*** kilo has joined #gnuenterprise *** SachaS has quit IRC *** johannesV has joined #gnuenterprise good morning morning johannesV did you read the logs?? good morning morning kilo happy birthday again :) ah, thank yuo again 8-)) happy birthday kilo ! :) lekma, i'm just reading it thx johannesV * johannesV just wrote down lekma's first bug he found ... first one i found??? hmmm... :) (in the log yesterday ...) ok, lekma i'll dig into that fetch () right now ... stay tuned :) *** sjc has joined #gnuenterprise johannesV: thx got it ... will fix it now ... bah, changing the sort-elements to dictionaries was a quite bad idea .... *grr* blame that Austrian guy... 8-)) *** reinhard has joined #gnuenterprise ok, looks like it's working again ... but i need a bit more testing (with dirty-reads and the like ....) kilo: congrats! how old are you now? very very old, i dont have whisky that is older than me... oldest one i own is 30 yrs... so with little maths anyone can compute that i am 34 ahhh you're still so young * reinhard is 34 1/2 ;-) lol *** btami has joined #gnuenterprise lekma, please svn update and re-try your tests ... i can't right now people are working on my test env :( at noon lunch time i will ok, np i'm sorry i haven't though of the cache-merging before ... reinhard: what about your demo the other day? did it work ok? kilo: yes, thanks clients on windows, server onlinux? yes worked perfectly apart from some mysql issues (which we sorted out during the demo) in fact we were able to set up a similar thing without any problems here *** jproby has joined #gnuenterprise *** ajmitch has quit IRC *** ajmitch has joined #gnuenterprise *** sjc has quit IRC do we hit revision 7000 today ? *lol* we will in fact, it looks like i've done everything i was told for today ... :) johannesV: re --connection did you also change the sample scripts? and can you please grep for "--database" through the doc/ directory? the sample-scripts ? you mean setup-db.sh ? they're all using --connection ... :) hi all hi ajmitch reinhard, if i have something like 'connection = foobar' in my gnue.conf running gcdcvs sample.gcd should be using the 'foobar' connection, right ? imho yes (so this is a bug of readgcd/gld then) since it does not right now * johannesV will fix that *** sjc has joined #gnuenterprise *** jproby has left #gnuenterprise reinhard: i'm trying to write a startup script for gnue-appserver putting itself into a taskbar icon without a console window open but some print's to console block me to do it may i put this 2 startup message into silent any way ? a --silent option would be great I would be ok with a --silent option but take care newest versions of appserver also output class repository info *** sjc has joined #gnuenterprise alternatively you could just > /dev/null reinhard: i correct myself no need for --silent if i start my script from an icon it's ok so, forget what i asked ok bbl *** aries has joined #gnuenterprise hi "Added '--connection' option to appserver and gnue-gsdgen and made '--database' to be depreciated (but still working)" why is it depreciated? I did committ 7000! :-P aries: for consistency all other tools except appserver have always used --connection appserver has so far used --database reinhard: I think the whole gnue should be use one, at most two db but on the stable environment only one for future aries: not sure if I understand you I don't see why we should limit users to a single database now all of the modules/contribs are sepearated to db-es but most of the cases, we should combine them (eg. packages) you can run them in separate db's or in the same db as you like i think, most of the users are not too experienced in the cli environment, or in the GNUe's internals we should provide compact packages which works, an almost complete solutions. they don't care about programning magic, but the look&feel, and the usefulness aries: yes, exactly aries: we have more or less the unwritten law that the database connection is named "gnue" and the appserver connection is named "appserver" my oppinion is that, now GNUe is full of magic, but there's now really usable module which use it at some points I believe that even the debian packages will create the connections.conf automatically i think we should change this aries: excellent aries: we are happy about anyone helping in package development bbl reinhard: ok, but you should improve the docs :) (no, I don't have exact questions now :) hi aries kilo: szia! bye *** aries has quit IRC *** sjc has quit IRC *** kilo has quit IRC *** jamest has joined #gnuenterprise *** johannesV has quit IRC johannesV: the sort ignorecase feature is working well thx a lot well hmmm... i may have been a bit fast *** sjc has joined #gnuenterprise "Équipe tournage" is at the end of the list after "VHS vierge" unicode hitting again?? most probably DB000: 0:03:44.464 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.492 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.522 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.550 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.579 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.607 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.637 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.665 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) DB000: 0:03:44.694 [DataObject:150] WARNING: non-unicode passed to the dbdriver (fr) and tons of this in debug level 1 is the new debug level system already in ?? no, this has been there for ages the debug system is setup on the new breakdowns but the code doesn't reflect it just do a --debug-level=1-100 to get all stuff lekma: the problem is the sorting that probably made no sense :) É is a higher ascii code than E but funny enough I just talked with johannes today on the phone that the problem with ó ò ö ô etc will most probably be the next issue we face with sorting lol :) however this is close to unsolvable for example in German, you sort Österreich at the place of O in Swedish, they sort Öberg after the Z hmmm what db's does gnu-schema support? hmmm locale.strcoll seems to do exactly what we want... *** dsmith has joined #gnuenterprise *** ajmitch_ has joined #gnuenterprise lekma: how would you tell the SQL server to sort according to the locale? AFAICT postgres sorts Ö after z no matter if you order by lower(x) or by just x lekma is using MS SQL Server? no postgresql *** ajmitch has quit IRC in postgresql (if memory serves) it's linked with the way you initdb which is strange cause i thought i did it well :( i have to find out if... but i think the pb now lies in db not in gnue lekma: you sure? i think you can tel pg which charset to use nope lekma: some parts of the sorting are done by gnue so you can just select directly to the db and see how it sorts if you sort by a calculated field, appserver does the sorting also if you have dirty records ok Howdy gnueers. I was just curious to see if you were still having subversion problems, or what did you do to fix things. where the field you order by has uncommitted changes hi dsmith reinhard: Hey AFAICT we have no problems with svn anmyore COol. not entirely true but close :) Not cool. (but jamest and jcater would be the men to tell) jamest: Did you move to the other backend? no i think an update took care of a lot of our issues Ok. i had 1 issue in the last month where it told me to recover the db i tried to recover and that also told me to recover Yoiks whcih made me a tad nervous :) and a 2nd run, and everything was fine but as long as people like reinhard aren't seeing the issues they used to see then I'm happy *** tiredbones has joined #gnuenterprise and I'm running 100% error free at work (less commits than gnue) *** sjc has quit IRC for some reason, running svn .debs on ash is a no-no but running source compiles is perfectly stable that's what I'm seeing which makes no sense, but I'm not tracking it down :) Intresting. I run from .debs at work though so no clue why this would be Yeah, me too. We've had no problems. must be some cruft from a long time ago somewhere than I Need to delete jcater: btw, as we are talking about ash But we don't stress things like you guys do. the daily builds seem to not work ah btw another one is viewcvs Traceback (most recent call last): File "/usr/lib/python2.3/site-packages/viewcvs/viewcvs.py", line 3235, in main request.run_viewcvs() File "/usr/lib/python2.3/site-packages/viewcvs/viewcvs.py", line 268, in run_viewcvs import vclib.svn File "/usr/lib/python2.3/site-packages/viewcvs/vclib/svn/__init__.py", line 27, in ? from svn import fs, repos, core, delta ImportError: No module named svn *** sjc has joined #gnuenterprise http://www.gnuenterprise.org/cgi-bin/viewcvs.cgi/gnue/trunk/ btami: hi, will you release last updates for win32? when+ i did it oh * mnemoc slaps himself thanks :D np have to go, bbl *** btami has quit IRC runtime-env doesn't include common anymore? :\ mnemoc: never did, AFAICT last week, when i installed it on my boss machine there wasn't a gnue-common.exe package and gnue-forms worked :) Another thing, what's the name of that tool that makes standalone .exe's from python? there are 2 mcmillian (sp?) and py2exe which i think is what btami uses to winderfy gnue py2exe rings a bell Have great sorrow for me. I'm stuck doing vb under w2k nowadays. :( :( ok, about to change the debug levels in common's code last chance to speak up against go for it *** titopbs has joined #gnuenterprise *** kilo has joined #gnuenterprise *** sjc has quit IRC *** wendall911 has joined #gnuenterprise reinhard, you already started on that? I wonder if we should add constants to python namespace for debug levels, so it's clear for new coders what each means (and old coders who haven't been in there in a year) *cough* not that I know any of those i.e., DEBUG_INFO, DEBUG_SQL, DEBUG_WARN *or* ah, nevermind *** sjc has joined #gnuenterprise *** johannesV has joined #gnuenterprise hi so we hit that DB000: 0:00:42.313 [DataObject:150] WARNING: non-unicode passed to the dbdriver (AT) again ? :) oh, i need to say --debug-level=0-2 instead of --debug-level=2 now ... :) *** sjc has quit IRC *** sjc has joined #gnuenterprise reinhard, looks like there's something wrong with conditions ... (not functionally but having to do with that unicode-warning) aha the improved debugging scheme shows its first fruits ;-) well, if i create a language-interface app doing a simple find ('address_country', {'address_code': u'AT'}) i get the above warning and that shouldn't be there ... yes, looks strange need some more time to investigate though ... *lol* funny bug in gdebug code debug-level=0-3 shows levels 0, 1, and 2 but not 3 whoops you fix or should I? forgot to +1 the range fixed it and committed it off for today bye all *** reinhard has quit IRC uhu ... looks like we've got a problem in xml-rpc passing dictionaries ... unicode-type values get changed into stringtype values ... bye *** lekma has quit IRC *** SachaS has joined #gnuenterprise hi everyone hi kilo apparently a topic with sap is MDM standing for masterdata management the problem if you have 10 thousands of products, or variants that you get a mess and that you have a product or variant entered several times and outdated variants and dead variants etc. etc so apparently what they try is at the time when the user creates an item (or customer or item, or supplier) that the system intelligently checks the given database to try to find if there is a similar item entered already and to prompt the user to check if he/she is not rentering a give item. so something for your packages to consider ;) *** titopbs has quit IRC i guess these problems apply to other erp solutions as well boy have i seen that problem it happens with shipping addresses for a customer alt address i.e. 1 order for 250 printers deliver to 220 addresses, the entry clerks dont check for existing its just 2 difficult data cleaning afterwards is also needed people use the address fields differently and the addressee does not always give the correct full addr + commonly applications dont give enough space or format for UK addresses hehe /rant *** sjc has quit IRC *** sjc has joined #gnuenterprise I'm trying to test appserver by running the module sample.gfd, It doesn't seem to br in the directory. Is there another module I can use? I think I found one. I ran this command: gnue-forms appserver://appserver/form//address_person and got the following error - Gtk-Warning ** cannot open display: 0.0 you running remote? ahh, gotcha .. found a bug in xmlrpc.typeconv looks like tiredbones runs the command in a terminal-window as that's a sigh your DISPLAY is set to something you can't actually access So, I don't run in xterm? sure you can do a echo $DISPLAY in an xterm the reply from echo $DISPLAY is :0.0 is that from a freshly opened xterm? yes. try running xwininfo from that xterm the cursor should change and it'll ask you to click in a window doesn't matter where you click, though clicking on the background will give you your resolution and color depth and confirm that your X is allowing gui apps to display properly What option should I give with xwininfo? none just xwininfo tiredbones, do you run the command on another machine ? like having an xterm running an ssh/telnet to a remote computer ... ? no or xev would be another one to try all we're after is verifying that your X is allowing connections and these apps are tiny and usually always installed xwininfo just gives me a list of options. vs say typing gimp :) when I type xev I get , xev unable to to display 0.0 s/to/open/ another part of the last error is Xlib: connection to 0.0 refuse be server ok, this means something is messed up with your X setup Xlib: No protocol specified if you are running in a GUI as the same user that started up the gui on the same machine I'm running in a gui hmm, is this a bug in python ? the keys of keyword-arguments must not be unicode, like def foo (**kwargs): ... and doing something like: foo (**{u'fred': 'bar'}) if you've done any su on this xterm I have not done su. is this computer stand alone? I do have 3 xterm running. and what OS are you running yes with what desktop gnome this is sarge dis *** chillywilly has quit IRC tiredbones: this may sound like a cop out but I'd be tempted to stop X and restart it as something is wack while stopped I'd log into my account *** chillywilly has joined #gnuenterprise and remove the .Xauthority file if it exists How about a reboot if that's easier i won't promise it fixes this but what you're seeing is odd *** tiredbones has quit IRC *** tiredbones has joined #gnuenterprise *** sjc has quit IRC ok, I ran the command again, gnue-forms appserver://appserver/form/address_person, I have a huge traceback. tiredbones, gimme the last few lines if from ConfigParser.py line 240 raise NoSectionError NoSectionError: No Section: appserver *** sjc has joined #gnuenterprise tiredbones, are you using svn ? I have a button on the bottom No, debs hmm that I can reply "OK", I haven't yet tiredbones, you're using wx ? I want to see if you want to see the error. I'm using python, I think tiredbones, well, of course you're using python :) but do you use wx as form's uidriver or gtk2 ? How do I check that? in your gnue.conf, what's the value of DefaultUI ? give me a sec. and if you're in gnue.conf ... do you have a secion [appserver] in there ? All DefaultUIs are commented out. What do you recommend? I do have a section [appserver] i think you're using wx then, and, if there's such a section i wonder why it can't be found but i've to say i've never played with the debs though ... It looks like everthing is commented out. that's no problem, since this means everything is set to it's default value I'm going to run the command again, and this time I'll answer ok to the error. let's see what happens. Nothing happen, It just ended. Is there an error log some place? tiredbones, try to do the following: gnue-forms -u gtk2 appserver://appserver/form/address_person you should get the same exception, but now you can copy and paste :) (if you've gtk2 availalbe) the error reads as follows An unexpected internal error has occur: No section : 'appserver' this means you found an error in GNUE. Please report to ... that's it tiredbones, press on the >> Detail button the detail gives me the same error. but it should contain a traceback, doesn't it ? NoSectionError: No Section: 'appserver' yes. It's the same traceback as before. only this time with a different screen. yeah, ok, but it's quite strange that it does not find the 'appserver' section ... do you have an appserver up and running ? the module in question is ConfigParser.py line 240 raise NoSectionError(section) AH!! (cause appserver would use the same config-section) I just enter the comman gnue-appserver and it ell me that it's already running. ok, so this is really weird ... can you try to open 'sample.gfd' ? you'll find that in appserver's sample directory *** holycow has joined #gnuenterprise