sacha: "The dependency that broke the camel's back" and "The schemecoder of Notre Dame" were both taken :) good night *** reinhard has joined #gnuenterprise *** johannesV has joined #gnuenterprise good morning *** btami has joined #gnuenterprise good morning good morning all I just can't believe gnue traffic is back! this is just so great *** johannesV has quit IRC *** johannesV has joined #gnuenterprise *** sjc has joined #gnuenterprise *** kilo has joined #gnuenterprise good morning reinhard: was this error fixed by you before? File "c:\python23\Lib\site-packages\gnue\common\rpc\drivers\xmlrpc\ClientAdapter.py", line 101, in request p.feed (data) File "c:\python23\lib\xmlrpclib.py", line 528, in feed self._parser.Parse(data, 0) File "c:\python23\lib\xmlrpclib.py", line 775, in end return f(self, join(self._data, "")) File "c:\python23\lib\xmlrpclib.py", line 816, in end_double self.append(float(data)) ValueError: invalid literal for float(): 78,79999999999999 decimal point versus decimal colon i can produce it, if add prequery="Y" to address_person in the sample.gfd I can't remember exactly hm, btami, the traceback also occurs if you just select all records (without prequery), doesn't it ? *** yure has joined #gnuenterprise if i select all records (without prequery), there is no traceback, but i get no records! (Python 2.3.5) hm, works fine for me will install hu_HU to double-check *** siesel has joined #gnuenterprise hm, btami it still works if i'm using hu_HU.UTF-8 as locale (although i cannot understand the menus anymore :)) to try it, you only have to change decimal point to colon, no hu_HU needed, at least i think... regional settings/customize/decimal signe johannesV: can you reproduce it on your XP ? haven't tried on xp just a sec ok, have changed the decimal-point to "." and it works just fine yes, i know, but we use "," btami: we usually have the decimal point at "," too so do i (usually) so i had to change it from a comma to a point to check :) btami: you run appserver and forms on same pc? yes johannesV: did you also test with appserver running on win? no, appserver is running on ubuntu ok I think it's a server side problem that scenario is ok server puts , in xmlrpc where it should be . the pb is appserver on XP can check it in the afternoon ok reinhard, fixed that encoding error in forms * johannesV off for lunch I'm not shure which encoding error you mean, but I have difficulties using wx2.6 with unicode. non ascii characters are not accepted when using utf-8 encoding (like äü or f.e. 我 ) they are displayed in the text box, but disappear when leaving the box. erm what does the last one stand for? *** sjc has quit IRC siesel, there's an error if someone enters german characters like äöü in query mode which occurs on F9 (execution of a query) I'm about to post the fix to svn No, I don't use the query mode. using such characters in normal operation is no problem nothing dissapears, everything gets stroed erm, stored I just change characters in edit mode kilo: chinese character, means I / me / myself johannesV: do you use de_DE.UTF-8 ? yeah i'm using ubuntu breezy with de_AT.UTF-8 using specials on xp, ubuntu and os x works just fine (across all os using the same backend) hmmm, I use dapper and you're not able to enter an ä for example ? can you paste a bit of the traceback ? ni. no I input the character, but when I leave the field, it disappears it doesn't raise an error ? that's strange you're using latest svn or package ? I fear, that the event subsystem "verschluckt" the character latest svn hm, but that would mean it could not work for me too ... ? but it does siesel, what about using another driver - just to doublecheck like gtk2 ? I will try copy and paste, but have to fix mysql driver first yes, gtk2 works ok. hum, that's strange ... that's really strange ... which wx are you using ? btw. what's the pb with mysql driver ? johannesV/siesel: i'm not too familiar with server/thread programming, maybe it's trivial... in the geasRpcserver 1 therad per adapter started in a for loop, but i see one adapter.shutdown in the except is it correct? line #235 wx version: 2.6.1.2ubuntu2 import wx; wx.PlatformInfo pb with mysql driver: charsets other than latin1 not supported at the moment. ('__WXGTK__', 'wxGTK', 'unicode', 'gtk2', 'wx-assertions-off') ah, i see well, the wx is nearly the same as here ... i'm using 2.6.1.1.pre maybe you should use a --debug-level=2 when running the form that should at least give some details about the __keypress handler (which is quite interesting here) I just get some DELETE and BACKSPACE events you mean COMMAND ? btami, too many threads running in parallel atm ... ? wrt your question about geasRpcserver ... i could have a look at that later, but i think it's ok ... long time since i did some work on that threading/forking stuff seems suspicious to me too johannesV: thx, bbl *** btami has quit IRC johannesV: ok, I fixed my local mysql instance and can test now. using dbg lvl 2 I just get normal commands not more so no lines like: DB002: 0:00:09.045 [entry:312] --> _keyress: 228 ? that looks like a bug in wx2.6.1.2 then ... which seems to not generate a wx.EVT_CHAR for such characters ... oth the characters seems to get displayed initially ... I dont get events for any other characters like asave and you're sure to use -u wx26 (not wx which is the old driver) so I wonder why it works for ascii code yeah oops I though you have changed it, because wx shows it uses version 2.6 in the about dialog. bah nevertheless, I got another error with wx26 that's only what's reported by wx.VERSION i think ... sorry no problem when I insert an umlaut it shows the keypress, but when I insert a chinese charater, no debugging message is send ok, additionally if I insert umlauts switching between different entries doesn't work alright ls what happens ? I use the zipcode example. Having 3 records in the table. prereq: the focus is in the first "city" field I click on the third "city" field cursor is transfered to the third field but keypresses still go to the first field marking the region also just affects the first field. all city names contain umlauts ok, does this also happens if you're using 7-bit values only ? но s/happens/happen s/но/no/ ha If I jump to field of another entry, like from city to state or back, the right field will be selected can reproduce that, but the bug is in there no matter of characters so it does not work having 7-bit values only too i think it has something to do with the rows .. brb *** jamest has joined #gnuenterprise does anybody uses mysql 4.0? if yes, please test the latest svn mysql driver. if you have set the correct table/column charset, full unicode support should be available now. johannesV: concerning complex character input into wx26 driver: EVT_CHAR doesn't support this. Is there any other event which is listening to changes of the entry field itself? char input handling also isn't real good in dealing with ctrl- combos then inserting text via middle mousebutton would work too for the user defined triggers on keystrokes it'd be nice to find a way to deal with that as well yes, I fully agree :) *** jcater has left #gnuenterprise siesel, there is EVT_TEXT but that is a bit different siesel, i wander why you encounter troubles with EVT_CHAR, as i'm using nearly the same setup and it's working here erm i mean wonder :) umlaute are working for wx26, but complex characters are not working ah, ok you mean double-byte characters complex characters = jap. chin. etc. characters inserted via XIM (X input method) as ä is a double-byte charater when you use utf-8, I tried to avoid the term does inserting character (normal ones) via middle mouse button works for you? is there any chance we can enter such a character at all without buying a new keyboard? just a sec :) yes, siesel, that works fine *** kilo has left #gnuenterprise strange as it don't work for me reinhard: yes, you can produce theses characters by switching the gtk input method with a right click, you should find a submenu "input methods" in the context menu yes cyrillic should be fine too кйхуиязуерхфащ воа if I use cyrillic in gtk2, I can really see these characters appearing... woa боа didn't know that yeah, all is working almost out of the box now. and by installing uim you will also get asian input methods nevertheless, if i try to input russian into the wx26 uidriver forms, I just get latin characters instead of russian ok this should help to reproduce the problem or at least part of it yeːs хмм it seems the wx26-entry widget is completely ignoring a change of the input-method ሎል as it works fine with -u gtk2 at least it looks like i can check if this is a gnue-forms related problem or a wx-problem as such that would be great ok, going to sleep now. cu guys *** siesel has quit IRC FWIW, I think we should make more use of our bug tracker as well as of the bug trackers of our dependencies for example the "lower does not work with umlauts with sqlite" should be documented in our bug tracker (with a gnue specific description like "case insensitive search does not work with non-ascii characters with sqlite backend") and we should file a bug for sqlite, and we should place a link to that bug in our issue *** derek has quit IRC *** jcater has joined #gnuenterprise *** derek has joined #gnuenterprise reinhard, have you filed that bug regarding collation ? no sigh is it just me you know when I say "we" I usually mean "johannes" ;-) or are we getting more spam on gnue-commits than actual commits? I think you are right however I think for a commit list we could use sender filters perfectly hmm, good point do you know if they support domain-level filters? like, .*@gnuenterprise.org ? reinhard: do you do much mailman administration? I'm looking on the admin page and damned if I can find anything like that hmm I do see topic filters though *** btami has joined #gnuenterprise jcater: let me look I found it ok I would use sender filters and allow all *@gnuenterprise.org and simply discard the rest well, I found "List of non-member addresses whose postings should be automatically accepted. (Details for accept_these_nonmembers)" to which I added "^.*@gnuenterprise.org" yeah right, I meant exactly that (the help says to start the line with a ^ to denote a regex) and at the very bottom of the same page, you can tell to discard for all the rest okay, done while you are at it, you could remove the 40 kb size limit, we had some commits in the past where we hit that limit hmm max_message_size is set to 0 I wonder if it's a transport-level setting gnu.org set up? ok, nevermind hm, i cannot break that form using the jump-to-record *** yure has quit IRC *** sjc has joined #gnuenterprise *** jamest has left #gnuenterprise *** jamest has joined #gnuenterprise *** klasstek has joined #gnuenterprise johannesV: don't worry, this bug is really not critical *** ajmitch__ has joined #gnuenterprise hey now stop fixing the mailing lists it is how i make sure my mail is functioning ;) *** ajmitch has quit IRC off for today *** reinhard has quit IRC *** johannesV_ has joined #gnuenterprise *** johannesV has quit IRC *** jcater has quit IRC *** bigbrother has quit IRC *** jcater has joined #gnuenterprise *** bigbrother has joined #gnuenterprise *** jamest has left #gnuenterprise *** jamest has joined #gnuenterprise *** sacha_ has joined #gnuenterprise *** sacha has quit IRC *** yure has joined #gnuenterprise *** yure has quit IRC *** yure has joined #gnuenterprise *** sacha_ has quit IRC anyone stateside interested in working for google in phoenix? better than google in SF bay area well except for the hotness its not hot its relative derek: doing what? writing billing systems they are developing all their internal business applications here in phoenix financial, billing, hr, etc i think the biggest focus is on billing systems at the moment they asked me to come down and visit, but right now too committed to own venture they asked me to send them any enterprise developers i might know figured i would ask here :) *** johannesV_ has quit IRC *** psu has joined #gnuenterprise *** jcater has quit IRC *** jcater has joined #gnuenterprise *** btami has quit IRC *** ajmitch__ is now known as ajmitch *** psu has quit IRC jcater: have you seen EnterpriseDB at all? yeah using it at all? no was in postgres asking others about different postgres issue and they brought up that enterpriseDB is almost fully oracle compatiable now yeah but costs as much as oracle in fact, sony just announced they are switching to it and threw a ton of money at it jcater: looks free to me :) no create a nice little cluster farm and SAN device and you are good to go a one-cpu version if free for developer purposes, or if you db is less than a few gb *** krizz_ has joined #gnuenterprise at $1,000 per cpu how does that stack up to oracle? jcater: you doing any failover or loadbalancing on postgres? my support contract is $3k/yr for 4 cpus + clustering fwiw, all new stuff is on postgres I don't w2ant to invest any more energy into the old oracle stuff :) I've beat my head against the desk on that issue rofl the desk, the filing cabinet, the floor, the sink and the server rack... i suspect I'm 100% vested in postgres here now, though several major departments run on postgres just in... postgres purchased by oracle.... * derek runs really really fast and hides ;) btw: that would be my kind of luck.. spend time moving off one platform to another only to have former buy later I'm waiting for the announcement :) well, at least I didn't go sap-db :) which I was close to *** yure has quit IRC *** derek has quit IRC *** derek has joined #gnuenterprise *** klasstek has quit IRC *** jamest has left #gnuenterprise *** derek has quit IRC *** derek has joined #gnuenterprise *** sjc has quit IRC *** sacha has joined #gnuenterprise