Action: chillywilly needs sleep Aiyee sleep needs me seems to me distutils is a lot bettter than it used to be it has support for scripts 64F outside! that rocks data_files Action: Vee2d2 goes to obliqe sleep with his presence nite night Action: SachaS did mirror (still is) irc logss and found the cvs command for gnue-sb without asking lol jamest (~jamest@adsl-64-216-106-144.dsl.tpkaks.swbell.net) left irc: "Client exiting" Action: alandd is away: Bed time! HTTrack mirror complete in 1 hours 35 minutes 23 seconds : 808 links scanned, 807 files written (29170503 bytes overall), no files updated [2959 3770 bytes received at 5171 bytes/sec], 1.7 requests per connection HTTrack mirror complete in 1 hours 35 minutes 23 seconds : 808 links scanned, 807 files written (29170503 bytes overall), no files updated [2959 3770 bytes received at 5171 bytes/sec], 1.7 requests per connection Nick change: SachaS -> SachaAway reinhard (~reinhard@M1264P004.adsl.highway.telekom.at) joined #gnuenterprise. dimas (~dimas@195.218.177.46) joined #gnuenterprise. reading logs hmmm wendall911 (~wendallc@198.31.172.218) left irc: Remote closed the connection siesel (~jan@xdsl-213-196-194-74.netcologne.de) joined #gnuenterprise. morning reinhard: you here? siesel (~jan@xdsl-213-196-194-74.netcologne.de) left irc: "Client exiting" siesel (~jan@xdsl-213-196-194-74.netcologne.de) joined #gnuenterprise. siesel (~jan@xdsl-213-196-194-74.netcologne.de) left irc: Client Quit dcmwai (~dcmwai@219.95.169.225) left irc: Remote closed the connection gypsymauro (~colorioma@host103-11.pool21757.interbusiness.it) joined #gnuenterprise. ciao Arturas (~arturas@gsk.vtu.lt) joined #gnuenterprise. Hello gypsymauro (~colorioma@host103-11.pool21757.interbusiness.it) left irc: "leaving" SachaS (~Sacha@dialup-196-168.wasp.net.au) joined #gnuenterprise. SachaS (~Sacha@dialup-196-168.wasp.net.au) left irc: Client Quit Nick change: SachaAway -> SachaS ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. thought to do some more work but feel to trired alandd (~alandd@h-66-167-132-163.PHNDAZ91.covad.net) left irc: "Client Exiting" gypsymauro (coloriomau@host7-26.pool80180.interbusiness.it) joined #gnuenterprise. hello ToyMan (~stuq@smtp.dstoys.com) left irc: Read error: 104 (Connection reset by peer) xwt.linux.gz not there :( wrong window icltlfatppl (~icltlfatp@157.22.13.232) left irc: "Client Exiting" jamest (~jamest@gw.math.ksu.edu) joined #gnuenterprise. ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. jamest: just commited the ultimate setup.py ;) works nicely with site-packages as well as gnue/lib/python works with --prefix, with --home and --root ToyMan (~stuq@smtp.dstoys.com) left irc: Excess Flood ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. checks gnue-appserver to be installed in the same directory as gnue-common and some nice features more appserver would be ready for release however sieses asked me to hold back until monday so he can finish his web-frontend and we can also put that into the release - yea, that simply rocks. I noticed he was commiting it all these commit messages of late are making me giddy.. reinhard: monday we'll have a new release of gnoe-designer too? gypsymauro: i wouldn't bet on that reinhard: cool is that across all the projects or just appserver? it is just appserver if you or jcater have time please tell me what's wrong with it as it would be too good if it really worked perfectly... wrong with it? and if you find nothing we can do the same principle for other tools there _must_ be something wrong otherwise we would have to take back that distutils suck ;-) wtf!?!?!?!?!?! what the hell is with all these lines that start with # rofl the basic idea is that i subclass install instead of build_scripts jamest: u mean in the designer? and in install i have the _real_ values for all installation target directories available so no matter wheter i used setup.cfg or --prefix or whatever self.install_lib is the path where .py files go i even can do ./setup.py install --prefix=/usr/local --install-lib=/there/goes/python and it works and the scripts are generated correctly (wrt search path) that's cool do you know if data_files is a drop in replacement for Data_Files data_files can't copy recursively that's the only difference i found that could be of interest to us honestly it took me 12 hours or more to build this setup.py it looks nice but now i very much believe that distutils doesn't suck at all you just have to tame it or learn how to use it ;) but you know auto* is so much more evil...hmm 12hrs...I think one could leanr autos* in 12 hrs too ;P Action: reinhard remembers good ol' geas configure.in reinhard: it would be great if you prove distutils doesn't suck our distutils usage is probably old school I know no noe wants to tell me but whsat is gnue-dbtools? ;P gypsymauro: no designer release monday chillywilly: it's tools that are handy for db stuff but not full featured like our "real" apps I'm putting in a text file importer gnue-loader and an sql loader? that is a very dumbed down gnue-integrator that doesn't do all the db to db w/ triggers and conversion jamest: if you want i can try to help with gnue-common setup.py for the next version reinhard: that's cool w/ me i hate the file but i would need your help with the special stuff in there s/the/that anyway I see what you mean...we write db tool scripts all the time for reiniting the db, etc. some special files like site-config and what's their purpose reinhard: ok, i'll have to re-read it myself chillywilly: jcater is adding gnue-sql maybe you might want to do a new release soon for jbailey reinhard: that's fine w/ me too if we can get the setup.py's in shape it's worth a release to make him happy 0.5.1 contains a missing import statement that makes gnue-common unusable when you don't have locales installed gack! .gld? another format? gnue loader def right I figure that figured i've got a flat file importer that uses a python dict to store file structure the USPS zip data http://savannah.gnu.org/cgi-bin/viewcvs/gnue/gnue-dbtools/samples/loader/usps/ctystate.gld?rev=1.1&content-type=text/vnd.viewcvs-markup I think there's a typo in that file it's unusable at this time well it's in xmlrpc so i guess it only affects appserver < extra '<' i just checked it in so I didn't lose everything if the HD fried while I sleep concider that file notes jamest: fwiw jcater's fix yesterday for jbailey didn't actually fix anyting reinhard: ok reinhard: if you've got setup.py mastered I'm all for cleaning them all up i imagine we can use appservers as a template because now, if it doesn't use the setup.cfg then it generates a traceback yes maybe chillywilly: i think you'll see the gnue-schema stuff go into gnue-dbtools as well as it's the only "app" in gnue-common jamest: there are also some files that are installed in a rather wrong position (IMHO) reinhard: ? like doc should go in $prefix/share/doc, not $prefix/doc images and all that should go in $prefix/share/gnue not $prefix/shared that's all fine w/ me and maybe some more but i'm not sure how to handle that as it would break compatibility maybe with 0.6.0 were's you get the idea to subclass install? where 'd that's the result of 10 hours trying, thinking and reading the source of distutils but mainly trial and error :) heh must bring kids to training bbl Nick change: reinhard -> rm-away that sounds funny "training" jcater (~jason@w202.z065105010.mem-tn.dsl.cnc.net) joined #gnuenterprise. hi jc /msg jcater it worked!!!! reinhard was suckered into fixing up a new setup.py style for gnue! =) cool he spent 12 hours w/ the distutils source poor fool asked us to look thru the setup.py in appserver to see if anything looked bad err, I mean that rocks the only thing I didn't like were all the lines starting the # they're unnatural we can grep those out for our setup.pys couldn't you make it more generic and write a function that has a list of mods that it tried to import and throw that into on of the hooks methods in the install child? was that english? one* i think it was chillish Action: Vee hugs chillywilly i wanted a basic setup.py framework in common that the other apps could import ;P so do I I was thinking about that last night though and how would that work w/building debs? as at that time, the stuff isn't actually getting installed is that our problem? lol true dat Action: jamest looks around for jbailey Action: chillywilly sets jamest's hair on fire HA!!!! I pitty the foo that's head is on fire i don't have all that much to burn! hehe.. me either.. damn monitor emissions actually it's all there just thin so it'd be more of a flash fire I shaved most of mine off we're talking head not back my wife wouldnt let me ToyMan (~stuq@smtp.dstoys.com) left irc: Excess Flood ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. lupo__ (lupo@pD954256E.dip.t-dialin.net) left irc: Remote closed the connection Arturas (~arturas@gsk.vtu.lt) left irc: "Bye :)" Nick change: rm-away -> reinhard ToyMan (~stuq@smtp.dstoys.com) left irc: Excess Flood ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. dsmith (~dsmith@mail.actron.com) joined #gnuenterprise. if I have installed gnue from cvs but with the setup.py installation procedure in each tools directory and then I am going to do a setup-cvs.py in gnue-common will that cause a problem? it shouldn't cvs setup installs into your home dir ok. then run the tools from ~/bin/g* it will call the code from the cvs repo will try that. :) gypsymauro (coloriomau@host7-26.pool80180.interbusiness.it) left irc: "leaving" is gnue-dbtools a new module? yep damn its hard to keep up here ;) sigh mysql> create table test (i1 int, i2 numeric(6,2) not null); Query OK, 0 rows affected (0.01 sec) mysql> insert into test values (12345678901234,123456.23); Query OK, 1 row affected (0.01 sec) mysql> select * from test; +------------+----------+ | i1 | i2 | +------------+----------+ | 2147483647 | 99999.99 | +------------+----------+ 1 row in set (0.00 sec) posted on the postgresql mailing list huh? if you try to put a value into a field that is larger than it can hold mysql silently drops your value and sets the field to the largest value the field can hold no error message or anything? nothing at least thats the word on the pgsql mailing list cool hi reinhard i'd make my debts numeric(2,2) hi SachaS brb Everyone knows mysql is evil. If you use it, it's because you Don't Care. what? what kills me is people do finacial stuff w/ mysql use the proper type just today postgresql is puking on an invoice a the part time place because their lame ass dbadmin put in a numeric numeric(8,4) instead of (8,2) btw: i'm their lame ass dbadmin ;) it's scary to think what accounting would have done to me if the system went mysql on their data and where's their patch to fix this? how can I print a "connections" object in python nicely? print connections.toString or something? jbailey (~jbailey@atlas.fundserv.com) joined #gnuenterprise. jbailey: reinhard has been working on new setup.py system w00t! appservers is supposed to bring you great joy now or something oooo, he's still here he can tell you :) how to I prettyprint an object in python? print connections is there something connections.to_string ?? or ptinr connections.print-all-methods print `obj` ? connections is an object or some crap what you might want is print connections.__dict__ that shows you all properties and all methods cool. i try that reinhard: i am trying appserver but there's a "toString" type python method reinhard: in gnue-appserver/src/classrep/__init__.py should: jbailey: were you successful yesterday with packaging common? if not os.path.isfile (basedir + "repository.ini"): be: if not os.path.isfile (basedir + "repository.ini"): ? you mean if not os.path.isfile (basedir + "/repository.ini"): "/repository" yes yep reinhard: No, I ran into some further craziness. Just a sec. Major broken stuff here. trying gcvs test.py in gnue-appserver/src db is up SachaS: what's your E-Mail? reinhard: schlegel@cs.curtin.edu.au thanks ?? please update cvs, should be fixed yep SachaS: everything works? when I created the gnue database with the setup-pgsql.sh I had issues with the new gnue user so i dropped the user and created a new user with password :( unfortunately is the password hardcoded in testLoginHandler ... reinhard: Me update cvs? now i get another error i am chasing and is that for common, or appserver? jbailey: no, that was for SachaS DB000: File "/usr/local/src/gnue/gnue-appserver/src/classrep/Namespace.py", line 109, in splitName jbailey: if you are lucky you have a working setup.py for common this evening your time DB000: raise EInvalidName (name) DB000: classrep.Namespace.EInvalidName: '__nonzero__' is not a valid, fully qualified identifier. brb in 15 min reinhard: Ah.. That type of luck that means I have to do more work. =) back SachaS: you still here? reinhard yes you still get that error? new_classes = ClassDict (session, new_modules, module = new_modules ["gnue"]) def __init__ (self, session, moduleDict, module = None, definition = None): module = None .... does that overwrite or in case module is not given it sets it to None eg when you call the constructor with only 2 attributes yep still got that error the latter it's only a default ok. otherwise it wouldn't make much sense to make it a parameter anyway :) true on that baseDir + /repository thing can you give me more lines of the traceback? i've not read the code but there is a os.path.join os.path.join yes i know :) that is cross platform i use it extensively in setup.py you just remind me on that :) ToyMan (~stuq@smtp.dstoys.com) left irc: Excess Flood reinahrd how much traceback do you want? at least 10 lines ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. shall I paste here? you can paste in private or here i don't mind i've got nothing to hide :) DB000: File "/usr/local/src/gnue/gnue-appserver/src/geasSessionManager.py", line 127, in load DB000: return s.load (classname, obj_id_list, propertylist) DB000: File "/usr/local/src/gnue/gnue-appserver/src/geasSession.py", line 171, in load DB000: list = self._createList (classname, ["gnue_id"] + propertylist, []) DB000: File "/usr/local/src/gnue/gnue-appserver/src/geasSession.py", line 102, in _createList DB000: return geasList.geasList (self, classdef, propertylist, sortorder) DB000: File "/usr/local/src/gnue/gnue-appserver/src/geasList.py", line 49, in __init__ DB000: propertydef = self._classdef.properties [propertyname] DB000: File "/usr/local/src/gnue/gnue-appserver/src/classrep/Base.py", line 74, in __getitem__ DB000: return self.get (key) DB000: File "/usr/local/src/gnue/gnue-appserver/src/classrep/Property.py", line 101, in get DB000: (moduleName, propName) = splitName (name) DB000: File "/usr/local/src/gnue/gnue-appserver/src/classrep/Namespace.py", line 109, in splitName DB000: raise EInvalidName (name) DB000: classrep.Namespace.EInvalidName: '__nonzero__' is not a valid, fully qualified identifier. ok thanks can you pleas add in geasSessionManager.py before line 127 the line "print propertylist" and tell me what that gives you? ['__nonzero__'] hmmm i need even more traceback marcello (~marcello@adsl-249-73.38-151.net24.it) joined #gnuenterprise. can you paste the rest? hi, which version of python should i use to compile 0.5.1? i got common forms and designer running but pysablot seems to have problems. i'm with a powerpc/debian unstable marcello: 2.3 thank you jbailey ToyMan (~stuq@smtp.dstoys.com) left irc: Excess Flood marcello: You may get 0.5.1 packages tonight in unstable, though. ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. jbailey hehe such good timing to try to install from source huh? :) those packages are multiplatform, i guess? yup and, to uninstall gnue i should just rm /usr/local/gnue/*? You'd have to check with the others. I've never done a standard gnue install. k thanks mouns (mouns@kali.mouns.org) joined #gnuenterprise. gnue? tonight? in unstable? jamest: you know what this means, right? we need to release tomorrow actually someone suggested we release soon with the setup improvements to help out jbailey plus whatever new bugs you've guys been diligently working on designer isn't particularly in a releasable state iirc well, at least it has my Focus Order button that doesn't work not that a non-functioning button has ever stopped me before change it's label to "dont press this!" ;) lol I need to uncomment the code that segfaults when you press that button and also change the label temporarily to a Jolly Roger then it'd be perfectly functioning or a pic of dynamite "segfault" or something similar "quit" ;) i have a bad bad feeling and if that becomes true there won't be much work on setup.py tonight :(( jamest: I can say cvs for now. reinhard: oh so you'll tease us with a little setup goodness and leave us hanging eh? :) should we try to release soon for jbailey's packages>? jcater: How 'bout after I get the packages going so that we can make sure it's all good. ok I'd personally like to avoid getting back in the habit of packaging up CVS but certainly until you get the process down Yeah. Like I don't want you to have to release 3 times next week. =) marcello (~marcello@adsl-249-73.38-151.net24.it) left irc: Read error: 113 (No route to host) I wouldn't worry about that we strive for at least 3 a year Make your quota in a week, and then relax. how can i tell debian to use python 2.3 as default version? Debian unstable does already. ToyMan (~stuq@smtp.dstoys.com) left irc: Excess Flood and testing? update-alternatives --config python i have both installed testing is blocked on gnue. ToyMan (~stuq@smtp.dstoys.com) joined #gnuenterprise. tamas (~btami@ngprs.pannongsm.hu) joined #gnuenterprise. Seriously. As soon as gnue is 2.3, it will be default in testing. nope jbailey: it's just for me to try something I've, erm.. Had some stern emails about it. =) besides derek? :) Yeah. Like the python DDs who want python 2.3 in sarge. =) when is sarge slated to be released?