*** krizz has quit IRC
*** johannesV has joined #gnuenterprise
*** yure has joined #gnuenterprise
<johannesV> for the logs: jamest, looks like i've finally managed to create a postgresdb without using createdb/createuser but with our common-based driver; this means using two different sets of credentials for owner and connecting user is no longer a problem ...
<johannesV> i'll finish this work on Behavior.py today in the evening or at least tomorrow ... sorry for any inconveniences this might has caused ...
<johannesV> bbl
*** jamest has joined #gnuenterprise
*** jamest has quit IRC
*** johannesV has quit IRC
*** jamest has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
<jamest> hmmm
<jamest> jamest@calvin:~/svn/gnue/gnue-common$ gfdes
<jamest> DB000:    File "/home/jamest/svn/gnue/.cvsdevelbase/gnue/common/external/plex/Machines.py", line 14, in ?
<jamest> DB000:     from Transitions import TransitionMap
<jamest> DB000:  SyntaxError: Invalid syntax.  Assignment to None. (Transitions.py, line 85)
<jamest> is anyone running gnue under ubuntu dapper?
<jamest> i guess no one is running GNUe under python 2.4?
*** btami has joined #gnuenterprise
<jamest> hi btami
*** gerold has quit IRC
<btami> hi jamest
<btami> have you seen the gnue-invoice sample app in the gnue-contrib repo ?
<jamest> no
<jamest> sigh
<jamest> that would have been nice :)
<btami> seems it's an appserver like part of your new sample app in gnue-samples
<btami> but it depends on gnue-packages a bit, as i know
<btami> developed by kilo
<btami> but it's stalled....
<jamest> all i'm after is a testkit for gnue
<jamest> not really something fleshed out completely
<jamest> but something we can all use in unit tests, samples, etc
<btami> i see
<jamest> reinhard and I talked a bit about cleaning up the samples
<jamest> and making something standardized
<jamest> i see your name in authors
<jamest> how far along is this?
<jamest> as i wanted to create an appserver class def to match the two-tier stuff
<btami> i just tried to apply my char based report filter
<btami> and wrote a .grd
<jamest> do you use gnue-reports?
<btami> just want to...
<btami> but i'ts a bit behind other gnue tools
<btami> no designer support is the most wanted feature for me
<btami> s/no/
<btami> but i'm not good enough to code that in wx
<btami> and had 0 time in the last 2-3 moths for gnueing :(
<jamest> i think I know the feeling
<jamest> :)
<btami> but seeing jcater, jan,lekma,psu and you again is a raising sun feeling :)
<jamest> :)
<jamest> you full timer gnue'rs aren't nervous about us messing things up after reinhard and johannesV and the rest of you just got things cleaned up?
<btami> it was only johannesV and reinhard (cleaning up  things)
<jamest> and adding lots of features
<btami> yep
<jamest> we standardize on python 2.3 right?
<btami> yes
*** lupo__ has joined #gnuenterprise
*** yure has quit IRC
*** btami has quit IRC
<johannesV> jamest, i'm running gnue using python2.4
<jamest> really?
<jamest> nothing worked until I changed our plex library
<jamest> is was generating a warning in 2.3 and a full traceback in 2.4.2
<johannesV> hm, strange ... where do we use that plex library ?
<jamest> in the currently unused input masks
<jamest> at a minimum
<johannesV> ah, i see
<johannesV> so i had no situation where i've been using an input-mask then ...
<jamest> it was in gfdes
<johannesV> which one ?
<jamest> i couldn't run gfcvs or gfdes at all
<johannesV> hm, that's strange as i can do it quite perfectly :)
<jamest> which python?
<jamest> go to gnue-common/src/external/plex
<jamest> fire up python
<jamest> then import Transitions
<jamest> line 85 should at least give a warning if you haven't svn up
<johannesV> SyntaxError: Invalid syntax.  Assignment to None.
<johannesV> Python 2.4.2 (#2, Sep 30 2005, 21:19:01)
<jamest> that was the error I got when running gfdes or gfcvs
<jamest> even w/o a file name
<jamest> it's fixed in head
<johannesV> haven't updated svn till now
<johannesV> but i'm able to start both, designer and forms
<jamest> that is wack :)
<jamest> i migrated from debian to kubuntu at home last night
<jamest> as I was sick of fighting debian dual head setup
<jamest> and kubuntu (almost) just works
<jamest> and hit that bug with a fresh checkout of gnue
<johannesV> hm, that's strange
<johannesV> i can't see where that masks-stuff is used in forms ... ?
<jamest> because it's only in the unit tests in common
<jamest> i never commited the forms stuff as it is wack
<johannesV> so how does this affect gnue-forms then ?
<jamest> it doesn't
<jamest> ooooo
<jamest> my forms still has uncommited input mask stuff
<jamest> which would make plex import even in designer as it uses forms ui dirvers
<jamest> so no one would have seen this :)
<jamest> do you know where the table create/alter statements are sorted for depedencies
<johannesV> you mean in gnue-schema ?
<jamest> i've found data in readgsd and i'm in DBSIG2.Behaviour now looking
<jamest> yes
<jamest> the two-tier sample I commited breaks gnue-schema
<jamest> it creates the PK FK alter table statements in incorrect order
<jamest> so I'm trying to find where those are ordered properly
<johannesV> it should be sorted in readgsd.py (line 760)
<jamest> um
<jamest> my readgsd.py only goes to 759
<johannesV> it's quite complicated and it took me several hours to do that
<jamest> line 759 is
<johannesV> oh, sorry
<jamest>   gsdClient ().run ()
<johannesV> 370
<jamest> that's for inportDate
<jamest> importData
<jamest> this is an importSchema error I'm fairly certain
<jamest> if you go into samples/two-tier/schema
<jamest> and do a gsscvs -c gnue *.gsd
<johannesV> ah, hm, man .... one could think i'm a bit dumb ...
<jamest> it'll try to create the FK reference from invoice_items.sku to items.sku
<jamest> but it hasn't created the items.sku unique constraint yet
<jamest> so postgresql pukes
<jamest> i *think* all I'd need to do it make sure unique constrains are done prior to other alter tables
<jamest> but I can't find where that's sorted
<jamest>  .... one could think I'M a bit dumb ...
<jamest> :)
<johannesV> hm, that key being referenced is (city, state, zip) right ?
<johannesV> a key having that order does not exist in zipcode table, does it ?
<jamest> that's fine
<johannesV> if that order doesn't matter that reference should match the pk then
<jamest> gsscvs -c gnue -u dbadmin -p <snip> -o out.sql *.gsd
<jamest> readgsd: ERROR:  there is no unique constraint matching given keys for referenced table "items"
<jamest> ALTER TABLE invoice_item ADD CONSTRAINT fk_invoice_item_sku FOREIGN KEY (sku) REFERENCES items (sku)
<jamest> if I --file-only -o out.sql
<johannesV> ah, maybe i've got an old zipcode table installed ...
<johannesV> ok, got it reproduce
<johannesV> s/reproduce/reproduced
<jamest> line 20 is
<jamest> ALTER TABLE invoice_item ADD CONSTRAINT fk_invoice_item_sku FOREIGN KEY (sku) REFERENCES items (sku)
<johannesV> yes
<jamest> line 21 is
<jamest> ALTER TABLE items ADD CONSTRAINT unique_sku UNIQUE (sku)
<johannesV> no no no ...
<johannesV> don't use unique
<johannesV> but do something like this:
<johannesV> just a sec ...
<jamest> but, but, but!
* johannesV searching dtd
<jamest> i read the docs!
<jamest> well, the files ending in .py
<jamest> and unique seemed right! :)
<jamest> that does remind me that technote 00012.txt mentions a type="key" which doesn't work at all
<johannesV> we need indexes
<johannesV> damn, where's that dtd
<johannesV> yep
<johannesV> that works
<jamest> i've been using the xmlElements in GSchema.py
<johannesV> <indexes>
<johannesV>         <index name="unique_sku" unique="True">
<johannesV>           <indexfield name="sku"/>
<johannesV>         </index>
<johannesV>       </indexes>
<jamest> as my docs
<johannesV> instead of that <constraint>
<johannesV> brb
<jamest> if you changed that .gsd file can you commit it
<johannesV> please run gnue-dtd and build it for schema (or gsd ... ) (not sure right now)
<johannesV> ok
<jamest> perfect, thanks
<johannesV> no problem
<johannesV> btw. i've changed postgres.Behavior to no longer use createdb and createuser scripts
<johannesV> but to do things directly using template1
<johannesV> as there was no chance to specify a password while using createdb or createuser
<johannesV> ok, have to run now
<johannesV> have a nice day/evening
*** johannesV has quit IRC
<ibr> dimas: Thanks much!
<ibr> Bye!
*** ibr has quit IRC
*** chillywilly has quit IRC
*** SachaS has quit IRC
*** SachaS has joined #gnuenterprise
*** chillywilly has joined #gnuenterprise
*** yure has joined #gnuenterprise
*** yure has quit IRC
*** yure has joined #gnuenterprise
*** kilo has joined #gnuenterprise
*** yure has quit IRC
*** lupo__ has quit IRC
*** kilo has left #gnuenterprise
<jamest> i've added a thing called objectWrapper to gnue-contrib
<jamest> it's a port of something I wrote for work that is in the spirit of SQL*Objects
<jamest> but uses GNUe datasources for persistance
*** jamest has left #gnuenterprise
*** chillywilly has quit IRC
*** jcater has left #gnuenterprise