at this rate, you'll be designing under 0.5.0 Action: jcater ducks scary Action: ajmitch has nothing to commit today sigh in that case both of you are fired :) Action: ajmitch packs his bags nite jcater (~jcater@cpe-066-061-083-220.midsouth.rr.com) left irc: "til tomorrow" ajmitch: did jcater say what our severance package would be like? probably a stale donut psu (psu@manorcon.demon.co.uk) joined #gnuenterprise. hello psu lists have been busy lately :) yep how's the hunt for willing vic^H^Holunteers? slim pickings at the moment people might be a bit more forthcoming when they realise I mean it ;-) yeah To be fair, I don;t want to drag people off coding as that's really the wrong way around but no-one else is probably interested enuf to care you've been doing a great service oh people are interested but too scared ;) i.e. the ones who value KC enough to help are already over-committed anyway yep jcater (~jcater@cpe-066-061-083-220.midsouth.rr.com) joined #gnuenterprise. wb jcater thx btami (~tamas@ip102-205.ktv.tiszanet.hu) joined #gnuenterprise. jcater (~jcater@cpe-066-061-083-220.midsouth.rr.com) left irc: reinhard (~reinhard@M693P020.adsl.highway.telekom.at) joined #gnuenterprise. psu (psu@manorcon.demon.co.uk) left #gnuenterprise. Vee2d2 (~vin@66-169-136-33.ftwrth.tx.charter.com) left irc: "Got a group logo? http://www.topsew.com/" Vee2d2 (~vin@66-169-136-33.ftwrth.tx.charter.com) joined #gnuenterprise. fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) left irc: Read error: 110 (Connection timed out) SachaS (~Sacha@dialup-196-251.wasp.net.au) left irc: Read error: 113 (No route to host) reinhard (~reinhard@M693P020.adsl.highway.telekom.at) left irc: Read error: 104 (Connection reset by peer) reinhard (~reinhard@M693P020.adsl.highway.telekom.at) joined #gnuenterprise. ariel_ (~ariel@ppp-217-133-241-185.dialup.tiscali.it) joined #gnuenterprise. fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) joined #gnuenterprise. fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) left irc: Read error: 110 (Connection timed out) dtm ([zZ+Q2kzEP@ip66.promontory.sfo.interquest.net) got netsplit. dtm ([zZ+Q2kzEP@ip66.promontory.sfo.interquest.net) returned to #gnuenterprise. dtm ([zZ+Q2kzEP@ip66.promontory.sfo.interquest.net) got netsplit. dtm ([zZ+Q2kzEP@ip66.promontory.sfo.interquest.net) returned to #gnuenterprise. jbailey (~jbailey@CPE014260028338.cpe.net.cable.rogers.com) joined #gnuenterprise. SachaS (~Sacha@dialup-196-95.wasp.net.au) joined #gnuenterprise. fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) joined #gnuenterprise. reinhard: hi, are you there? yes i just printed out the whitepaper, i want to check it better before (eventually) comment it. can i go on with the analisys of class repository? i begun to put down some python classes to play around yes please keep me informed :) sure :) Action: SachaS is following the appserver story with interest but is busy himself Action: SachaS wants to have an ERP for himself :) siesel (jan@131.220.92.82) joined #gnuenterprise. hi jbailey (~jbailey@CPE014260028338.cpe.net.cable.rogers.com) left irc: "Client Exiting" jamest (~jamest@leibniz.math.ksu.edu) joined #gnuenterprise. dsmith (firewall-u@cherry7.comerica.com) joined #gnuenterprise. reinhard: how the language interface stores NEW objects? do we need another api function say getNextAvailableID hmmm not sure or just pass a standard ID say 0 to store maybe we can define that storing an object with ID="" means to store a new object or like that yes exactly what you say :) so appserver knows that this is an insert (and not an update) and provides by itself to generate new id yes at least that would be my evil plan ;) ok remember to signal this on an explaination of the api. yes thanks for pointing out this issue fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) left irc: Read error: 110 (Connection timed out) when i get the code from cvs and go into common, forms, designer and do an ./setup.py install everything gets installed into /usr/local/gnue right? no other directoreis... yes but if you want to just test w/o any install there is a setup-cvs.py in the topmost directory run that and it'll config to run from cvs and install a few commands in $HOME/bin if you plan on messing with cvs head then I suggest you do that as head is very unstable at the moment ok. if i want to get rid of "production" gnue (install from common dir, forms dir etc) I just 'rm -rf /usr/local/gnue' . is that enough? that's how I do it cool. how do i get the last stable from cvs? i dont want to mess around ;) btami (~tamas@ip102-205.ktv.tiszanet.hu) left irc: checkout cvs as normal ok. then cd to common and cvs update -r common-0-4-1-branch then cd to forms and cvs update -r forms-0-4-1-branch then cd to designer cvs update -r designer-0-4-1-branch that should get you what will soon become the 0.4.2 release ok and then after the cvs update -r xxx-0-4-1-branch back into each dir and './setup.py install' right? for a production install yes maybe you could say how to: warning: install: modules installed to '/usr/local/gnue/lib/python/', which is not in Python's module search path (sys.path) -- you'll have to change the search path yourself setup the path in python. as i never use python... is it the general PATH env variable? please read that again it should say something to the effect ignore the above error ah true sorry it's confusing I know yep gnue-forms forms/samples/helloworld2.gfd comes up good fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) joined #gnuenterprise. Chipaca (~john@ADSL-200-59-86-93.cordoba.sinectis.com.ar) joined #gnuenterprise. hi all how was the attribute of entry that made multi-row entries skip N rows on the screen? rowSpacer tks appserver to db connection is xml-rpc? appserver to db connection is xml-rpc based? is uses the same drivers that are in gnue common to connect to the db err... gnue.common.GParser.MarkupError: Error processing tag [required attribute "block" not present] something changed? Chipaca: what are you running? cvs head or a branch? Action: Chipaca remembers an email Action: Chipaca goes to the branch sorry :) head is wacked for the old forms but gnue/forms/util contains a convert script ah we're seperating the logic from the layout and making a real mess in the process expect head forms/designer to be broken for a while I'm writing a small gnue thingie to keep track of the LUGs CD requests, so I'm doing it against gnue's cvs, and I kind of miss texentries :) err textlookups, or whatever they're called broken = it works but who knows how well :) heh what was the name of the branch? checkout cvs as normal ok. then cd to common and cvs update -r common-0-4-1-branch then cd to forms and cvs update -r forms-0-4-1-branch then cd to designer cvs update -r designer-0-4-1-branch that should get you what will soon become the 0.4.2 release i'll be merging more patches in hopefully before Dec ack had a client bump a deadline up from the 1st week of dec to this tuesday they promised a client of their this w/o talking to me at all for i in common forms deskigner; do cvs update -Pd -r ${i}-0-4-1-branch $i/; done designer lol getting fancy :) Action: Chipaca hates typing all these script whizzes :) Chipaca: if you can stomach issues then I might suggest you run head with a little patching I have it running most my stuff here jamest: ok, will do -- tonight the only warning is designer is wierd Action: SachaS is thinking of a personal need for gnue ..... I've never used designer Chipaca: then you'll be golden it was too green when we started, and I haven't checked lately but I suspect it'll still piss me off the way most designers do Chipaca: You should, it makes things quite comfortable :) good night Nick change: SachaS -> SachaS_sleep SachaS_sleep: you make me feel tired ;) hey reinhard its 23.00 here. good time to go to bed. see you soon. Action: Chipaca waves Chipaca (~john@ADSL-200-59-86-93.cordoba.sinectis.com.ar) left irc: "bbl" jcater (~jason@w202.z065105010.mem-tn.dsl.cnc.net) joined #gnuenterprise. fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) left irc: Read error: 110 (Connection timed out) drochaid_ (~drochaid@pc2-kirk1-3-cust16.renf.cable.ntl.com) joined #gnuenterprise. drochaid (~drochaid@pc2-kirk1-3-cust16.renf.cable.ntl.com) left irc: Read error: 104 (Connection reset by peer) jcater (~jason@w202.z065105010.mem-tn.dsl.cnc.net) left irc: "rebooting" jcater (~jason@w202.z065105010.mem-tn.dsl.cnc.net) joined #gnuenterprise. Nick change: drochaid_ -> drochaid reinhard: are you still around? ariel_: in and out ariel_: just giving dinner to the kids sorry, can i disturb? in the store function of the api we don't pass the TYPE of the object to be stored if this is an update the type can be reconstruted from the id but if inserting a new object... a solution is that the property list must contain "type" revDeke (www-data@alt1.libertydistribution.com) joined #gnuenterprise. and the data list must contain the appropriate value ariel_: even a new object has to be defined in the object repository siesel if the object is a object definition? ariel_: so the type would be kind of "pre-defined" siesel: suppose you start with the empty repository then it uses the the buildin values i really would like to interject you want just to add empty module 'foo' by the api we defined you (the client) open a session then call store(sessionhandle, ....) with reinhard we agreed that if the id is an empty string this means adding a new object hmmm Action: reinhard thinks maybe we have to redefine that johannes__ (~johannes@M693P018.adsl.highway.telekom.at) left irc: Read error: 54 (Connection reset by peer) IMHO : store(xxx, ['ID','objtype','modulename','objname'] ['','class','foo','fooobj']) reinhard: did we pass the class type when calling store? ../we pass/ we defined it to pass/ siesel: not so defined in hallowen it was store(session,id-list,prop-list,data-list) passing type in prop-list and value in data is also what i think johannes__ (~johannes@M693P018.adsl.highway.telekom.at) joined #gnuenterprise. ok, that means just passing an empty ID don't work, because appserver don't know in which class it should store something gotcha! the original specification we agreed upon for appserver never once mentioned objects or object to relational i guess im a bit miffed that this is now the focus when we never agreed upon it and pretty matter factedly stated that it should wait fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) joined #gnuenterprise. hooha but we could pass an ID like "classname@" for new objects and "classname@32234" for already created objects maybe you could not focus on objects at all? and follow the agreed upon requirements? Action: siesel has to run siesel (jan@131.220.92.82) left irc: "later" Action: revDeke guesses that speaks volumes revDeke: please read the halloween mail and the new appserver whitepaper ariel_: PLEASE read the requirements written in Feb 2002 Action: revDeke is sorry if the halloween emails were not able to understand the requirements we specifically asked that the meetings be done on irc (while in person) to avoid confusion we had a standing requirements definition which NO WHERE included object2sql sql2object yet this is the primary focus so much so that now in diagrams appserver is called OBJECT SERVER fixe (sfad@dsl093-081-253.chi2.dsl.speakeasy.net) left irc: "pamf" reinhard: i'm sorry jan has left, IMHO i prefer the first solution because ids don't have to contain meaningful information (i.e classname@ as prefix) revDeke: can you point me exactly to the doc you are speaking about? yeah let me find it reinhard_ (~reinhard@M693P020.adsl.highway.telekom.at) joined #gnuenterprise. http://www.gnue.org/~dneighbo/gnue/geas_notes.txt specifically items 1 - 12 reinhard (~reinhard@M693P020.adsl.highway.telekom.at) left irc: Read error: 110 (Connection timed out) revDeke: ok let me see it, but if this is neither on cvs or website i suppose it's something that derek is working on it and not definitive ok there there is a task for reinhard_ to prepare a proposal, this is exactly what he did with the withepaper and the emails there were answers on the emails by variuos people raising many issues, mainly on what will be in common and what in appserver, but the object-relational mapping issue was not raised Action: jcater is not gonna get into the middle of it but derek is right the discussions we all had were to not mess w/objects at least initially or you end up having GEASv1 all over again bbl Action: jcater is away: lunch anyway the main goal is not to do construct an oo-db, rather giving the user the possibility to manipulate what we call business objects roughly speaking tables that can have calculated fields or performing procedures on their data reinhard_: did you miss what i say after jan has left ariel_: i just have read the logs i was disconnected Nick change: reinhard_ -> reinhard i was saying maybe we have to redefine that like we said that an object of type "foo" returns an id of "foo.121203798" we could say that a new object of type "foo" has an id of "foo.0" or only "foo" or even "foo.new" so actually i had the same proposal as siesel btw for your information revDeke == derek :-) reinhard: IMHO i prefer the first solution because ids don't have to contain meaningful information (i.e classname@ as prefix) what is the first solution (i think i missed that) ? passing type in prop-list and value in data ariel_: sorry i had to run and for the record that list was discussed publicly and i believe posted to the list you mean that every class has an implicit property "type" and probably on teh website i gave url to my home dir because thats what i remembered ok there there is a task for reinhard_ to prepare a proposal, this is exactly what he did with the withepaper and the emails revDeke: this list is part of the whitepaper you are correct here but im stating that the whitepaper is not within the requirements and you all seem to be ignoring that if we end up with a repeat of geas v1 dont be suprised revDeke: can you explain what appserver will do in your opinion other than be a db abstraction library ? it will be remote db and triggers over rpc nothing more nothing less basically triggers and database are in common and get called locally will it call triggers automatically all that appserver should really need i.e. will it define events and fire automatic procedures? is to provide a mechanism to make those same calls via RPC no objects involved if you are familiar with borland and its tools think mida er think midas and remote datasources revDeke: sorry i'm not familiar with midas ok think gnue forms and think of making datasources instead of inline to a form being 'remote' or accessed via rpc think same thing for triggers and you have the basis of appserver initial version where are the triggers defined? revDeke: kinda ODBC trough RPC? ariel_: its more than that because code can be accessed via rpc as well think remote trigger repository and disconnected data source some tricks you can play with it are things like 'brief case mode' but thats a whole other discussion in many ways borlands BDE is similar to what we have done with common for data access the draw back is having to have bde on client installations midas made it so that you need not have bde on clients but rahter remotely accessed bde and in conjunction opend up the idea of server side logic in a middle tier that was compatiable with 2 tier modes ive not done borladn in a while i guess midas is now called datasnap http://www.borland.com/delphi/dsnap/index.html thank you it was illuminating, still i think i need more clarification, i will try to study it a little Action: ariel_ has to go ariel_ (~ariel@ppp-217-133-241-185.dialup.tiscali.it) left irc: "Uscita dal client" the way the implemented was they had 3 components 1. data broker (remote datasources) 2. constraint broker (remote triggers) 3. object broker (what you all are driving for) my suggestion was and still is lets start with 1 & 2 which are HIGHLY useful but should require little work as common already is half way there to providing them already 3 would require some work but shoudnt be impossible to add-on in future versions here is an old faq sheet http://info.borland.com/midas/papers/overview/midas.html btw: im not saying we should do midas im simply using it as a reference point to try to better convey the general concept siesel (jan@xdsl-195-14-204-21.netcologne.de) joined #gnuenterprise. Action: jcater is back (gone 00:57:35) jbailey (~jbailey@atlas.fundserv.com) joined #gnuenterprise. revDeke: can you give me that scripter command line again? Debian Sid still has CVS from November the 5th in it. Should that be updated? Also, does anyone know a good tool for moving stuff into mysql from a dbf file? =) dsmith (firewall-u@cherry7.comerica.com) left irc: "later.." revDeke: found it, google rules jbailey: well, certainly don't package up cvs head Okee. =) Do you folks actually do CVS branches and all that? yeah forms-0-4-1-branch common-0-4-1-branch designer-0-4-1-branch Cool. are the stable branch It also looks like I'm supposed to defer byte-compiling until install time. why? i swore the pyc files were cross platform Dunno - but the new dh_python file does it in the postinst. It creates both pyc and pyo files. Are you guys expecting to continue to patch the 0.4.1 branch? morning hi ajmitch, and a good morning :) jbailey: i have a dbf2mysql around here somewhere - there's one floating around on the net, the one i have is hacked to extract from visual foxpro :) hello siesel :) dsmith (firewall-u@cherry7.comerica.com) joined #gnuenterprise. jbailey: you mean add stuff? ajmitch: Cool, thanks. i'd like to add a few more things jamest: Or if there's actualy bug fixing happening on the branch and stuff. no bugs that I know of where know of == that I intend to fix in 0.4.2 ;) you wanting to package up some new .debs? I was thinking of it. hmmmm I'm doing more hacking on this database for my work. i've heard very little about 0.4.1 hmm, is that a good thing? but i guess we could wrap up a 0.4.2 release jcater? why roll 0.4.2 when there's no changes? there are changes, just not alot of them ? oooo, that's not true something was broken with 0.4.1 installs but I don't recall what it was? oh curses didn't install someone complained about that that was it and i know I fixed a few bugs in there i think jcater did as well merged in some papo stuff too jcater: you do the curses improvements in the branch? hmm yes menu works in curses mostly curses mostly works now in 0.4.2 I would like to finish it up the biggest push for 0.4.2 would probaby be to have the .debs in sid be valid but I honestly don't see me having time to do that anytime soon jcater: what's broken? just really odd things buttons don't fire menus don't exit once they've fired an event right now the curses thing is good for demoing but not for using ah just ood things dropdowns just look like dropdowns but they act like text fields ;) are these issues in cursing or issues in uidriver or unknown? Action: jbailey is away: I'm busy Action: ajmitch runs off to work dsmith (firewall-u@cherry7.comerica.com) left irc: Remote closed the connection tcom (~icechat@58.mufdf.lsan.la6ca01r1.dsl.att.net) joined #gnuenterprise. tcom (~icechat@58.mufdf.lsan.la6ca01r1.dsl.att.net) left irc: Client Quit dsmith (firewall-u@cherry7.comerica.com) joined #gnuenterprise. dsmith (firewall-u@cherry7.comerica.com) left irc: "later.." sjc_ (~sjc@modem-3204.elephant.dialup.pol.co.uk) joined #gnuenterprise. sjc_ (~sjc@modem-3204.elephant.dialup.pol.co.uk) left #gnuenterprise ("Client Exiting"). jamest (~jamest@leibniz.math.ksu.edu) left #gnuenterprise. dsmith (firewall-u@cherry7.comerica.com) joined #gnuenterprise. jan (jan@xdsl-195-14-205-192.netcologne.de) joined #gnuenterprise. siesel (jan@xdsl-195-14-204-21.netcologne.de) left irc: Read error: 104 (Connection reset by peer) jan (jan@xdsl-195-14-205-192.netcologne.de) left irc: Client Quit jan (jan@xdsl-195-14-205-192.netcologne.de) joined #gnuenterprise. jan (jan@xdsl-195-14-205-192.netcologne.de) left irc: Client Quit jan (jan@xdsl-195-14-205-192.netcologne.de) joined #gnuenterprise. jan (jan@xdsl-195-14-205-192.netcologne.de) left irc: Client Quit siesel (jan@xdsl-195-14-205-192.netcologne.de) joined #gnuenterprise. siesel (jan@xdsl-195-14-205-192.netcologne.de) left irc: Client Quit Just go ahead and write your own multitasking multiuser os! Worked for me all the times. -- Linus Torvalds siesel (jan@xdsl-195-14-205-192.netcologne.de) joined #gnuenterprise. siesel: you're having fun ? ;) kind of, I've switched to use xchat ;) ah didn't hear much from you wrt halloween mails and whitepaper update I'm quite busy whith some real world stuff :( jamest (~jamest@adsl-64-216-107-182.dsl.tpkaks.swbell.net) joined #gnuenterprise. can i interpret this as a silent approval? :) you're not following at all at the moment? or just skim over everything and found nothing worth disagreeing? Its quite time consuming to follow. Especially if it comes to security ;) yes i share this experience :) and I realized again how bad my english is :( speaking of the appserver discussions, I haven't found anything to disagree with :) But I haven't looked at the revised version of the appserver whitepaper yet. Tonight I've to decide if I begin to code parts of the new appserver api, or have a look at the whitepaper ... what about Johannes and ariel, has the class repository improved meanwhile? please dont start coding until there is resolution you all keep saying you want feed back and 3 of the core gnue team have told you they disagree with the whitepaper yet you are willing to start coding this is most disturbing where is this whitepaper? if the only people that need to agree on the paper are those that wrote the paper why bother asking for feedback? revDeke: sorry, I haven't read the whitepaper yet, so I can't comment on that. I just wanted to change the appserver RPC api from its current (very slow) to a better implementation. so you are going to code appserver w/o reading the whitepaper? Tonight I've to decide if I begin to code parts of the new appserver api :) i see so its not an appserver thing its an RPC thing you are working on Action: revDeke can understand that difference the appserver RPC api? or the gnue rpc code? i.e., stuff in common? Possibly both. But I hope just the first one ;) i guess only the first one otay jcater: and derek: to keep you informed after we have remote data access about 100% working in appserver 0.0.1 we found out that the RPC interface definition for that part is not optimal performance wise and in Frankfurt we agreed on a better interface which is what siesel is talking about (siesel: please tell me if i'm wrong with this) http://mail.gnu.org/pipermail/gnue-dev/2002-November/000307.html no. siesel: "no you're wrong" or "no, you're not wrong"? ehhmm hmmmm the second ;) ok good :) siesel: i think it might be a good idea to code the new interface routines into new .py source files so for the time being both interfaces are possible as i think we don't want to implement the new "dbdriver" for appserver before 0.5 release of common to not break common (and forms and reports) before 0.5 release btw. jcater: I've some code for reading DBF files, which could be a basis for a DBF (DBase, Foxpro) dbadapter. Where should it go: in common/contrib or in common/src/dbdriver/dbase? reinhard: yes, I'll will do it that way. it's the basis for a .dbf file driver I would think in a dbase dir excellent i think i can sleep much better now ;-) does dbase do sql? night all no. I don't think so. reinhard (~reinhard@M693P020.adsl.highway.telekom.at) left irc: "Don't contradict a woman -- wait until she does herself" night reinhard just curious but I just used dbase III :) I think the DBF driver would only be useful for integrator. Especially because this heap of DBF code just can read DBF jcater: if 32 <= keystroke <= 255 and chr(keystroke) in string.printable: does not pass non-ascii it is done differently in wx driver first row is from curses driver Could not load MySQLdb. For MySQL support, please install mysql-python 0.9.0 or later fromhttp://sourceforge.net/projects/mysql-python This probably shouldn't be a fatal error. what you doing to get that? hey guys I have an idea for a splash screen for designer he needs help and would like to get y'alls feedback http://www.gnuenterprise.org/~jcater/goats/designer-splash.png roflmao lemme know what you think? looks great. jamest: using a mysql provider without it installed. =) Action: siesel is just not shure what the goat is doing with the computer siesel: use your imagination what do we joke about so much? think "mad goat rape" chilly willy? jbailey: i guess it wouldn't be too hard to emulate a mysql db if the driver is missing cat file | cut -d\: -f dbfile :) j/k i probably butchered the cut args anyway :) jamest: Emulating it is probably not ideal. Scaring people into doing the right thing is better. Have a picture of rms appear in that widget I was talking to you about the other day. :) i was running one form and at the same time i was unable to alter table from psql, could anyone comment? DB000: Traceback (most recent call last): DB000: File "/usr/lib/gnue/lib/python/gnue/designer/TemplateParser.py", line 184, in OnNextStep DB000: self.setStep(self.nextStep) DB000: File "/usr/lib/gnue/lib/python/gnue/designer/TemplateParser.py", line 151, in setStep DB000: self.stepInfo = self.parser.template.GetStep(step) DB000: File "/usr/lib/gnue/lib/python/gnue/designer/templates/forms/FormBuilder.py", line 86, in GetStep DB000: self.variables['newPage%s' % iteration] = self.variables['newPage'] DB000: KeyError: newPage Aside from the refusal to continue, is this bad? =) So.. Since the wizards consistantly don't love me. It *is* possible to setup these forms without it right? =) ra3vat_: postgres is smart enough to not allow table struct updates if someone is using the table (iirc) i.e. if your form had a connection to that table open you couldnt delete a column on that table jbailey: what are you using? cvs head? cvs branch Cvs head from November 9th. 0.4.1 release? the sid packages. dogfood kinda stuff. hmmm 9th might have been ok would have to see change logs That was the concensus at the time. revDeke: i was trying to add same thing if you were trying to 'alter' the table i think it will reject it if someone has a cursor open to it (iirc) The botched version number thing keeps me from updating to the final release. did it give you an error? I could CVS sync to the release branches, I guess. jbailey: i would suggest that i think as i know they are pretty solid Hmmm.. Okay. ok, not a problem We are going to create the docs needed for the source distro. You'll need lyx installed. Press Control-D to abort or any other key to continue Space doesn't work... Nor in fact does anything other than the enter key. Action: jbailey is back (gone 03:37:26) jbailey: sync to the release branches you'll be glad you did In process. ok Blech. I don't know what to do with the version number for common forms and designer I see were bumped to 0.4.1.99 common should be too that stands for cvs i had a prob installing common 0.4.1 release today on my redhat 7.3 yeah larsu: what was the issue? to the 0.4.1.99 not the 'problems' [rfox@linux GNUe-Common-0.4.1]$ python2.2 setup.py install running install Traceback (most recent call last): ... File "/usr/src/build/87651-i386/install/usr/lib/python2.2/distutils/sysconfig.py", line 313, in _init_posix distutils.errors.DistutilsPlatformError: invalid Python installation: unable to open /usr/lib/python2.2/config/Makefile (No such file or directory) jamest: I don't want CVS head, I'm trying to sync to the release branches. jeffb@desert:~/cvstree/gnue/forms/CVS$ cat Tag Tforms-0-4-1-branch common-0-4-1-branch larsu that looks like a bad disutils/python combo jbailey: hang on jeffb@desert:~/cvstree/gnue/common/CVS$ cat Tag Tcommon-0-4-1-branch that's what i gathered from the error. :) i'm just using a default redhat install i think. Chipaca> for i in common forms deskigner; do cvs update -Pd -r ${i}-0-4-1-branch $i/; done what is this disutils you speak of? speliing designer properly of course larsu: i bet there is a python-dev rpm of some type larsu: disutils is an 'installer' that comes with python on debian you must install it separately on rh6.2 python2.x it came native BUT yeah, sounds like jamest is right. will look. maybe as jamest is hinting too they have separated it into the -dev packages right which is more debianesque if you dont have the -dev installed, install it and if problem persists let us know will do or if you want to really fix the problem you could upgrade to debian ;) :) Action: revDeke warns at some times there are more debian developers than gnue developers around here if nothing else it makes for a good place to get debian support :) Action: jbailey hides. i was going to install debian on an rs6000 box last night, but it didn't like some scsi stuff. it runs netbsd now instead There are no Debian developpers here. These are not the droids you're looking for. hehe move along oh, jbailey.. you're a debian developer? let me send you the errors from last night.... j/k :) larsu: i think that scsi problem was probably the damn maintainer or libc not having his act together you should use reportbug libc and file a bug ;) revDeke: Sure. At the rate we're going, we're not going to see libc in testing anytime soon. =) it was on a weird ass ibm 43p rs6000 box.... that it even booted off the debian floppies was impressive actually there are some tricks i had to make a special boot disk and do a bf2.4 with some params to get a compaq machine here at work to recognize the raid controller worked fine once i found the magic commands netbsd was no picnic either. i had to dd a boot sector to the hd :) Action: revDeke scribbles down not to buy any rs6000's any time soon ;) fortunately it was surprisingly well documented. it only took about 5 installs to get it right i didn't look for the docs until halfway through tho :) Action: jamest looks forward to finally seeing sarge w/ kde 3.1 gcc 3.x and gnue move to stable in 2007 grrrrrrr [17:58] Last message repeated 5 time(s). Action: jcater is working on the code editor in designer which uses the kick ass scintilla / wxStyledTextCtrl widget yes, i was missing python2-devel....rpm. installing it fixed the prob and trying to find stuff on their events http://www.google.com/search?hl=en&lr=&ie=ISO-8859-1&q=EVT_STC_UPDATEUI grrrr jcater: isnt it great when we are the leading experts on the stuff we are clueless about :) according to google jamest: If I understand right, there's a push to have a sarge release next spring or so. revDeke: it makes life very difficult because to find my friggin answers, I get to peruse C code :) someone from my lug told me they were searchig for some package in debian and the answer was in our irc log :) ah shit the other one I need to find info about http://www.google.com/search?q=EVT_STC_MARGINCLICK&btnG=Google+Search&hl=en&lr=&ie=ISO-8859-1 Action: revDeke grunts sorry no dice i went at lunch Action: jcater throws his hands up in disbelief alrighty, time for my own fscking website for this control so I can find info later on when i Need it grr piss moan but I guess it beats doing perverse goat pics have you seen this? http://wiki.wxpython.org/index.cgi/wxStyledTextCtrl jamest: Can you confirm that I've still got the right branch, even if it's not 0.4.1.99 ? revDeke: yes it is code but gives at least list of events you see this http://www.rexx.com/~dkuhlman/#pyeditor down below he has a wxEditor based on the wxStyledTextCtrl perhaps his code is worth looking at? Action: jcater has a full-featured editor in there complete with syntax highlighting keyword completion etc Action: jcater was just trying to find some undocumented events but I think I found it does anyone has a unimportant / rubish DBF file lying around? larsu: you having any luck over there? Action: siesel wants to test the still uncommited DBF dbdriver :) jbailey: that branch should have been called -0-4-x but hindsight is 20/20 the branches will be 04.2 0.4.2 cvs head should be 0.4.99 IIRC we went to this setup case that crazy debian maintainer of ours packaged up cvs again :) jcater: so wxeditor is of no use as a reference Action: jcater thinks I based mine off the same stuff he did chillywilly (danielb@mke-24-167-202-50.wi.rr.com) joined #gnuenterprise. ok Action: revDeke didnt realize you were that far along Action: revDeke thinks you should 'base' some stuff off of boa work (the cvs integration) ;) or maybe i will play with that is that stuff sufficently abstracted to where one could do a plugin? um I looked at boa and talked w/the author his stuff is horribly tied into his boa product we talked about how it would be neat if he could abstract so others can use he thought a good idea but that would be a helluva lot of work after looking at it jamest: I was more wondering why the setup version that common produced was stil 4.1 and the forms and designer produce 4.1.99 but we have his blessing to "reuse" his code yeah, installing the python2-devel...rpm fixed it jbailey: because in common/src/__init__.py we forgot to flip _release off commiting now commited should now be 0.4.1.99 bigbrother joined #gnuenterprise. sigh revDeke (www-data@alt1.libertydistribution.com) left irc: "CGI:IRC 0.5 (EOF)" try again Action: jamest is working on house and doing drive by cvs commits isn't that dangerous? only to the code man i can't take time to aim properly All good - getting ready to spin some new debs. And report more glibc bugs. =) jeff jeff he's our man if he can't d it nobody can! s/d/do/ what's boa? web server I heard of it but I don't remember what it is k and what was derek and jason talking about? jamest (~jamest@adsl-64-216-107-182.dsl.tpkaks.swbell.net) left irc: Read error: 60 (Operation timed out) just out of curiousity http://boa-constructor.sourceforge.net/ what I meant was what the hell does boa have to do with gnue in regards to their convo the reverend said somehting about not realizing that they were that far along ah, whatever Action: chillywilly is away: food hah! from wxGrid docs (this is new) Note: there is a bug in the current grid implementation that makes it impossible to edit cells which are wider than the grid's displayed client area (i.e. the size of the grid window minus the labels). For this reason you should ensure that your renderer's return conservative values from GetBestSize or avoid using automatic sizing until the bug is fixed. so I wasn't on crack KeithJagrs (trilluser@63.74.24.9) joined #gnuenterprise. hello! dsmith (firewall-u@cherry7.comerica.com) left irc: Remote closed the connection jamest (~jamest@adsl-65-71-168-128.dsl.tpkaks.swbell.net) joined #gnuenterprise. siesel (jan@xdsl-195-14-205-192.netcologne.de) left irc: "night" mdean (~mdean@mkc-65-28-72-76.kc.rr.com) left irc: Read error: 113 (No route to host) File "/usr/lib/gnue/lib/python/gnue/common/cursing/Notebook.py", line 69 ctrl = ^ SyntaxError: invalid syntax And a few others like that. Anyone around to help me fix? Remosi (dahoose@port-219-88-246-86.jet.net.nz) joined #gnuenterprise. Nick change: SachaS_sleep -> SachaS IsoosI (dahoose@port-219-88-247-227.jet.net.nz) left irc: Read error: 113 (No route to host) jbailey: committed any others? Yup. =) SyntaxError: invalid syntax File "/usr/lib/gnue/lib/python/gnue/common/dbdrivers/mxodbc/DBdriver.py", line 127 def _buildQuery(self, conditions={}forDetail=None,sql=""): ^ SyntaxError: invalid syntax File "/usr/lib/gnue/lib/python/gnue/common/cursing/Notebook.py", line 69 ctrl = ^ SyntaxError: invalid syntax File "/usr/lib/gnue/lib/python/gnue/common/dbdrivers/mxodbc/DBdriver.py", line 127 def _buildQuery(self, conditions={}forDetail=None,sql=""): ^ SyntaxError: invalid syntax sigh On the upside, the designer wizard works *much* better. =) hang on someone already fix cursing? I fixed the cursing one jbailey: guess what, I just broke it again in cvs head :) commited mxodbc fix Action: jcater hides jcater: Lovely. jbailey: necessary evil :( Keeping in mind the compile is now happening at install time.. or would that be >:) Error messages like this will generate a ton of bugs. =( why are we doing at install time again? That seems to be the way dh_python wants it done. dh_python? debhelper's python tool. It auto detects what version of python you need and sets up the build depends right. It also decides if your modules need to be compiled, and puts in rules to do so at install time. Emacs byte compiles at install time, too. ok So is now a good time to do a cvs update? And is it all in common? Bah, mousewheel doesn't work in designer. how bizarre, it works in the code editor window but not the layout or property editors hmm that's weird i bet we have to handle those scroll events isn't the code editor based upon some other hunk of code yeah gnue-designer doesn't seem to exit cleanly when called from the command line. wa? ctrl+\ always works for me what's mroe clean than that *lol* I'm thinking from file->exit actually, what is it doing? it usually closes for me jeffb@desert:~/debian$ gnue-designer and it's just sitting there. that's a print statement hidden in the code i'll find it ! KeithJagrs (trilluser@63.74.24.9) left #gnuenterprise. Building another common package with new CVS commits. good morning jbailey: did you run any wizzards or anything? Yes, the designer wizard. the multipage/block one? That sounds right. gnue-common installs without error this time. Are you looking for the printf? i've found it err.. print ? i'm looking for more :) *lol* committed which module? designer Tx. Wow. *two* patches. =) Hmm. I should write a manpage for gnue-designer. it's built in run gnue-designer --help there is an option along the lines of --generate-man-page Oh yeah, that's right we talked about this. And the problem is that I can't run it until after it's installed. you are building from cvs? setup-cvs.py then gfdes --generate-man-page I'm doing this: GNUE_VERSION_SUFFIX=-`date +"%Y%m%d"` ./setup.py sdist you won't have to install it That generates my release tarball. And I put that into the package. If I go straight into gnue-designer and back out it exits fine. so we need to add generation of man page to our base setup then right? Yes, please. jamest: it's there if you've run ./setup-cvs.py Action: derek is away: I'm busy then ./setup.py sdist will use ~/bin/g?cvs --generate-man-page Action: derek is back (gone 00:00:01) hmm od rdean (~rdean@12-221-229-247.client.insightBB.com) joined #gnuenterprise. I have a really pedantic wishlist item for you guys.. It would be *really* nice if designer were Xinerama aware. =) Action: jcater groans what is Xinerama? scroll wheel support and xinerama :) Xinerama is the beautiful setup where you have two monitors hooked up side by side and X thinks it's one big desktop. But non-xinerama apps have this habit of popping windows up in the middle of the screen. Action: jcater thought that was a function of the window manager, not an application? So to use them you have to drag them to one or the other monitor. oooooo the answer to this is please send the dual monitor setup to
so we can properly test it out :) ;) Actually, it's surprisingly cheap. The Matrox card was $140 CDN. Action: jcater had a xinerama setup until debconf ate my xf86config-4 file and I never got around to rebuilding it Yeah. You have to turn off manage with debconf. =( Do you want a copy of mine? so I wasn't on crack <-- me thinks you still are gee thanks Hmm.. gnue-designer doesn't seem to want to hang the same way for me on exit now. Also, I don't know if it's the compiling or if it's the released version, but the speed is now usable. http://www.gnuenterprise.org/~jcater/screenshots/designer-docks2.png example of the syntax-hiliting code editor Weird, I can't resize the wizard windows. Is this expected? hmmm... my fonts aren't really that blurry on screen for now, yes :-/ Woohoo! I can actually spend time making this work tommorow. =) cutting it close to the wire aren't you? Incredibly. I also have to build a tool that generates layer 2 maps of our networks. All by Friday 4pm. bah who needs sleep? it's highly overrated I couldn't have really started before today, since the box I need to work on couldn't see the routers and such until Friday. :) And as it was from Friday night to this morning I only got, ummm.. 18 hours of sleep. =) "Works well under pressure" jbailey: the upside seriously is how many other projects do you get patches real time "It's a pitty he creates it himself" Action: derek loves that about gnue derek: Day or night, it seems. =) wxwindows has no explicit multi-monitor support yo yo yo jcater: that hilighting windows is r0x0r Action: chillywilly is back (gone 02:03:18) jamest: Oh well. I'll wait until the gtk2 port shows up and harass you again. =) DB000: Exception exceptions.AttributeError: "wxEvent instance has no attribute 'thisown'" in > ignored DB000: Exception exceptions.AttributeError: "wxEvent instance has no attribute 'thisown'" in > ignored damn wx I didn't notice when that happened, sorry. i imagine that one will have to wait Sure, that's fine. I'm less worried about spewage to console than I am about install time errors. i'd prefer to work from the 0.5.0 base on that as I've noticed some other oddities disappear in head s/dis// Well, designer produced a .gfd. And it seems to load in forms. Very cool. jbailey: test away I think I'm going to go home now and get more of this tommorow. Do you think it's reasonable to upload the packages I have now? um what will they be called? jeffb@desert:~/debian$ ls *deb gnue-common_0.4.1.99-20021125_all.deb gnue-designer_0.4.1.99-20021125_all.deb gnue-forms-wxgtk_0.4.1.99-20021125_all.deb that's ok w/ me is that gonna work? Action: jcater thought the last batch was gnue-*_0.4.1a-...... jcater: It passed the inital filter here. Apparently . is lexically greater than a. But we'll find out, I guess. =) otay we're close to a 0.4.2 anyway IMHO jbailey: fwiw the gtk2 already exists please package it too oooo i don't think it's ready jamest: I didn't know you were doing an 0.4.2 before 0.5.0 jamest: i think of it like curses i swore it wasn't feature complete I don't either packaging it saying 'testing' only is a good thing get some people looking at it only reason i havent tested is have been too lazy to get all dependencies Action: jcater doesn't need bug reports on stuff I know doesn';t work derek: I'd really rather hear consensus on 'upload this' first... a deb package woudl be nice jbailey: i would say it depends on how soon jamest is ready for 0.4.2 if more than a day or two probably its good to upload it if he will have it ready before turkey day no sense likely in putting one out there just yet The one that's out there is broken: Designer's wizards don't do anything useful. we can package 0.4.2 tonight what? cluttering up the console isn't useful? jamest: If you're comfortable with that. let me try and squeez one that thing in jamest: atta boy jamest: Is the 0.4.1 branch going to turn into 0.4.2 branch? nothing says we cant have a 0.4.3 if we find more stuff no I don't foresee a 0.4.3 from now on I only want one branch if possible jcater: i would rather have a 0.5.0 than a 0.4.3 believe me 0-4-x-branch forms 0.5.0 is stable 0-5-x-branch jamest dont do branches do tags jamest: It would be nice if I could track the release branch. derek: tags are a given no im saying we have to have those to do a release you branch for the first one and after that its just tags jamest: So that I know that the 0-4-x-branch is stable plus bug fixes, or non invasive feature adds. 0.4.2 0.4.3 0.4.4 etc are all just tags not branches as in the future saying we will only have one x.x.X release wont likely fly derek: that's what I said as major changes will take longer and longer the bigger the user base becomes ? 0-4-x-branch i thought you said you only want ONE 1 branch marked 0-4-x yes agree to one branch from that branch i tag on release oh crap im sorry i missed this critical line jamest: Is the 0.4.1 branch going to turn into 0.4.2 branch? and your response was 1 branch if possible i didnt see that and saw only nothing says we cant have a 0.4.3 if we find more stuff no I don't foresee a 0.4.3 from now on I only want one branch if possible Nick change: jcater -> cjcarter jbailey: at this time cvs is anything but stable changes HEH and was thinking the i dont foresee a 0.4.3 was you and not jcater talking maybe ppl will get us less confused w/this? the alterations between 0.4.0 and 0.4.1 IIRC were significant Nick change: cjcarter -> jcater cj! Action: derek slaps self with trout yip cjcater dat's da name your first name is not Jason? no my first name is C. as in C. Jason Cater brandy said the c stands for 'crabby' but i said not likely for Codemaster Codemaster Cater hmmm well Scotty's name like my Dad's is D. Scott ;) i have it on file to FSF as Codeslave Cater Chris? d for daddy? Action: jcater always thought it was Cashless Cater uh, no chillywilly: yes well, Christopher Douglas Scott Action: jcater is C Cater my sisters married name is moore pronounced 'more' Action: jcater wanted to name TJ C++ Cater she teased about kids names and the next one C# Cater Lon Moore but wife didn't like that Noah Moore etc etc etc That would be an awful name for a child! jeffrey jcater: you need to see an eye doktor them screenshots done make it seem like you are using 400dpi font or something ;) dude how do you change GTK font sizes? seriously at home, it's tooooo friggin small to read GTK under KDE his momma told him he'd go blind use the gnome dialog? here at work ;) it's huge! but he didn't listen gtk requirements: libpng gnome Home time *poof* jbailey (~jbailey@atlas.fundserv.com) left #gnuenterprise ("Client Exiting"). jamest: ? what i do (not that its right) is comment out the line in XF86Config-4 that uses the 'font server' and then put 75dpi above 100dpi in the list and it makes tolerable fonts for me gack samuel_xuxu (~samuel_xu@hy.cz.jsinfo.net) joined #gnuenterprise. Action: jcater uses scaled fonts anyway as the font server done picks fonts for the blind I'll be damned Action: derek is weird about font size there's a ~/.gtkrc-kde where kde overrides gtk font sizes kick arse Action: derek notes im leaning towards kde because it has smoother fonts Action: jcater hugs his KDE Action: derek wonders why he hugs an application that steals settings hi guys dude I wish it'd steal even more normally that is a 'bad' thing I wish it would apply my KDE style to the GTK apps samuel_xuxu: hi hi samuel_xuxu samuel_xuxu: howdy jcater: would be nice if you could set compatiability mode like you can in sawfish you can say sawfish go check the gtk theme if there is sawfish theme with similar name try to apply it or the likes Action: chillywilly jst opens preferences:/// in nautilus click on fonts ;) same with gkrellm would be cool if kde/gnome would do that with apps derek: gnome 2 has purty fonts i.e. if using one or the other set compatiablity mode for themes samuel_xuxu: anything on topic we can do for you? gnome2 wont install for me i.e. i apt-get gnome2 stuff and dont get gnome2 i gave up kde3 installs for me apt-get install gnome they got rid of the "2" on the ends for unstable Nick change: jcater -> kater whahahaha oooo I like this nick anyone knows of a XML schema file (xsd) to a class (ruby) convertor ? Nick change: kater -> jcater Nick change: chillywilly -> gchillywilly Nick change: gchillywilly -> chillywilly deke! how do I install java on debian for galeon again? yo get the binary from the ftp site and make a symlink lies nope yes I installed it with apt I am telling the truth you are a liar! I never used apt with java as I cheat virtual rms that way ;) he can't yell at me then ;) gasp! SachaS: sorry, not too many ruby ppl here :( nicr the blackdown java has a apt repository ruby == sux0rz ;) more lies! ruby is cool how can the gune forms be distributed? nickr, jamest: www.blackdown.org I think ruby is cool I havce no opinion of it I want deke to give me the easy answer. :) it's python-esque in many ways I just wanna start some shit if I wasn't on python, I'm sure i'd be on ruby but I'm not ftp://metalab.unc.edu/pub/linux/devel/lang/java/blackdown.org/debian/dists i started with ruby a couple days ago, looks nice actually. havent done anything with python, except installed them for ngue someone smarter than i can turn this into a sources.list entry s/ngue/gnue samuel_xuxu: i don't understand the question? I think python and ruby try to accomplish the same goal from what I hear4 suck less? yeah 1 application = several forms, how to release them? by a single exe file? ngue rocks! i'd use inno compiler to include all the stuff we include in the exe and add in your gfd files all our inno stuff is in cvs last I knew in gune dev env, where is inno compiler? nickr http://www.gnue.org/~dneighbo/notes/java_plugin.html danke samuel_xuxu: http://www.jrsoftware.org/isinfo.php is the inno installer though recently their packages were broken (looked like they were restructuring them) http://www.mcmillan-inc.com/install1.html is the actual "compiler" per se so gnue app should be released with the python library? derek: I like the "disclaimer" ;) jamest: what about gear files he wanted it all in 1 .exe file the only way to do that would be to rebundle our install w/ his extra stuff bah if downloading form seperate was ok then a gear file should work derek: they are broken :( can't connect jrsoftware jamest: im more thinking gear + inno so to speak like making gear files and then just bundling our .exe's with the gear files hmmm jrsoftware.org is up for me samuel_xuxu: do you have a gnue app ready to ship or rather just curious about how to do it if you did? I'm just test it's usability from my old dev exp. what is old dev experience i.e.platform, tool database my previous development Action: derek has used many i might be able to offer some comparison in erp systems is there a working GNUe CVS "blame" anywhere? jcater,what's cvs blame? show who added a specific line in our source code yeah so you can assign blame oh. anybody can contribute to the cvs? shows a cvs file, with each line having a cvs file version number beside it we take patches chillywilly (danielb@mke-24-167-202-50.wi.rr.com) left irc: "Free Your Enterprise! - http://www.gnuenterprise.org" Action: jcater was just trying to do some research on some changes I found jcater: i think lxr does that well, but its old derek: url? we should get lxr and bonasi back up and working was lxr.gnue.org that not work :( hmm we should get those working again i used lxr all the time to track brokeness would you like to give me some recommend on how to join develop for here? http://www.gnuenterprise.org/community/involve.php maybe start from python language? sorry didnt mean to be cold in just pasting a url we suggest all contributors regardless of background download and install the tools as a starting point and get the a sample or two working i've download it , and try them this gives a little idea of what exists and at what level the tools are at cool then usually the next question is but maybe u can recommend me the python resource. after installing and looking at them, what looks like an area that interests you? http://www.python.org is a good jump point on that front page there is a link to a good tutorial if you are looking to write applications with the tools and not actually extend the tools i think you will be impressed to find you dont need much if any knowledge of python ok. i have to struggle with the big snake now. though knowledge of python is certainly helpful i also think you will find this channel particularly helpful in answering python questions :) python is being used in many areas? for gnue its only language we really use oh.thx.when i got question, i will go to here for help. why we choose python, not C or C++? * speed of deployment * readability of code * ability to prototype * easy, native cross-platform support um and because python rocks our socks :) but seriously wooo Monty Python :) rock sock? sorry, I guess that's an English idiom the developers really like python it's proven a good choice from the beginning Why python must win? ? oh. i c. I go for my lunch now. see u all. hehe SachaS (~Sacha@dialup-196-95.wasp.net.au) left #gnuenterprise ("Client Exiting"). samuel_xuxu (~samuel_xu@hy.cz.jsinfo.net) left irc: SachaS (~Sacha@dialup-196-95.wasp.net.au) joined #gnuenterprise. we need to add that C/C++ question to our FAQ we have Why not Java but no c/c++ ok stupid time as opposed to? Action: jcater hides on redhat how can you see what packages are installed damn Action: derek hasnt done redhat in a while you're too quick jcater rpm -qa? or something try it! duh man rpm rpm -qa IIRC Action: jcater heard rpm -qa `rm -rf /` works well um do you know how to use up2date don't cut-n-paste that btw or whatever their 'apt like tool' is no jamest (~jamest@adsl-65-71-168-128.dsl.tpkaks.swbell.net) left #gnuenterprise. Action: drochaid is away: not ere, pretending to sleep jcater (~jason@w202.z065105010.mem-tn.dsl.cnc.net) left irc: "home" SachaS (~Sacha@dialup-196-95.wasp.net.au) left #gnuenterprise ("Client Exiting"). rdean (~rdean@12-221-229-247.client.insightBB.com) left irc: "Client Exiting" --- Tue Nov 26 2002