and how is foo feeling? lonely :) :) *** btami has joined #gnuenterprise *** dalen has joined #gnuenterprise hi everyone I'm trying to get the appserver running but I'm having some troubles and would be happy if you could help hi dalen what kind of troubles do you have ? well, the appserver is running and I've been able to do gnue-readgcd on the address_person example without errors but when I do: gnue-forms appserver://appserver/form/address_person it gives an error: gnue 'module' object has no attribute 'DataObject' in a forms dialog this is using the packages from debian unstable dalen, are you able to run sample.gfd (within appserver's examples dir) ? (that one uses address_person too) nope same error ? yup when I try the curses interface it says: Cannot load plugin 'appserver' hm anybody around who's using the packages too ? The complete error is: The following plugins failed: * gnue.common.datasources.drivers.appserver.appserver: AttributeError: 'module' object has no attribute 'DataObject' hm, i don't know anything about the packages from debian unstable, but that one looks quite old since gnue-common underwent some heavy changes wrt datasources hmmm ajmitch ? dalen: did you have an older version of gnue installed and did upgrade lately? as i see things it should be "gnue.common.datasources.drivers.other.appserver" yes ok please uninstall gnue completely (saving your config files before) then see what is left in /usr/lib/gnue I fear uninstall doesn't remove the pyc/pyo files *** dalen_ has joined #gnuenterprise seems like something weird happened to the irc client reinhard: ok reinhard: please uninstall gnue completely reinhard: (saving your config files before) reinhard: then see what is left in /usr/lib/gnue reinhard: I fear uninstall doesn't remove the pyc/pyo files reinhard yes, I've had older versions installed there was loads left after the uninstall :) all .pyo files? you did uninstall all gnue, right? gnue-common, gnue-forms, gnue-appserver... mostly .pyc files I've now uninstalled them all and deleted the /usr/lib/gnue directory ok however I've just seen that latest gnue-forms hasn't made it into sid yet :( so the versions of gnue-forms and gnue-common in sid don't match damn gnue-forms 0.5.12-1 is the one I'm using. not the gnue-forms-wxgtk 0.5.11-2 erm yes scratch that, it's correct after new install it should work and we have to tell ajmitch that uninstall doesn't remove everything correctly yep, now it worked. thanks for the help I'll now proceed with reading the rest of the documentation :) I'm trying to make a replacement for a Filemaker Pro based system for subscribers of a paper have fun and please come back here whenever you have questions or ideas for improvement just out of curiosity where is .nu ? has there been any work done on the financials pakage? would be nice to tie it directly to the general ledger well, I'm in .se dalen_: I've been working on the financials package, among other things .nu is some carribean island I think Niube or something but it goes slowly :( dalen_: ah :) At the moment I'm using a commercial windows only system for the accounting SPCS Aministration (I think they just changed name to something else) it would be nice to replace it with something free. but gnu cash just doesn't cut it for a business yep right quite exactly my view it's more a lack of resources now... but I've had somebody here yesterday that was pretty interested in helping out so maybe things speed up a bit When I've learn't a bit more about how it all works I might be able to help. I'm both a python programmer and work with accounting and administration excellent I'm looking forward to any kind of input from you where in .se are you? Stockholm I've been in Gothenborg once nice country yeah, it's quite nice. you're in .de? .at ahh, ok. food time yes, good idea *** johannesV_ has joined #gnuenterprise *** johannesV has quit IRC * ajmitch saw the uninstall issue, file it as a bug so I can keep track of it ok filed against gnue-common. but it probably applies to the other gnue packages as well thanks * ajmitch can probably clean up .py[co] files in the postrm this should theoretically apply to all python based packages I wonder if there is something in this python magic stuff you use probably I'll check with jbailey for suitable cdbs magic dalen_: filed in debian, or ubuntu? debian ok * ajmitch will wait for the bts pages to show it :) yeah, I've seen it's not up. but I have no return mail either (yet at least) ajmitch: if we did a new release this week any chance to get it into breezy? yes breezy is in 3 weeks so I can still push in new packages if I talk fast ;) this is just a new upstream version which we froze for in july, but is far less strict for universe reinhard: what components? common & appserver? ajmitch: now it's there thanks common, forms, appserver several bug fixes done ok * ajmitch can get that in excellent, thanks it'll be a good test of my new bzr playground for packaging ;) * johannesV_ needs some coffee *** btami has quit IRC do you have to use a OnValidate trigger to chack that a property is unique? *** sjc has quit IRC *** sjc has joined #gnuenterprise *** iro has joined #gnuenterprise *** sjc has quit IRC dalen_: you could do a find() with that property as condition to see if there is already another record with the property set to the same value also IIRC there is a way to have a unique index over a property, meaning the uniqueness would be enforced by the db backend but I can't remember the syntax for that off-hand johannesV_: do I remember correctly? *** kilo has quit IRC yeah, it would probably be faster to have the backend check it. and easier. btw, there's some errors in the application developers guide on page 9 (in the pdf) it should say "less than 10000" instead of "greater than 10000" and on page 12 the property "czc" shouldn't have a closing slash, right? dalen_: yes, both is correct changing it right now, thanks a log lot you make me happy - finally somebody is really *reading* my docs ;-) reinhard: now you have one more thing to fix they are good otherwise. some sections need a bit of expanding and everything isn't covered. but I suppose that's already on your todo list are the packages documentation still up to date, on: doh, I meant pasting http://www.gnuenterprise.org/packages/docs.php hate that windows doesn't copy text on selection.... reinhard, you can add an -tag to a class (in a gcd) to enforce uniqueness on a given set of properties like of course there could be multiple definitions, as well as multiple -tags per index *** jamest has joined #gnuenterprise jamest, could you please have a look why commit-emails are not working ? sure, in a bit ok, thanks ... don't need to hurry :) dalen_: can't say much about packages except that they are either in alpha or in planning state (AFAICT) dalen_: regarding documentation, in case you have any proposals, patches are *highly* welcome jamest: I have planned an attack on you today, please let me know when you have a minute reinhard, I'll try. a section on class extension would be pretty good describing those indexes would also be nice great just send me what you have reinhard at gnue dot org erm not sure if that works with new server to make sure take reinhard at bytewise dot at *** dalen has quit IRC ping pong ah reinhard, you've seen my private window ? no * johannesV_ thought he's a ghost again ... and now ? no hm try /msg reinhard blahblah *** johannesV_ has quit IRC *** johannesV has joined #gnuenterprise *** iro has quit IRC *** johannesV_ has joined #gnuenterprise *** johannesV_ has left #gnuenterprise reinhard: fire away, i'll be at my desk for a while o k first point I looked at the undelete feature and had following idea to make things clearer to the user * introduce a new message requestUNDELETE * make a new menu item "undelete record" directly under "delete record" where only one of both can be active at a time * show the "delete" button pressed in if cursor is on a deleted record * clicking the delete button then again issues a requestUNDELETE what would you think about this change? i think that would be great ok committed :-) second under what condition should it be possible to insert a new record? so far it was possible always except the current record was the last record and it was empty i changed that to always except the current record is empty (no matter if it is the last one or not) but I am not sure if that is logical as I still can insert tons of empty records in a table *** jcater has joined #gnuenterprise bbl *** btami has joined #gnuenterprise *** dneighbors has joined #gnuenterprise reinhard: i don't follow also, you can force behaviour with a config var to only create new records per user request sorry for being unclear if I hit "new record" 10 times, I get 10 empty records is that what we want? is that a new behaviour? I don't think so but as I said I'm not entirely sure what is the desired behavior I'm used to only being allowed one empty record but that's out of habit not sure if I have a good reason for it the existing behavior was allow only one empty record if you are at last record allow as many empty records as you want if you are not at last record interesting where I don't see any reasoning behind that would have been a bug i would imagine as the idea was don't allow a new record if the last was empty the not last record allow as many as you want was a bug i would think it would only allow 1 tops and if you have an empty anywhere and a new record is created then "old" empty new record would be removed and a new empty opened at the current location well, that removing of the old empty record isn't that easy... would you remove it as soon as a new empty record is created or as soon as the user navigates to another record? i think the latter would be easier (latter might make sense, too) and achive the same result so as soon as you move out of an empty record, ditch it? with minor user confusing about "where'd my record go?" if they accidentally moved out of record might be reasonable, yes at one time, I'm pretty sure that's what forms did on the last record if you down arrowed into a new record it created an empty one ok, I'll try to achieve that if you then went back up one record, it'd delete that empty record but I don't recall when that changed jcater: yes, I also remember it being that way or if I'm remembering some imaginary world (both are possible) i don't recall that behaviour i thought it just ignored the last empty on commit anyway as I have both of you on the line another question I tried hard to find out the effect of the "edit record" button in the toolbar and failed can't even find where requestENTEREDIT is handeled handled i think that was a jcater addition where all fields were read only unless the user selected that button so that users could not accidentally change a record does that depend on a parameter? I don't remember it was one of derek's "must have" features it might have been a parameter to enable I don't recall it's been too long :( I couldn't find *anything* i had nothing to do with the last record thing dneighbors: we're talking about the "edit record" button in the toolbar jcater: i swore this was for a data entry app in house ok Changed behavior so navigating out of an empty record discards it back to requestENTEREDIT I herewith claim that this event is processed nowhere and neither the button nor the menu entry nor the F2 key have *any* effect and should all be removed (or at least commented out) to not confuse users okay anybody see a chance to confirm or disprove this? $ find . -name "*.py" | xargs grep -i enteredit ./uidrivers/_commonGuiToolkit/ToolBar.py: (u_('Edit Current Record'),'ENTEREDIT'), ./uidrivers/_base/UserActions.py: event="ENTEREDIT", ./input/GFKeyMapper.py: (vk.F2, False, False, False) : 'ENTEREDIT', ... is what I did to come to this conclusion yep btw. jcater,may i ask you to restore some old fuctionality on the server wiki,logs, search, cvs-view ok, commenting out button, menu entry and hotkey now... if you have time, of course reinhard: iirc you said clicking on scrollbar don't have to move the focus from another entry after johannesV changes, it moves not a big issue, just noticed... ok, thanks which ui? just tried win32 ok maybe it's time to suggest again to set up DCL too :) maybe or a different issue tracker btami: if we are at talking about "if you have time"... ;-) any chance you can look at enabling dropdowns to be entered with the keyboard? I suggested to johannesV to try to change CBS_DROPDOWNLIST to CBS_DROPDOWN, but he tested and it didn't help *** dvoss has joined #gnuenterprise Yesterday, I tried to download the Windows release of GNUe from http://www.gnuenterprise.org/downloads/current.php. You guys were having some sort of server issue, so I couldn't. I still can't; is the server still down? The wiki is down too. (or at least not accessible) In both cases, I get a 403 Forbidden error. reinhard: yep, i'v spent 1 hour with CBS_DROPDOWN without any success :( darn btami: :( the pb is LSB_NOTIFY dvoss, I haven't moved the wiki over dvoss: server is being moved to a new machine apache should be running now though, so you should be able to download the setup btami: btw, scrollbar is moving cursor on gtk2, too reinhard: if you remove it, the CBS_DROPDOWN is ok, but you can't grab WM_CHAR and other events from edit part of dropdow so it's a bit complicated wil fight with it tomorrow again will sigh jcater, it's not working, though. btami: thanks a lot btami: just let me know if you think if it is worth waiting with the release if you think you can fix it within 2 days or so I'd wait please "Forbidden -- You don't have permission to access /downloads/current/gnue-runtime-environment-2005.07.14.exe on this server. Additionally, a 403 Forbidden error was encountered while trying to use an ErrorDocument to handle the request. " I'd *happily* wait, as my customer wants it... :-) ok :) will look into scrollbar meanwhile ah hmm btw.another bug with scrollbars on win32 introduced with latest commit dragging the thumb to the bottom doesn't moves to the last record maybe it was there before, can't remember... dvoss, http://www.gnuenterprise.org/downloads/releases/gnue-runtime-environment-2005.07.14.exe should work seems i'm having issues with symlinks in the current/ folder Yep, it works. Thanks! btami, this is because the trackposition isn't reported with the last recordnumber, but stops at count - page btw. i've just found out that LBS_NOTIFY seems to introduce the pb wrt combos (as you've mentioned before) :) reinhard, all translations are already up to date for "de" johannesV: thanks I am looking at the scrollbar issues now btami: cursor moving on scrollbar click should be fixed now after playing with scrollbar a bit I'm not sure if its behavior is what people would expect intuitively or if people would rather expect to move the visible portion within the grid rather than the current record I mean clicking on the scrollbar I would normally expect something to *scroll* not just moving the cursor into the next line what do you think? *** johannesV has quit IRC *** sjc has joined #gnuenterprise It looks like the GNUe Windows package includes Python 2.3 dll's. Will GNUe use Python 2.4 if I install it? probably not k *** michael301080 has joined #gnuenterprise reinhard: could be better, imho wrt "scroll" the wiki is back albeit possibly not editable nevermind, editable too jcater: Can you make it possible to access the irc-logs? please ? :-) I'm trying but apache2 is winning I specify Option Indexes, be she no listen OK thanks *** dvoss has quit IRC reinhard: agreed (scroll) ok, I'll look into it thx i have no svn here too :( yet no, i have it again wohooo jcater: the wiki is forbidden and logs too working on it thanks jamest. ;-) try now on the logs woohoo, works like a charm, here in Belgium! fwiw svn is browsable at http://svn.gnuenterprise.org/gnue/ and the viewcvs links should be working as well but that http://svn.gnuenterprise.org/gnue/ is for anonymous checkout svn co http://svn.gnuenterprise.org/gnue/trunk/ gnue, iow OK, thanks jcater :-) and to whomever brought down our old server I hope you have a non-peaceful, eternal rot in the depths of hell not that I get bitter about such things /msg jcater maybe they'll accidentally read some of the reports code, it'd have the same effect hmm maybe that's what put them over the edge in the first place *** dneighbors has quit IRC *** kilo has joined #gnuenterprise *** michael301080 has left #gnuenterprise reinhard: something weird happning with delete when i delete a rec in states.gfd then press a down key the deleted record disappear, and it really deletes, without pressing a save icon yep that's because the block is set to autocommit :) grrrrrrrr :) I have to ":)" because I stumbled over exactly the same thing about an hour ago and it cost me 15 minutes or so to figure out :) ok :) should have waited 15 mins then before telling the clue :) 'night *** btami has quit IRC *** yure has quit IRC kilo: you still awake? bored? want to try something cool? just committed scrollbar change i connection with gnue? lol, yes you could test scrollbar on win32 canot gnue on this machine... ah ok nevermind win32 would mean to start up another machine, but that is like a B737 np, we can test tomorrow ok sorry I will drink a southern comfort on your health now :) good night all *** reinhard has quit IRC *** jamest has quit IRC *** jamest has joined #gnuenterprise *** jamest has left #gnuenterprise *** jcater has quit IRC *** dalen_ has quit IRC *** kilo has quit IRC *** sjc has quit IRC *** jcater has joined #gnuenterprise *** jcater has quit IRC hi reinhard last commit for toolbar does not work for pyGTK < 2.4.0 File "/home/ds/svn/gnue/.cvsdevelbase/gnue/common/events/EventController.py", line 107, in dispatchEvent