*** btami has joined #gnuenterprise
*** kilo has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
<SachaS> good morning austria
<SachaS> ;)
<johannesV> hi SachaS
<johannesV> have you noticed the latest changes in appserver ?
* johannesV looking for some coffee
<kilo> hi SachaS
<kilo> hi austria
<kilo> hi btami
<kilo> hi universe
<johannesV> :)
<kilo> coffee, that is a good thought
<kilo> in number(10,2) 2 positions of decimals should be?
<kilo> eg it should be 5.25
<johannesV> yes
<kilo> and if i enter eg 10  then it should be 10.00 or 10.0?
<SachaS> johannesV: appserver?
<SachaS> what changes?
<SachaS> company was introduces and somewhat removed agaiN?
<johannesV> well, we did a lot the last few weeks
<SachaS> introduced
<johannesV> it's not removed but replaced by a much more flexible thing
<johannesV> but besides this
<johannesV> OnInit () is now working as intended :)
<kilo> SachaS: automatic timestamping...
<SachaS> can u access properties?
<johannesV> yes
<SachaS> automatic timestamps, cool.
<johannesV> you even can set defaults there
<SachaS> OnInit does it work with accessing properties of super classes?
<johannesV> (which will be displayed in the form from start)
<SachaS> thats good news too
<johannesV> you mean in address.OnInit access foobar-props ?
<SachaS> yeah if the module of address extends the foobar modules address
<johannesV> yes, it should work, as long as you keep the following in mind:
<SachaS> somewhat wrong what I wrote ...
<johannesV> OnInit of the defining module runs first
<johannesV> all other modules having an OnInit for the same class run in arbitrary order
<SachaS> :(
<SachaS> "arbitrary"
<johannesV> this mean you can access properties as you like, but you cannot determine if another OnInit has set it's values
<SachaS> so if modules x extends modules y which extends module z
<johannesV> well, SachaS, there's no rule of thumb to say which module's OnInit for *one* class should run
<johannesV> SachaS, *modules* do NOT extend other *modules*
<johannesV> but
<johannesV> module foo defines a class "bar" (--> foo_bar)
<johannesV> and module "sachas" can extend that class "foo_bar"
<johannesV> so foo_bar.sachas_prop would be a valid property
<SachaS> so if class xx of module x extends class yy of module y which extends class zz of module z
<johannesV> but it is only *one* class, namely "foo_bar
<johannesV> SachaS, a class cannot extend another class
<johannesV> there is only *one* class, having a bunch of properties
<SachaS> z = base; y = finance; x = finance statistics
<johannesV> beeing introduced by one or *more* modules
<johannesV> ?
<SachaS> x needs class properties which modules y introduced
*** Mango_Madness has joined #gnuenterprise
<johannesV> does it ?
<johannesV> i mean, how can it know of them ?
<SachaS> yeah if I want to install a finance statistics module it must be made sure that the base and the finance modules are installed
<SachaS> guess you could have something like in the .deb files ...
<SachaS> setting the dependency
*** cilkay has quit IRC
<SachaS> well i do not know
<SachaS> but I was thinking that could be the case
<SachaS> if no dependency is set, it can be any order
<johannesV> SachaS, this dependency is given in the gcd-files already
<Mango_Madness> i just stumbled across this project; i have to say, it is the most interesting i've seen in quite some time
<SachaS> dont you think there could be a dependency?
<johannesV> SachaS, there *is* a dependency which is already given by GCD
<SachaS> what? by the order they are installed?
<johannesV> Mango_Madness, thanks  :)
<johannesV> SachaS, no by the code they define :)
<SachaS> or which class is extended
<johannesV> look at a gcd
<johannesV> giving a <module name="foo"><class name="bar"> ...
<johannesV> this defines a class "foo_bar"
<johannesV> using something like: <module name="SachaS"><class name="foo_bar"> ..
<johannesV> means: extend the class "foo_bar" (from foo module) by my own props (all beeing from the module 'SachaS')
<johannesV> so this *exactly* defines a dependency between foo_bar and foo_bar.SachaS_*
<SachaS> do you think there will be dependency which goes down more than 2 levels?
<johannesV> the latter one does *not* introduce a *new* class
<johannesV> it just adds properties to an existing class using another context !
<SachaS> johannesV i know :)
<SachaS> but
<johannesV> there are no levels
<SachaS> if there is a cool module
<SachaS> and you write another cool module which extends a class of the first cool module
<johannesV> we have no real OO here, because we must *flatten* everything to db-structures
<SachaS> and here I come and want to write another cool module which REUSES your cool module
<johannesV> no problem
<SachaS> which builds on top of your cool module
<SachaS> in the OnInit I might have to access your cool modules default values
<SachaS> so if the OnInit is arbirary
<SachaS> my uebercool module might fail to initilize your cool module
<SachaS> because the OnInit of my module is called second after the first cool module and before your cool module
<kilo> hi Mango_Madness
<johannesV> right
<Mango_Madness> hola
<SachaS> do you think that could become a problem?
<johannesV> SachaS, the only thing which is for sure is, that the def. module's OnInit comes first
<SachaS> well thats a start ;)
<johannesV> SachaS, no i don't think so yet
<johannesV> because in OnInit () usually no 'tricky' things happen
<johannesV> but only setting defaults
<SachaS> damn i am even using gnumeric to do expenses instead of a gnue app
<johannesV> if you define such code which is connected to a given module in such a manner (you described)
<johannesV> the question would be: why not modify that module directly ?
<SachaS> maybe because I do not want to touch your cool module
<johannesV> SachaS, shame on you ! *lol*
<SachaS> shame on you
<SachaS> :)
* SachaS fingers johannesV's cool module
<SachaS> ahmm
<johannesV> ok, after all, i got your point
<SachaS> ;)
<SachaS> yeah
<SachaS> its 1.00am
<SachaS> here
<johannesV> but i've to think if an additional attrib (of the class- or the procedure-tag) would solve such problems
<kilo> SachaS: gnumeric??? beware, i will say Ni on you
<SachaS> i was doing expenses with gnumeric on my sharp zaurus cs-860 in the plane ;)
<SachaS> was kind of cool
<SachaS> i thought
<SachaS> :)
<Mango_Madness> geek alert : )
<SachaS> was looking at word documents with abiword
<SachaS> the xpdf screwed up on pdfs becuase it misses some fonts, that sucked
<SachaS> other than that it was cool
<SachaS> reminds me to send the expense reports actually ;)
<kilo> expense report? 'Dear Sir, I have BIG expenses, sincerely, SachaS'
<SachaS> ;) kilo yes ;) i am in business now
<kilo> activity report "Dear Sir, I flew a big plane to 'merica, met derek (expensive night), hopped over to Vegas (expensive night), got new girlfriend(s) in California (expensive nights). yours truly SachaS"
<SachaS> ;) plus internet costs us 100
<SachaS> apropos derek I missed to see him. i think i went to the wrong soccer stadium
<SachaS> hope to catch up next week with derek
<SachaS> well going home dudes
<johannesV> ok, have a nice time SachaS
<kilo> going home? you mean it is 1 am and you're in the office?
<SachaS> yeah
<kilo> go home
<SachaS> ok night
<kilo> bye
<kilo> austria: gross = self.product.ITEM_vat.TAX_grossFromNet(self.valueNet)
<kilo> grossFromNet has 1 parameter
<kilo> and if i call it, get an error that i pass 2 parameters to it, though it expects only 1
<kilo> why?
<reinhard> you forgot to add self to parameter list?
<reinhard> err
<reinhard> forget it
<johannesV> using that self.product ...
<johannesV> this implies a 'self'-parameter
<reinhard> but every appserver procedure has a self parameter automatically
<reinhard> doesn't it?
<johannesV> dunno
<johannesV> i've just detected a big prob when pressing 'rollback'
<johannesV> this 'removes' the 'new instance' silently
<johannesV> and when user wants to commit again: *bang* instance ... does not exist
<kilo> should i add self to the param list???
<kilo> i thought it was dealt with automatically
<reinhard> kilo: no
<johannesV> kilo: i'm looking into it, just a sec
<kilo> ok, i stare at my coffee then for a while
<johannesV> i get the same error, that's good ...
<johannesV> just another sec
<reinhard> off to customer for the rest of the day
<reinhard> l8r all
*** reinhard has quit IRC
<kilo> johannesV: i have a bag full of things like this... 8-)))
<johannesV> kilo: got it solved
<kilo> oh, 't was quick
<johannesV> what's the name of the param of your grossFromNet function
<kilo> inNet
<johannesV> ok, then the proper syntax would be:
<johannesV>  gross = self.product.ITEM_vat.TAX_grossFromNet(inNet = self.valueNet)
<kilo> i think i see the reason...
<johannesV> the arguments to procedures are *always* keyword-args !
<johannesV> these get flattened into a paramlist in language-adapters
<kilo> somewhere we should write it down with CAPITALS
<johannesV> i think reinhard will do this in his dev-guide
<kilo> that is badly needed
<kilo> btw what's up with lekma???
<johannesV> dunno
<kilo> i would be very interested how his project is. and more, how the users like it
<kilo> johannesV: thx for solving this
<kilo> should i explain an other one?
<johannesV> if it's not too complicatd
<johannesV> i've to leave in about 30 min
<kilo> hmm, ok, maybe then it is a good one for the weekend 8)))
<kilo> a little bit hard to explain
<kilo> non-nullable properties are the main actors in this drama
<kilo> given an invoice
<kilo> it consists of invoice head and a number of invoice items (lines)
<kilo> invoice items have a property, Head, that contain the gnue_id of the invoice head they are attached to
<kilo> i think you can see it in the gcds i've sent you earlier
<johannesV> ok
<kilo> on the form i've set up a M/D relation, master=invoice head, detail=invoice items
<kilo> and on commit it gives an error that Head property is null...
<kilo> then i went further
<kilo> i have cascading forms, ie if a dropdown doesn't contain a given entry, one can open a form and expand the items
<kilo> if the property the dropdown is handling is non-nullable, i get an error when closing the expander form...
<johannesV> kilo, well, that m-/d-stuff should work automatically; so i've to dig into that bug
<johannesV> the latter one sounds cool
<johannesV> :)
<johannesV> haven't tried that
<johannesV> which uidriver ?
<kilo> for easier bug-search i will zip up my sample and send it to you later today
<kilo> gtk
<johannesV> ok, great
<kilo> complete with schema and forms
<johannesV> but i can look into this not before sunday (i'm leaving for germany today and i'm back on sunday)
<kilo> or hmmm, should i commit it to gnue-contrib?
<johannesV> as you like
<kilo> maybe i will wait with that...
<kilo> ok, will send it to you
<johannesV> ok
<johannesV> i'm off for the weekend now
<johannesV> have a nice day !
*** johannesV has quit IRC
<btami> bbl
*** btami has quit IRC
*** Mango_Madness has quit IRC
*** cilkay has joined #gnuenterprise
*** Mango_Madness has joined #gnuenterprise
*** kilo has quit IRC
*** jamest has joined #gnuenterprise
*** kilo has joined #gnuenterprise
*** dcmwai has joined #gnuenterprise
*** jcater has joined #gnuenterprise
*** holycow has quit IRC
*** mixi has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
*** sjc has joined #gnuenterprise
*** dcmwai_ has joined #gnuenterprise
*** dcmwai has quit IRC
*** holycow has joined #gnuenterprise
*** holycow has quit IRC
*** holycow has joined #gnuenterprise
*** dcmwai_ has quit IRC
*** SachaS has quit IRC
*** SachaS has joined #gnuenterprise
*** SachaS has quit IRC
*** SachaS has joined #gnuenterprise
*** Vee2d2 has quit IRC
*** Vee2d2 has joined #gnuenterprise
*** Ahewes has joined #gnuenterprise
*** jamest has quit IRC
*** kilo has quit IRC
*** sjc has quit IRC
*** jcater_ has joined #gnuenterprise
*** reinhard has quit IRC
*** jcater has quit IRC
*** holycow has quit IRC
*** MiXi^ has joined #gnuenterprise
*** mixi has quit IRC
*** Ahewes has quit IRC
*** SachaS has quit IRC
*** jcater_ has joined #gnuenterprise
*** Mango_Madness has quit IRC
*** jcater has quit IRC
*** Mango_Madness has joined #gnuenterprise