but you need that ib_udf library then ...
<btami> without ; at the end
<btami> sure
<siesel> hi SachaS
<siesel> haven't been close to Bonn yet.
<btami> it's an official part of interbase/firebird
<siesel> were do you're leaving to?
*** SachaS_ has joined #gnuenterprise
<SachaS_> siesel not much time left to meet
<johannesV> hmm, the above code doesn't work for me ...
<btami> johannesV: i have it in my UHU-Linux distro in /usr/lib/firebird/Udf/ib_udf.so
<btami> what do you get with dpkg -L firebird-bla-bla ?
<johannesV> /usr/lib/firebird2/UDF/fbudf.so
<johannesV> so it's called fbudf
<btami> there is 2 .so files, fbudf.so and ib_udf.so
<btami> fbudf.so is a minor addition
<btami> lower is in original ib_udf.so
<johannesV> ah, ok so this must be in another package then ...
<btami> anyhow, after adding lower to gnue.fdb i get this error
<btami> Traceback (most recent call last):
<btami>    File "<address_person.address_show>", line 8, in show
<btami>    File "c:\python23\Lib\site-packages\gnue\appserver\language\Session.py", line 389, in message
<btami>     raise MessageNotFoundError, self.qualify (messageName)
<btami>  MessageNotFoundError: 'address_nameis' üzenet nem található
<johannesV> will try the code again (maybe i need to write it in a single line ....
<johannesV> do you have sample-C.gld imported ?
<btami> don't think so
<johannesV> well, so you do not have that message defined, do you ?
<johannesV> you can double-check if you do a 'select gnue_ name from gnue_message'
<btami> ok, added it, and now it works
<btami> the problem was
<btami> python gnue-readgld --connection=gnue sample-*.gld
<btami> in setup-cvs-db.bat
<btami> it doesn't works, so i changed it to use sample-C.gld only
<btami> sigh
<johannesV> hmm, that's strange ... since sample-*.gld should work fine
<btami> c:\python23\scripts>python gnue-readgld --connection=gnue sample-*.gld
<btami> gnue-readgld: Nem lehet megnyitni a [Errno 2] No such file or directory: 'sample-*.gld' bemeneti fßjlt.
<johannesV> ?!
<johannesV> so your shell has a problem ...
<johannesV> doesn't sample-*.gld get's replaced by the shell ... ?
*** SachaS has quit IRC
<btami> isit ok on any of your win32 box
<btami> ?
<johannesV> i haven't one up and running ...
<btami> :)
<johannesV> who's using win32 ? *lol*
<johannesV> but i can try this tomorrow ...
<btami> just my kids for playing AOE
<btami> ok, have to go, bye
*** btami has quit IRC
<johannesV> off for today
*** johannesV has quit IRC
*** mnemoc has joined #gnuenterprise
*** kilo has joined #gnuenterprise
*** mnemoc has quit IRC
*** reinhard has joined #gnuenterprise
*** reinhar1 has joined #gnuenterprise
*** reinhar1 has left #gnuenterprise
*** reinhard has quit IRC
*** reinhard has joined #gnuenterprise
*** mnemoc has joined #gnuenterprise
<SachaS_> reinhard your still working?
<reinhard> kinda
<SachaS_> i wish i had your stuamina
<mnemoc> reinhard is a bot
*** jamest has left #gnuenterprise
*** mnemoc has quit IRC
<kilo> who can stop reinhard?
<SachaS_> :)
*** tiredbones has joined #gnuenterprise
*** kilo has quit IRC
*** jcater_ has joined #gnuenterprise
*** jcater_ has quit IRC
*** jcater has quit IRC
*** mnemoc has joined #gnuenterprise
* SachaS_ thinks reinhard and johannes are on steroids
* SachaS_ wants some as well
<SachaS_> ;)
<SachaS_> night
*** SachaS_ has quit IRC
*** tiredbones has quit IRC
<reinhard> night all
*** reinhard has quit IRC
*** sjc has quit IRC
*** seb_ has joined #gnuenterprise
<sebastian> hello.anybodu at home?
*** sebastian has left #gnuenterprise
*** kraut has joined #gnuenterprise
<kraut> anyone here?
*** kraut has quit IRC
*** dimas has quit IRC
*** Nineth has joined #gnuenterprise
*** Nineth has left #gnuenterprise
*** ajmitch has quit IRC
*** ajmitch has joined #gnuenterprise