09:06 -!- cnieves [~cnieves@201.pool85-57-129.dynamic.uni2.es] has joined #geda
09:06 < cnieves> Hi
09:10 <@Ales> good morning Carlos
09:10 <@Ales> well.. morning for me at least. :)
09:11 < cnieves> Good morning!
09:11 < cnieves> I expect you a bit later... 
09:12 <@Ales> yeah, I got a up a little earlier. :)
09:12 < cnieves> did you have your breakfast? :-)
09:12 <@Ales> not yet.. going to go get some cereal right now. 
09:13 < cnieves> then, take your time! we can start when you are ready :-)
09:13 <@Ales> I'm ready now. 
09:14 <@Ales> long story short, I did make a release last night
09:14 < cnieves> yeah, I saw the mail and the cvs logs... quite a bit of work, wasn't it?
09:15 <@Ales> no, not really, just the usual, version numbers, a few last minute patches, and doc updates
09:15 <@Ales> I did apply Patrick's dialog button reorder patch.
09:15 <@Ales> it was about 3 hours of work
09:16 < cnieves> yes, I liked that. I'm going to add the gtk dialog button reordering (later)
09:16 <@Ales> that's fine
09:16 < cnieves> well, regarding the patch we are dealing with today, I have the following plan:
09:17 < cnieves> Look at: http://sourceforge.net/tracker/index.php?func=detail&aid=1551233&group_id=161080&atid=818428 , first apply the least intrusive patches
09:18 < cnieves> which are:
09:18 < cnieves>   - improve_o_text_error_detection
09:18 < cnieves>   - dont_redraw
09:18 < cnieves>   - don't duplicate code in o_redraw
09:18 < cnieves>   - selection_glist
09:18 < cnieves>   - copy_using_complex_place_list
09:19 <@Ales> I'm giving the patches a once over look again
09:19 < cnieves> I have here an updated CVS version for each patch, except the last one, which depends on the others.
09:19 < cnieves> Yes, look at them (in this order). The first ones are quite simple.
09:22 <@Ales> first two look okay
09:22 < cnieves> ok, should I commit them?
09:22 <@Ales> yeah
09:24 < cnieves> I doubt about the fprintf(stder,"..."); lines. Should they go into the log as well?
09:25 <@Ales> It's a toss up.  those are kind of critical and I usually look at errors in the console window
09:25 <@Ales> at some point we should go through the entire code base and change all printfs to log
09:25 <@Ales> third patch looks okay, although, I will look at it again when it's commited using tkdiff
09:26 < cnieves> ok, I'll add it as is.
09:27 < cnieves> First one commited.
09:28 <@Ales> I'm rebuilding right now, should be ready to test it in about 7 mins
09:28 < Igor2_off> hi
09:28 <@Ales> hi 
09:28 < Igor2_off> >ales> i have a problem with gaf
09:28 < Igor2_off> autoconf-problem
09:29 < Igor2_off> it puts [config.h].in in the Makefile instead of config.h
09:29 < Igor2_off> or config.h.in or anything
09:29 < Igor2_off> if i symlink or copy my config.h.in to [config.h].in prior to starting the autoconf stuff, it works, but it doesn't seem intended :>
09:29 <@Ales> huh
09:30 <@Ales> what version of autoconf/automake?
09:30 < Igor2_off> (booting the comp to find out)
09:32 <@Ales> I don't recall ever hearing of such a problem
09:32 < cnieves> Second one too.
09:32 <@Ales> updating now
09:33 <@Ales> for some odd reason my cvs updates have been taking longer and longer
09:33 <@Ales> lemme test this change a little and then you can apply the next one
09:33 < cnieves> maybe because it is checking the wiki pages as well?
09:33 <@Ales> maybe
09:34 <@Ales> yeah, there are a lot of files in there
09:34 <@Ales> I should try to restrict my update to just code
09:34 < cnieves> ok. I'll prepare the 3rd while you test this one (or these two?).
09:34 <@Ales> k
09:34 < cnieves> You can do a cvs update gschem libgeda, since today we are going to change only those directories.
09:34 <@Ales> yup
09:34 < Igor2_off> $ autoconf --version
09:34 < Igor2_off> autoconf (GNU Autoconf) 2.60a
09:34 < Igor2_off> Written by David J. MacKenzie and Akim Demaille.
09:35 < Igor2_off> i'm logged in now so i can try reproducing the problem
09:35 <@Ales> some version now
09:35 <@Ales> same version as you 
09:35 < Igor2_off> or i can even give you shell account to test
09:35 < cnieves> Igor, I have the same version too.
09:35 <@Ales> what about automake?
09:36 < Igor2_off> $ automake --version
09:36 < Igor2_off> automake (GNU automake) 1.4-p6
09:37 < Igor2_off> now i try to reproduce the error to be able to copy the results here
09:37 < cnieves> I wonder if automake is too old. I have 1.9.6 version here, and I remember there were some problems with old versions....
09:38 < Igor2_off> we will see :)
09:38 < Igor2_off> now, i have run distclean in gaf/gattrib
09:38 < Igor2_off> $ ./autogen.sh
09:38 < Igor2_off> find: warning: Unix filenames usually don't contain slashes (though pathnames do).  That means that '-name ./CVS' will probably evaluate to false all the time on this system.  You might find the '-wholename' test more useful, or perhaps '-samefile'.  Alternatively, if you are using GNU grep, you could use 'find ... -print0 | grep -FzZ ./CVS'.
09:38 < Igor2_off> processing .
09:38 < Igor2_off> autogen.sh running: aclocal  ...
09:38 < Igor2_off> autogen.sh running: autoheader ...
09:38 < Igor2_off> autogen.sh running: automake  ...
09:38 < Igor2_off> configure.ac: 14: required file `./[config.h].in' not found
09:38 < Igor2_off> autogen.sh running: autoconf ...
09:39 <@Ales> ah, I have a much never version of automake as well.
09:39 <@Ales> I have 1.7.9
09:39 <@Ales> could you upgrade your automake?
09:40 < Igor2_off> checking
09:40 <@Ales> is this your first time building gEDA/gaf?
09:40 <@Ales> what linux distribution?
09:40 <@Ales> first two patches seem to be okay carlos
09:40 < Igor2_off> debian testing (etch)
09:40 < Igor2_off> yes, my first time
09:40 < Igor2_off> doing this because i need to put new versions to my geda live cd 
09:41 < Igor2_off> because xgsch2pcb for my students :>
09:41 <@Ales> yeah, definately upgrade your automake then, I'm using etch as well
09:41 < Igor2_off> # automake --version
09:41 < Igor2_off> automake (GNU automake) 1.8.5
09:41 < Igor2_off> yeah, it seems they have different packages
09:41 < Igor2_off> automake1.4 i had
09:42 < Igor2_off> now installed this automake1.8 and there's automake1.9 available
09:42 < Igor2_off> $ ./autogen.sh
09:42 < Igor2_off> find: warning: Unix filenames usually don't contain slashes (though pathnames do).  That means that '-name ./CVS' will probably evaluate to false all the time on this system.  You might find the '-wholename' test more useful, or perhaps '-samefile'.  Alternatively, if you are using GNU grep, you could use 'find ... -print0 | grep -FzZ ./CVS'.
09:42 < Igor2_off> processing .
09:42 < Igor2_off> autogen.sh running: aclocal  ...
09:42 < Igor2_off>   run info '(automake)Extending aclocal'
09:42 < Igor2_off>   or see http://sources.redhat.com/automake/automake.html#Extending%20aclocal
09:42 < Igor2_off> autogen.sh running: autoheader ...
09:42 < Igor2_off> autogen.sh running: automake  ...
09:42 < Igor2_off> oops, one line missing from the dump
09:42 < Igor2_off>  /usr/share/aclocal/glib.m4:8: warning: underquoted definition of AM_PATH_GLIB
09:43 < Igor2_off> so still not perfect, but the [] bug is gone
09:43 < Igor2_off> i believe autogen.sh should check for automake version and warn the user or something :)
09:44 < Igor2_off> (and i don't see any reference in the docs about which version is required, so it's not too trivial:)
09:44 <@Ales> I see similar warnings sometimes too, probably due to the new version of automake, but it should work
09:45 < cnieves> I'm getting a lot of "underquoted definition" warning too. Automake became stricter, and there are some macros out there which haven't been updated.
09:45 < Igor2_off> ok
09:45 < Igor2_off> then i check the find error
09:45 < Igor2_off> what about the version check, is it possible? :)
09:45 < Igor2_off> i think it would help newcomers like it would have helped me :)
09:47 < Igor2_off> ok, about find:
09:48 < Igor2_off> this is the original version, in the for:
09:48 < Igor2_off> find $srcdir -name $srcdir/CVS -prune -o -name $configure_script -print
09:48 < Igor2_off> if you use -wholename instead of -name, it works
09:48 < Igor2_off> (without warning)
09:48 <@Ales> but the problem is that -wholename might not be present in all version of find
09:48 < Igor2_off> i'm not sure if it's a portable solution, if you want, i can check it on a few hosts
09:48 < Igor2_off> ohh
09:48 <@Ales> I think it's a gnu extension. 
09:49 < Igor2_off> well, what scares me is not the fact it gives a warning
09:49 <@Ales> carlos: I looked over the commited changes again.  look reasonable.
09:49 < Igor2_off> but that it says it will fail anyway :)
09:50 <@Ales> this is kind of a fun... a mini code sprint. :)
09:51 < cnieves> Ales: yes, :-) . ok, should I commit the 3rd (o_redraw) one?
09:51 < Igor2_off> >ales> i'm for this kind of code sprints any time if it doesn't annoy you ;)
09:52 < Igor2_off> (trying to find a solution that actually works without a warning and is portable)
09:52 <@Ales> what is the purpose of the flag on o_redraw?
09:52 < cnieves> let the caller choose wether draw the selected objects or not.
09:52 <@Ales> yeah, I wouldn't mind getting rid of the warning as long as the solution is portable :)
09:52 <@Ales> okay
09:53 < Igor2_off> so far i couldn't think anything better than testing what version of find we have, hehe
09:53 < Igor2_off> hmmz
09:53 < Igor2_off> why do we need $src/CVS, why not just CVS?
09:53 <@Ales> carlos: commit away on o_redraw patch
09:54 <@Ales> Igor: just to filter the find down a little bit
09:54 <@Ales> keep in mind I didn't really write the autogen.sh scripts, I snagged them from glade a while ago and make some changes here and there
09:55 < Igor2_off> ok
09:56 < Igor2_off> `find $srcdir  -name $configure_script -print | grep -v "^$srcdir/CVS"`
09:56 < cnieves> Ales: o_redraw_patch commited.
09:56 < Igor2_off> what about this one?
09:56 < Igor2_off> it runs 2 processes
09:56 < Igor2_off> but filters out CVS :)
09:59 < Igor2_off> (less elegant than the original, for sure)
10:01 <@Ales> carlos: okay, this patch builds and seems to be okay
10:02 <@Ales> is selection_glist next?
10:02 < cnieves> Ales, the next one is to convert the selection list into a glist...
10:02 <@Ales> the big one.. 
10:02 < cnieves> maybe it's better if I send you the patch first, instead of commiting it directly.
10:02 < cnieves> not really. The last is the big one. :-)
10:02 -!- patb [~patb@AGrenoble-203-1-7-168.w80-13.abo.wanadoo.fr] has joined #geda
10:03 < patb> hello all
10:03 <@Ales> Hi Patrick
10:03 < cnieves> Hi Patrick
10:03 < Igor2_off> >ales> so is that grep-approach ok? :)
10:03 <@Ales> boy, everybody is coming. :)
10:03 < Igor2_off> hi :)
10:03 < Igor2_off> :>
10:03 <@Ales> Igor: send me an e-mail with the modifications and I'll take a look
10:03 < Igor2_off> >Ales> everyone comes here to overload you! :)
10:03 <@Ales> or better yet, post it to geda-dev
10:03 < cnieves> patb: we have some kind of mini code-sprint here.
10:03 <@Ales> no no, the more the merrier!
10:03 < Igor2_off> ok :>
10:03 < patb> cnieves: on your patch?
10:04 < Igor2_off> yeah, patb, you just came the right moment :)
10:04 <@Ales> Patrick: Thanks for fixing those gtk+ issues. 
10:04 < cnieves> patb: yes, we have just committed some few changes.
10:04 < patb> cnieves: for zooming/panning
10:04 < patb> ok
10:04 <@Ales> we are getting ready to do the big one
10:04 < patb> yes I noticed them coming
10:04 <@Ales> carlos: yeah send me the patch
10:04 < patb> cnieves: I was just reviewing your patches 
10:05 <@Ales> Patrick: any comments on the patches?
10:05 < patb> cnieves: but I failed to convert the time of the meeting
10:05 < patb> I expected it one hour later :)
10:05 < Igor2_off> bah, i hate cvs :)
10:05 <@Ales> I got here a little early, so we started earlier
10:05 <@Ales> my fault
10:06 <@Ales> I have to disappear around 11ish or so my time.
10:06 < patb> so I got the time conversion right? 
10:06 <@Ales> Yes you did.
10:06 <@Ales> perfectly
10:06 <@Ales> anybody object if I post the log of this mini-sprint after we are done?
10:06 < patb> Good, I really hate these time conversion, never sure I am right
10:07 < patb> I can not stay too much too
10:07 <@Ales> sorry about startting too early.
10:07 < Igor2_off> no objection on my side :)
10:07 < patb> in fact I would have suggested the patch to be applied to a branch instead of the main trunk
10:08 <@Ales> well so far the patches have been fairly straightforward
10:08 <@Ales> however, the next ones are the biggies
10:08 <@Ales> so we could still branch
10:08 < cnieves> Ales: no objection to the log.
10:08 <@Ales> k
10:09 <@Ales> I wouldn't be opposed to a branch for this
10:09 <@Ales> thoughts carlos?
10:09 < Igor2_off> >Ales> btw, xgsch2pcb doesn't have ./configure nor a script that i could run to generate it so it's not trivial what to do there (the doc sais i should just run ./configure)
10:10 <@Ales> Igor: take that up with PeterC or PeterB :)
10:10 < cnieves> agree. It's a better way.
10:10 <@Ales> okay, I'm going to create the branch
10:10 <@Ales> what should I call it?
10:11 < Igor2_off> (and cvs hangs)
10:11 < cnieves> Ales: we have two patches left: one to convert the selection into a glist, and the other is to make copy use the complex place list....
10:12 <@Ales> how about glist_dev?
10:12 < cnieves> ok
10:15 <@Ales> creating branch now
10:15 < cnieves> Ales: I'm having problems using a cvs diff. It doesn't do anything. does the CVS live in moria? a ping doesn't show any answer...
10:16 <@Ales> yup it's on moria
10:16 <@Ales> it might be a little slow right now
10:17 <@Ales> try it now
10:17 < Igor2_off> yeah, i can't access the cvs server either
10:17 < Igor2_off> or hmmz
10:17 -!- patb2 [~patb@AGrenoble-203-1-7-168.w80-13.abo.wanadoo.fr] has joined #geda
10:17 < Igor2_off> actually i can access it trough the university network
10:17 < Igor2_off> but not from home
10:18 <@Ales> branch created
10:18 <@Ales> please update your working directory by running: cvs update -r glist_dev
10:18 -!- patb [~patb@AGrenoble-203-1-7-168.w80-13.abo.wanadoo.fr] has quit [Ping timeout: 622 seconds]
10:18 <@Ales> that'll switch you to the new branch
10:19 < cnieves> no answer... :(
10:19 < patb2> well I am having problems with my connection
10:19 < patb2> good I was saying I am going to post my notes on the SF tracker
10:19 <@Ales> ack
10:19 <@Ales> maybe something has gone bad on the net
10:20 <@Ales> can you ping moria.mit.edu?
10:20 < cnieves> yes, it works.
10:20 < patb2> Ales, works for me too
10:21 -!- patb2 is now known as patb
10:21 <@Ales> how about a cvs diff?
10:22 < cnieves> it works too. Wait a moment and I'll post the diff.
10:22 < Igor2_off> ok, can access the cvs server now
10:22 <@Ales> okay, remember to switch to the glist_dev branch so that your changes go there
10:24 < cnieves> I was going to send you (Ales and patb) the patch...
10:24 < cnieves> before committing
10:24 <@Ales> k
10:26 <@Ales> or better yet, create a new checkout using cvs checkout -r glist_dev
10:26 <@Ales> and name the directory properly
10:28 < Igor2_off> sent the mail to geda-dev :)
10:32 < cnieves> I'm checking out the new branch. If the diff is the same, I can send it now.
10:33 <@Ales> okay
10:34 <@Ales> yeah, I'm checking out the branch as well
10:37 < cnieves> Ales: I'm trying to send you the patch using DCC... can you accept the connection?
10:38 <@Ales> got it
10:38 < cnieves> patb: you have the patch waiting for you...
10:39 < patb> got it too
10:40 <@Ales> I would say apply it to the glist_dev branch, unless somebody objects
10:41 < cnieves> the checkout is still running... so it could take a while.
10:49 < cnieves> I'm compiling. Patrick, I saw your comments on the SF page. I'll take a look at them as soon as the compiling finishes...
10:50 < patb> ok great
10:54 <@Ales> okay, I have to run. I look forward to playing with the changes 
10:55 <@Ales> take care everybody
10:56 < patb> have a nice day ales. I will also leave soon, I have walnut to gather and chestnut to prepare
10:56 <@Ales> enjoy
10:56 <@Ales> see ya Carlos and Igor.  this was fun
10:56 < cnieves> patb: your comment: why not keeping the SELECTION type as an alias to GList?
10:56 < cnieves> it would
10:56 < cnieves>     help differentiate selections from other kind of lists.
10:57 < patb> cnieves: yes?
10:57 < patb> we are already using GList for conns and we are likely to use it elsewhere
10:57 < cnieves> I don't object. What I think is that the code would be better if we avoid to embed the lists pointers into the objects.
10:57 < cnieves> Bye Ales, thank you.
10:58 < cnieves> I hope for objects too, sometime in the future. That will solve the current OBJECT* list and GList * duality.
10:59 < patb> I do not want that. What I meant was: typedef GList SELECTION;
10:59 < patb> BTW your patch is only against gschem, not libgeda
10:59 < patb> it complains it has not the right version of o_complex_add
11:01 < cnieves> I'll do a diff for libgeda.
11:02 < patb> well as long as you commit to the branch I have no problem with applying your patch
11:03 < patb> I consider it as some kind of sandbox
11:04 < cnieves> what do you mean?
11:06 < patb> you can try/broke things in a branch it does not mess with the main trunk
11:07 < patb> so you have all the time you need to fix at later time
11:07 < cnieves> yes, that's the benefits of branches...
11:07 < patb> indeed
11:07 < cnieves> If I understood correctly, you don't want to use glists for objects lists. why?
11:09 < patb> Maybe not for objects lists (the prev and next in OBJECT) but they are great for selections
11:10 < patb> As you said the problem is that we have to keep a reference on the GList in OBJECT if we want to traverse an object list from an OBJECT
11:10 < patb> There is no such problem with selections
11:12 < cnieves> The reason why I didn't use a typedef for the GList was that if, in the future, we use a GList for other things than selections, we won't need to change the name again.
11:13 < patb> yes but not in libgeda/o_selection.c and gschem/o_select.c
11:14 < patb> And it is purely cosmetic 
11:14 < cnieves> well, it's a matter of taste. If I'm going to use glib's glist functions with selections, I should know they are selections. The typedef hides it.
11:15 < patb> then just ignore this point of my comments
11:15 < cnieves> I mispelled "I should know they are selections": I meant "I should know they are glists"
11:16 < cnieves> Regarding the names of get_object_complex_bounds, and get_object_glist_complex_bounds..
11:17 < cnieves> what name do you suggest?
11:18 < cnieves> Maybe get_bounds_of_object and get_bounds_of_object_glist ?
11:18 < patb> at least scrap the complex from their names: they do not work specifically on complex
11:18 < cnieves> IIRC, the complex was there before, so I just kept it there.
11:19 < patb> what about get_list_bounds() and get_glist_bounds()
11:19 < patb> I know it was there
11:20 < cnieves> As they are a lists of OBJECTs, I'd prefer get_object_list_bounds and get_object_glist_bounds... 
11:21 < patb> sounds good to me 
11:24 < cnieves> ok. 
11:24 < cnieves> get_object_glist_complex_bounds() and get_complex_bounds_selection() are the same. I'll change that.
11:24 < patb> Ok I really have to go now. If you have other questions on my comments, mail them to me.
11:25 < patb> BTW it compiles fine but some of the crash I reported are still valid.
11:25 < cnieves> ok. I'll review your comments and mail you.
11:25 < patb> have fun
11:25 < cnieves> Thank you for your work.
11:25 -!- patb [~patb@AGrenoble-203-1-7-168.w80-13.abo.wanadoo.fr] has quit [Quit: Leaving]
11:27 -!- cnieves [~cnieves@201.pool85-57-129.dynamic.uni2.es] has quit [Quit: Abandonando]