*** dcmwai has quit IRC
*** reinhard has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
*** btami has joined #gnuenterprise
*** kilo has joined #gnuenterprise
<kilo> good morning
<btami> good morning
<kilo> hi btami
<kilo> johannesV: can i call a procedure from a calculated property, both in the same module?
<johannesV> why not ?
<johannesV> the module does not matter, does it ?
<johannesV> just do something like session.call (...)
<johannesV> [but i haven't tried]
<kilo> i have something like this
<kilo>     <property name="totalGross5"   type="number(10,2)" >
<kilo>       return self.totalVAT('Gross', 5)
<kilo>     </property>
<kilo> 
<kilo>     <procedure name="totalVAT" type="number(10,2)">
<kilo>       <parameter name="type" type="string(5)"/>
<kilo>       <parameter name="rate" type="number(10,2)"/>
<kilo> blabla
<kilo> and when trying to access totalGross5, it tells me 3 params are given, though it expects only 1...
<johannesV> hey, you want to make me crazy ?
<johannesV> :)
<johannesV> as i've explained this a few weeks ago ... did i ?
<johannesV> j/k
<johannesV> ok, try to do a call like this:
<johannesV> params should always be given as keyword args
<johannesV> like: type = 'Gross', rate = 5
<johannesV> but i've to check if we need a dictionary here or keyword args ...
<kilo> DDDDDDDDDDAAAAAAAAAAAAAAAAAAAMMMMMMMMMMMMMMMMMMNNNNNNNNNNNNNNNNNNNNNNNN
<johannesV> just a sec
<kilo> DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN DAMN
<kilo> i remember now
<johannesV> yes
<johannesV> you need keyword-arguments
<johannesV> so this should be valid:
<johannesV> return self.totalVAT (type = 'Gross', rate = 5)
<johannesV> please try it
* johannesV needs some more coffee ...
<johannesV> :)
<kilo> if you need coffee, what d oI need...
* johannesV back
<kilo> testing...
<kilo> johannesV: great, it works. thank you
<johannesV> great
<btami> maybe it can go into reinhards dev-guide :)
<kilo> btw it would be very nice if we could call appserver procedures from a report...
<johannesV> i'll continue on the dev-guide anyway ... (cause reinhard hasn't very much time left for it)
<kilo> right now we can only access properties, but not procedures
<johannesV> how would you call procedures in a report ?
<btami> reports need formula support implementation
<btami> there is a GRFormula.py in base/layout ,but
<btami> it's unfinished
<johannesV> so you would treat an appserver-procedure as a formula ... hmm
<johannesV> not a bad idea
<btami> yep
<ajmitch_> hi
<btami> this is the missing part for gnue-repoorts 0.2
<btami> ajmitch_: hi
<johannesV> i haven't looked into reports very deeply now
<kilo> hi ajmitch_
*** darioush has joined #gnuenterprise
<johannesV> hi ajmitch_
<johannesV> in which state are our deb's ?
<darioush> hello people
<johannesV> hi darioush
<kilo> hi darioush
<darioush> is anyone in the mood of explaining this thing to me?
<ajmitch_> johannesV: same state as last week, still trying to get them uploaded
<kilo> which thing
<darioush> hello?
<johannesV> ajmitch_, ok thanks
<ajmitch_> all the debs that I've done are up to date
<ajmitch_> I've just been trying to get the uploads sponsored
<kilo> darioush: what should be explained?
<kilo> johannesV: do you have 15 mins available?
<johannesV> kilo, for doing what ? :)
<kilo> set up gnue-pim for yourself, so that i could explain a bug in gtk interface
<kilo> s/interface/driver
<johannesV> what do i need to do so ?
<kilo> setup is easy, just like gnue-invoice
<kilo> ring back if ready
<johannesV> ok
<johannesV> ERROR POSTING RECORD # 1
<johannesV> ERROR:  value too long for type character varying(70)
<johannesV> INSERT INTO gnue_class (gnue_name, gnue_module, gnue_id, gnue_comment) VALUES ('RelType', '80692804774038157241424696145182', '97445745295931443811722574221329', 'Party-Person relation types (Employee, Contact, etc.), used in m:n links')
<johannesV> you're still using too long comments, do you ?
<kilo> because interbase does not give error...
<johannesV> well, kilo, but you do now gnue_comment does only have 70 characters for this defined, don't you ?? *lol*
<kilo> well, ok, my fault, havnet checked it so far
<johannesV> np
<johannesV> have everything loaded now
<darioush> i don't know, it is likely that i would like to contribiute to this project (or something like that)
*** darioush has quit IRC
* johannesV phone
* johannesV back ..
<johannesV> hmm, that was a quick call ... :)
<kilo> lol
<johannesV> kilo, so what shall i check ?
<kilo> up and ready, appserver running?
<johannesV> yes
<kilo> start form SelectAddr.gfd
<johannesV> ok
<kilo> if you start only, then click Close, it closes OK
<johannesV> right
<kilo> if you click on one of the listboxes, it will not close when hitting Close
<kilo> no way out, 'kill' is your friend...
<johannesV> ok, i'm out (without a kill)
<johannesV> but it was hard to do so ...
<johannesV> clicking close twice does the job
<johannesV> the listbox seems to dirten the block
<kilo> you were lucky then
<kilo> yes, seems like the block is always in 'dirty' state
<kilo> interesting fact is that if you start the same form in wx, it doesn't behave this way, you can always get out with Close button
<johannesV> so it's a bug in wx
<johannesV> *lol*
<johannesV> :)(
<johannesV> i try to understand your class-hierarchy first ...
<kilo> ask if something is not clear. look at doc dir
<kilo> but your explanation of the bug is rotfl
<johannesV> i do not understand your gfd at all
<johannesV> why are you using so much fk_...
<johannesV> like the block "blkBoundAddr"
<johannesV> it is 'self-referring', is it ?
<kilo> yes
<johannesV> why ? what do you want to do with this form ?
<kilo> list all available addresses on the left, and addresses bound to a person on the right
<johannesV> ok, "all available" addresses first then
<johannesV> where should they come from
<kilo> from ADDR_Address
<kilo> a listbox needs fk_ things iirc
<johannesV> so this is a 'select * from ADDR_Address' displayed in a listbox ?
<kilo> yes
<johannesV> ok
<johannesV> and we do want to see 'formatted' only
<kilo> aye
<johannesV> hmm, still trying
<johannesV> ok, which class should be edited in there ?
<kilo> ok
<kilo> edited? PIM_AdPe
<johannesV> ok, so new records get into PIM_AdPe
<kilo> with Add button, that calls PIM_AdPe.add
*** lekma has joined #gnuenterprise
<lekma> hi everyone
<kilo> hi lekma
<lekma> hi kilo
<kilo> news?
<lekma> we're at customer offices testing
<lekma> and things are not going very well
<kilo> oh
<lekma> we're so damn late
<kilo> bad news
<johannesV> kilo, do you want to see 'all' available addresses every time, or if one get's selected only the remaining addresses
<kilo> all
<johannesV> kilo, so why don't you use just a block with rows then ?
<johannesV> (instead of a listbox
<johannesV> )
<kilo> a listbox is cooler
<kilo> this is what customers are used to
<kilo> and i want to be a pioneer so that my name be engraved on the top of GNUe Mountain 8-))
<johannesV> :)
<johannesV> but a listbox is used to display a fk-relation
<johannesV> and in this case you don't have one ...
<kilo> it works with Ugly WX
<johannesV> as i've said before ... :)
<johannesV> it's a bug in wx ... *lol*
<johannesV> j/k
<johannesV> i'm still trying ...
<kilo> ah, i see, you want me to forget the problem, instead of solving it 8-))))
<johannesV> right *lol*
<kilo> good tactics 8-))))
<johannesV> ok, got the left listbox working ...
<johannesV> now i try to catch the bug ...
<kilo> as far as i looked into this, i see the problem is that the listbox causes equestREPLACEVALUE even if it should not, ie it has no items...
<johannesV> ah, now wx does not work too ... !
<johannesV> so it might not be a bug in uidrivers then
<johannesV> you can reproduce the bug in 'gauntlet.gfd' too ... !
<johannesV> will check with an earlier svn-checkout now ...
<johannesV> just start gauntlet.gfd and select an item in the listbox
<kilo> ah. is it good news then?
<johannesV> at least it shows up as 'non-uidriver-specific' bug
<johannesV> since wx has the same problem then
<johannesV> and this is nice, so i don't search for the bug at the wrong place
<johannesV> i think it must be something in GF* t
<kilo> @eightball where is the bug
<bigbrother`> kilo: NO!
<johannesV> wow, now this is really interessting
<johannesV> if you move away with the focus (from the listbox) everything's fine
<kilo> i am happy that you enjoy it 8-)))
<johannesV> kilo, have you noticed that you cannot select a value from a listbox without the mouse?
<kilo> yes
<kilo> affirmed
<johannesV> this is odd, isn't it ?
<kilo> @eightball is this odd?
<bigbrother`> kilo: One would be wise to think so.
<kilo> @zen
<bigbrother`> kilo: Flat is better than nested.
<kilo> @pydoc eval
<bigbrother`> kilo: eval(source[, globals[, locals]]) -> value. Evaluate the source in the context of globals and locals. The source may be a string representing a Python expression or a code object as returned by compile(). The globals and locals are dictionaries, defaulting to the current globals and locals. If only globals is given, locals defaults to it.
* kilo loves SupyBot
<lekma> johannesV or reinhard : 1 small question :is there a way to close at once all lists attached to a session (while keeping the session open)? or at least to know all list ids that are still open?
<lekma> another one: what are the parameters needed by abort() in a bound proc
<lekma> another: is there a way to make a Session .request condition case insensitive?
<lekma> lotsa questions
<lekma> :)
<johannesV> lekma, there's no way to access the lists of a session
<lekma> :(
<johannesV> abort () takes only one parameter, which is the message (as string)
<johannesV> common.logic.adapters.Base.requestAbort ()
<johannesV> session.request (): case-insensitve
<johannesV> is not available atm
<lekma> :(
<johannesV> although i would have needed it too
<johannesV> but i don't think all backends would have something similiar
<lekma> johannesV: did you ever try to session.rollback in a middle of an OnValidate??
<lekma> i get some strange error about None being an invalid value for gnue_id
<johannesV> no i haven't tried that
<johannesV> do you really want to do a rollback ?
<johannesV> or just abort the commit (so the user can fix and try again) ?
<johannesV> well, after all, gnue_id is defined as NOT NULL
<johannesV> so if you call 'rollback' in an onvalidate you set gnue_id of the current instance to "None" (meaning NULL)
<johannesV> which then violates the NOT NULL constraint
<johannesV> (which is checked after the *last* OnValidate trigger now)
<kilo> and that check is very good in place there...
<lekma> yes i agree
<lekma> but then is there a way to silently abort
<lekma> a comit?
<kilo> abort() not working?
<johannesV> lekma, no, i think there's no way to silently abort a commit
<johannesV> do you really want to hide this from the user ?
<johannesV> so he does not know wether his works has been saved or not ?
<lekma> well the initial problem is:
<lekma> we would like to call a procedure, without an object
<lekma> only parameters
<lekma> so i thougth about creating a fake class
<lekma> that would provide an object to the proc
<lekma> but never commit it
<lekma> which is kinda stupid i know
<johannesV> you want to simulate global procedures
<lekma> but i lack time to find a better solution
<lekma> global?
<johannesV> hmm, global in a sense of not bound to a given class
<lekma> well not bound to objects of a given class
<johannesV> rigth, i see
<johannesV> i haven't found a nice way to do this too
<johannesV> kilo: as to your problem from today regarding the listbox
<johannesV> this is not a listbox-specific problem
<johannesV> try to do this: start gauntlet.gfd, and enter "foo" into the first field
<johannesV> then try to quit the form ...
<johannesV> which won't work
<johannesV> now tab into the next field and try again !
<kilo> hmmm, strange
<johannesV> so you can see the problem is that an unbound field/block does interfere the exitApplication
<johannesV> and since a listbox is nothing else than a dropdown which is always open
<johannesV> chaning the selection in the listbox means 'change the master-records key according to the element shown in the listbox'
<johannesV> doing this the unbound "field" get's dirty !
<johannesV> which shouldn't be a problem anyway; this is a bug
<kilo> ah, so then another one
<kilo> but it wasn't an easy one to track down... was it?
<johannesV> not really, cause i was a bit confused with the listbox
<johannesV> but it's *not* a new one, although it's missing in BUGs of forms
<johannesV> anyway
<johannesV> i think the SelectAddress.gfd should *not* use listboxes
<johannesV> to accomplish
<johannesV> you can think of a listbox to be an open dropdown
<kilo> i'm sure you have seen hundreds of forms like this to select something. if nowhere else, then in designer's form wizard
<kilo> i just want to stick to that
<kilo> and i will be kinda sad if i can't accomplish that with gnue
*** btami has quit IRC
<johannesV> but you cannot compare the concept used in designer with gnue-forms
<johannesV> the listbox-widget is not thought to act as *grid*
<kilo> hmm, maybe, we get back to our old 'we should have a grid control' issue...
<kilo> bbl
*** kilo has quit IRC
*** MoR4euZ has joined #gnuenterprise
<MoR4euZ> guten tag
<johannesV> hi MoR4euZ
*** Vee has quit IRC
*** johannesV_ has joined #gnuenterprise
*** johannesV has quit IRC
*** lekma has quit IRC
*** chillywilly has quit IRC
*** jamest_ has joined #gnuenterprise
*** lekma has joined #gnuenterprise
*** chillywilly has joined #gnuenterprise
*** lekma has quit IRC
*** wendall911 has joined #gnuenterprise
*** dcmwai has joined #gnuenterprise
*** sjc has joined #gnuenterprise
*** johannesV_ has quit IRC
*** suitable has joined #gnuenterprise
*** Vee has joined #gnuenterprise
*** reinhard has quit IRC
*** dcmwai has quit IRC
*** Vee has quit IRC
*** jcater has joined #gnuenterprise
*** Vee has joined #gnuenterprise
* suitable wonders if he has the patience to lurk here until he figures out what to say
<jcater> I'm sure whatever you say is perfectly suitable
<suitable> I try to keep it that way, and will wait until I have something to contribute.
<chillywilly> hehe
*** btami has joined #gnuenterprise
*** MoR4euZ has quit IRC
*** SachaS has joined #gnuenterprise
*** sss has joined #gnuenterprise
*** yacc has joined #gnuenterprise
<sss> Is Gnue Integrator available for download?
<btami> there was no gnue-integrator release yet, it's in planning stage...
<btami> you can find some code in gnue-integrator dir in svn
*** suitable has quit IRC
*** suitable has joined #gnuenterprise
*** kilo_ has joined #gnuenterprise
*** suitable has left #gnuenterprise
*** btami has quit IRC
*** sss has quit IRC
*** cilkay has joined #gnuenterprise
*** cilkay has quit IRC
*** cilkay has joined #gnuenterprise
*** jamest_ has quit IRC
*** sjc has quit IRC
*** kilo_ has quit IRC
*** jcater has quit IRC
*** dcmwai has joined #gnuenterprise
*** wendall911 has quit IRC
*** dimas has quit IRC
*** jamest has quit IRC
*** SachaS_ has joined #gnuenterprise
*** SachaS has quit IRC