2007-03-08 GnuCash IRC logs

00:02:03 <warlord> Probably.
00:03:01 <hampton> There's a define for G_PI. Cool.
00:17:10 <warlord> I should be surprised, but I'm just not.
00:24:29 <hampton> hmm, a remote autopilot that once turned on cannot be turned off. Good idea or horrible idea?
00:25:13 <hampton> http://www.thisislondon.co.uk/news/article-23387585-details/New%20autopilot%20will%20make%20another%20911%20impossible/article.do
00:27:37 <warlord> Well, it's only activated by the pilot onboard, so it cannot be initiated remotely.
00:27:40 <warlord> THAT part is good.
00:27:52 <warlord> I do have to question the security of the remote control, though.
00:28:27 *** dbr has quit IRC
00:29:12 <hampton> I'm worried about what happen if the remote signal is lost.
00:29:32 <warlord> Eh, the FMS would already be programmed.
00:31:34 <hampton> true. all you would need to do once the pilot enable the remote is reprogram the FMS and lock out any local changes.
00:31:48 <warlord> Right
00:32:11 <warlord> Granted, the pilot could still control things like throttle, gear, and flaps.
00:32:29 <jsled> Ah. More movie theatre security.
00:32:36 <warlord> So, the could still crash the plane, but would have much less control over where.
00:32:47 <warlord> yeah
00:32:54 <jsled> "autoland functionality"
00:32:56 <warlord> "Die Hard", anyone? ;)
00:33:06 <hampton> does boeing still use hydraulics for that? Airbus it totally fly-by-wire so you could even prevent that.
00:33:15 <warlord> The 777 is fly by wire
00:33:38 <hampton> autoland is a staple of today's airline fleet
00:33:59 <warlord> yep
00:35:24 <hampton> I've been in an airport (Milwaukee, I think) where the visibility was zero-zero. Half the planes were indefinitely delayed, the rest were taking off and landing on schedule because they had autoland capability.
00:35:47 <warlord> Ahh, the joys of 0-0
00:51:01 <warlord> okay, bedtime.
00:51:04 <warlord> night all
00:51:06 *** warlord is now known as warlord-afk
00:54:07 <foo> warlord-afk: late
01:25:58 <hampton> night
01:26:04 *** hampton is now known as hampton|away
01:45:46 *** vikram has joined #gnucash
01:46:00 <vikram> hey
01:46:09 <vikram> anyone here
01:52:09 *** vikram has left #gnucash
01:52:15 *** vikram has joined #gnucash
01:52:40 *** vikram has quit IRC
03:25:23 *** RallyU has joined #gnucash
04:09:14 *** ErKa has joined #gnucash
04:23:21 *** RallyU has left #gnucash
05:17:22 *** Def has quit IRC
05:17:22 *** cortana has quit IRC
05:17:22 *** conrad has quit IRC
05:17:22 *** slicslak has quit IRC
05:17:22 *** wizkid238_ has quit IRC
05:19:00 *** wizkid238 has joined #gnucash
05:19:00 *** conrad has joined #gnucash
05:19:00 *** Def has joined #gnucash
05:19:00 *** cortana has joined #gnucash
05:19:00 *** slicslak has joined #gnucash
05:52:51 *** ceplma has joined #gnucash
06:32:10 *** egsavage has joined #gnucash
06:43:02 *** IanL has joined #gnucash
07:03:47 *** twunder has joined #gnucash
07:12:56 *** twunder has quit IRC
07:42:53 *** egsavage has quit IRC
08:54:31 *** warlord-afk is now known as warlord
09:07:37 <warlord> Hmm.. interesting.. I just got pointed to http://developer.imendio.com/projects/gtk-macosx/
09:10:12 <jsled> I thought that'd come up in here before.
09:11:58 <jsled> Hopefully that won't take as long as the gtk-win32 backend seemed to...
09:12:20 <warlord> I dont know.
09:12:58 <warlord> I'm not sure what the current status is; they've got a 'todo' list, but not a "here's what works" list.
09:14:45 <warlord> Still... this is off that livejournal entry posted here last night. I've invited her to come help with the port on gnucash-devel, but I also made it clear that "porting to cmake isn't on the table".
09:15:49 <jsled> Yeah. That's very silly.
09:16:06 <jsled> Of course, randomly moving the scheme code to mono is silly.
09:17:39 <chris> who's actually submitting the Google SoC app?
09:17:55 <jsled> warlord, right? How's it coming?
09:18:14 <chris> warlord: you have a google account?
09:19:18 <warlord> chris: yes, I do
09:19:36 <warlord> And.. right. the app. Damn, where as this week gone?
09:43:00 <warlord> Does someone want to be the backup organization administrator?
09:44:30 <warlord> Oh, and if anyone wants to help me answer the application questions: http://code.google.com/soc/org_signup.html
09:45:08 *** IanL has quit IRC
09:55:14 <jsled> I can be backup admin.
09:57:07 <warlord> thanks!
10:01:20 <jsled> Some of these questions are interesting.
10:02:00 <warlord> Indeed.
10:02:03 <jsled> I'm having a hard time imagining a project where "What criteria did you use to select these individuals as mentors? Please be as specific as possible." would need a very specific answer.
10:02:13 <warlord> Hehehe...
10:03:13 <jsled> I guess someone random could come by the project and be all "I wanna sponsor someone for SoC07" and the project's all "uh, okay, sure, whatever. have fun."
10:03:38 <jsled> Also, I suggest all 4 questions in the "About The Program" section should be answered simply by "Beatings.".
10:03:59 <warlord> LOL!
10:06:55 <warlord> I mean, I could certainly unilaterally fill in an application.. BUt I figured it would be "nicer" to get more input.
10:07:35 <chris> warlord: and look at the quality input you get from this esteemed crowd.
10:08:00 <warlord> Eh, I have good filters ;)
10:08:10 <jsled> Heh.
10:09:09 <jsled> Disappearing students is the trickiest one. I mean, if they're really incommunicado, I'm not sure what can be done.
10:09:25 <jsled> If they're just not being productive, we can enter into dialog with them.
10:10:09 <jsled> Get them on a work plan, cooperatively develop, &c.
10:10:39 <warlord> * nods *
10:31:10 *** ErKa has quit IRC
10:42:22 *** jpeach has joined #gnucash
10:42:28 *** jpeach has left #gnucash
11:16:02 <chris> gncbot: tell andi5 I spoke too soon. I still see those errors about PKG_CONFIG_MIN_VERSION.
11:16:02 <gncbot> chris: The operation succeeded.
11:23:50 <warlord> Lovely... One MIT mailserver is dead.. and everyone is losing as a result. I haven't received any mail in hours.
11:30:38 <jsled> :(
11:42:33 *** ErKa has joined #gnucash
11:49:14 *** egsavage has joined #gnucash
11:50:22 <warlord> oooh! 1 email has come through!
11:54:15 <jsled> heh.
12:12:22 *** ceplma has quit IRC
12:18:44 <warlord> BIAB
12:18:46 *** warlord is now known as warlord-afk
12:29:24 *** BlackBsd has joined #GnuCash
13:06:33 *** MrN has joined #gnucash
13:08:03 *** wizkid238_ has joined #gnucash
13:09:52 *** wizkid238 has quit IRC
13:10:12 *** twunder has joined #gnucash
13:20:49 *** warlord-afk is now known as warlord
13:34:09 *** Geot has quit IRC
13:53:21 *** hampton|away is now known as hampton
14:24:40 *** ceplma has joined #gnucash
15:04:14 *** twunder has quit IRC
15:42:59 *** andi5 has joined #gnucash
15:43:00 *** gncbot sets mode: +o andi5
15:44:22 *** benoitg has joined #gnucash
15:53:37 <andi5> chris: wrt autogen.sh: that is weird...... does it fail for you in a fresh checkout or is there a way to reproduce the error?... i was pretty sure it disappeared for me, but well...
15:54:42 *** cortana has quit IRC
15:57:32 <chris> andi5: "fresh checkout" ha!
15:58:29 <chris> oh well, running it twice is easy enough.
15:59:50 <chris> I can hardly believe it.
16:00:06 <chris> I just got date entry to work in the new register.
16:00:17 <andi5> ohh, nice :)
16:00:44 <andi5> well, is it true then?
16:00:56 <chris> "true"?
16:01:02 <chris> oh.
16:01:08 <chris> :)
16:01:14 <andi5> ;-)
16:01:27 <chris> yes, it's true.
16:01:43 <chris> now for the bad news.. hehe.
16:02:25 <andi5> you cannot open the register any more? ;-)
16:02:49 <chris> implementing gtkcelleditable on a subclass of gtkcalendar, and making a gtkcellrenderercalendar subclass of gtkcellrenderertext turned out to have some technical difficulties.
16:03:35 <jsled> It sounds like it should ... ETYPENAMETOOLONG, perhaps? ;)
16:03:50 <warlord> heh
16:04:14 <chris> Apparently, the only way to do such a thing is to implement a cellrenderer that raises a custom popup.
16:04:51 <chris> (That's a non-toplevel gtkwindow, but not a GtkMenu.)
16:05:15 <chris> now for the moderately better news...
16:06:06 <chris> Planner already has a rather clean implementation of all this.
16:07:04 <chris> The design is rather clever. An abstract cellrendererpopup subclasses cellrenderertext, and implements some show/hide popup signals.
16:07:43 <chris> Then, there's the custom popup - basically a GtkEventBox.
16:09:23 <chris> Then, a subclass - cellrendererdate - provides a GtkCalendar for the cellrendererpopup to use in the custom popup.
16:09:46 <chris> oh, important detail: the custom popup implements gtkcelleditable.
16:10:39 <andi5> chris: gtk 2.12 seems to have some calendar changes on its roadmap, they should not be a problem, right?
16:10:48 <chris> I forced the integration, but there must be some better way.
16:11:04 <chris> andi5: I don't know.
16:12:09 <chris> I basically copied 7 files into gnome-utils, added libplanner to the configure.in
16:12:28 <chris> and I had to make one change to a date-format string.
16:13:32 <chris> Planner development seems kinda stalled, so I don't think it would be easy to get them to add the copied files to libplanner.
16:13:46 <chris> actually there kinda is no "them", that's the problem.
16:14:22 <chris> I'm not going to deal with it for now.
16:16:48 *** ceplma_ has joined #gnucash
16:16:49 *** twunder has joined #gnucash
16:20:56 <jsled> chris: well, if it's generally useful, maybe you can propose it for gtk?
16:22:05 <chris> people in #gtk agreed that it was generally useful.
16:23:08 <chris> the cellrendererpopup and editable popup would be pretty easy to whip into gtk shape, I'd think.
16:23:49 <chris> the cellrendererdate uses some planner-specific date/time code where glib equivs would probably do.
16:24:29 <chris> In any case, I'm not up for carrying that flag.
16:24:45 <warlord> :(
16:26:50 *** andi5 has quit IRC
16:32:21 *** andi5 has joined #gnucash
16:32:22 *** gncbot sets mode: +o andi5
16:42:48 <chris> that's pretty much all the hacking I can manage in the near term.
16:43:39 <chris> that date-entry was the most technically-uncertain outstanding work for the register rewrite.
16:44:53 <chris> I think there's mostly just a lot of little UI-improvements to make it more convenient.
16:45:04 *** egsavage has left #gnucash
16:45:09 <chris> and bug-squashing, of course.
16:45:36 <warlord> Well, and then fitting in the various registers/ledgers into the UI.
16:46:48 <chris> well, that's needed to remove src/register, but not for using a gtktreeview to edit transactions.
16:46:58 <andi5> chris: i have no clue, so one stupid question: will it be technically possible to configure shortcuts for the calendar (next month, next year, a.s.o.) ...... (imagine: you have to type Ctrl+] in evolution on a german keyboard to go to the next unread message...)
16:47:41 <chris> andi5: sure.
16:48:08 <chris> andi5: It already works.
16:48:14 <andi5> yeehaw :)
16:48:41 <chris> it's ctrl+arrow
16:48:53 <andi5> you mean on r-r?
16:49:03 <chris> andi5: of course. :)
16:49:16 <chris> it's up/down for years, left/right for months.
16:49:35 <chris> andi5: check it out.
16:50:07 <andi5> will do
16:50:44 <chris> andi5: and fix some bugs, while you're at it, man!
16:51:00 <hampton> can we add the current accelerator keys? [t]oday, start of [y]ear, end of yea[r], start of [m]onth, and end of mont[h]
16:51:11 <andi5> hehe... i would love to, but i have quite some other buggies on my todo list ;-)
16:51:23 <hampton> and yes, i freely admit that those are english specific
16:51:50 <hampton> I use them *all* the time.
16:52:00 <andi5> . for today? ... i cannot remember where i used that, but somehow i try it in every other app ;-)
16:52:02 <chris> hampton: you should check out the widget, too. It's nice.
16:52:38 <chris> hampton: I'm sure we can add them.
16:53:18 <hampton> I'd love to. My current list though is 1) finish check printing, 2) indent, 3) convert price quotes to C, 4) look at the gobject branch.
16:53:49 <chris> but fair warning: there's still a lot of little (important) stuff that's not done. E.g. entering the date by typing.
16:53:59 <andi5> hampton: hej, i remember you said price-quotes are the 2nd on the list...
16:54:37 <chris> does 2) mean come up with settings or actually reformat?
16:54:54 <hampton> propose settings
16:57:13 <jsled> I worked out my preferred settings in terms of indent options the other week.
16:57:31 <jsled> One thing I could not find was a way to control whitespace around '*' ...
16:57:48 <jsled> It seems to want to do {Type * ident}, which is totally wrong.
16:58:45 <chris> jsled: I hope it doesn't have those if-braces on a line by themselves, indented 1/2 way between the inner and outer scope. :)
16:58:47 <andi5> it == indent?
16:58:55 <hampton> lol
16:59:01 <jsled> Sorry, "which is so obviously and totally inane that one would have to be a $insult1 $insult2 ..."
16:59:06 <jsled> chris: oh god no.
16:59:07 <hampton> what settings did you come up with?
16:59:11 <andi5> chris: that is gnu style, right?
16:59:44 <chris> andi5: sounds right.
17:02:51 *** ErKa has quit IRC
17:02:56 <jsled> http://pastebin.ca/386803
17:04:00 <chris> jsled: I'm not going to look at them, but I reserve the right to have a strong opinion after it's too late. ;)
17:04:09 <jsled> heh.
17:05:05 <chris> oh, I didn't expect to see a sample.
17:05:20 <andi5> jsled: is this (1) what you would love to code in ; (2) what you think we should use ; (3) what looks most like the existing code .... i guess (1) :-)
17:05:39 <jsled> 1 & 2.
17:06:00 <chris> I prefer braces on the same line as if/else/for.
17:06:06 <jsled> The obvious difference between the "common" existing is -bl -bli0
17:06:07 <jsled> right.
17:06:39 <chris> for everything except functions.
17:07:53 <jsled> The other difference in the existing code is -nprs, i.e., "function()" vs. {function ()}, which I dislike.
17:07:54 <andi5> http://pastebin.ca/raw/386803 might look better
17:08:25 <chris> looks pretty much the same. :)
17:09:11 <jsled> Er, -npcs , I guess. Hmm.
17:09:24 <chris> andi5: thank you for your novel contribution. :P
17:09:45 <andi5> meck meck... ;-)
17:11:22 <jsled> I used to be a fan of 1TBS, but no longer. vertical whitespace is free, and the visual clarity of the column-matched braces especially on code of any reasonable complexity, is a win.
17:12:15 <chris> wha? vertical whitespace isn't free!
17:12:38 <jsled> I think the other "contentious" issue is line width. 80 characters is too constraining when gnc_dealing_with_a_super_long_typename(EspeciallyWith, SomeArguments).
17:12:57 <hampton> I also strongly prefer braces on the same line as if/else/for.
17:13:00 <andi5> i see, we differ in some points already ;-)
17:13:01 <chris> I want to see a whole function on the screen at once.
17:14:01 <andi5> i guess brace placing and fill-column basically deal with to how large your screen is ;-)
17:14:49 <chris> jsled's got a projector for his wall, ya see?
17:15:21 <hampton> I run with a 188 line high emacs window, but I don't see any point in delineating blocks of code with both indentation and braces on separate lines.
17:15:26 <andi5> i am still waiting for my minority-report-alike-input-whatever-device....
17:15:29 <hampton> 118 not 188
17:15:53 <hampton> 21" screen, small font. ;-)
17:16:17 <andi5> yes, 71 lines here (and i have never had more)
17:16:21 <chris> I program on a vt100.
17:16:32 <chris> in ed.
17:16:40 <hampton> oh, you poor man. This is 2007, not 1989.
17:17:16 <jsled> I got a new punch, and a whole box of 80-column cards...
17:17:24 <hampton> lol
17:18:35 <chris> well, the nice thing is that each vt100 session only takes 2048 bytes, so I can open 500k of them at once.
17:19:03 <andi5> what is the consensus? ... discuss it here, by mail or on the mailing list?
17:20:16 <andi5> chris: emacs can handle more than one buffer
17:20:32 <andi5> do not know about vi though ;-)
17:21:24 <jsled> Well, ultimately the mailing list is the list of record, but it sounds like we're going to consense (?) on some variant of that "other" bracing style, at least.
17:22:20 <jsled> I still think hampton should propose something to the list at his convenience.
17:22:32 <hampton> sure
17:22:44 <andi5> ok
17:22:51 * hampton swaps tasks 1 and 2
17:23:20 <chris> andi5: emacs? I just put i++ << 11 into the base point of my video ram. viola, new window.
17:23:48 <jsled> O'
17:25:04 <andi5> hey, do you think we can buy chris an i386 pc from the money spent to the project?
17:25:06 <jsled> Really, the only two points I care a lot about are -nut -npcs -nprs and -bbo
17:25:21 <jsled> s/two/three/
17:25:55 <jsled> And -i4 to a lesser extent. -i2 is fine; -i8 is not.
17:26:19 * hampton wonders if this is going to turn into a Monty Python skit.
17:26:29 <hampton> :-)
17:27:01 <jsled> The only 5 points I care about are fear, surprise, a ruthless efficiency, an almost fanatical devotion ...
17:27:23 <hampton> rotfl
17:31:24 <andi5> -nprs and -nut .... yes, please... -i2 would be nice (i4 ok, i8 for linux), .. have never used -bbo before, but ok.... -npcs is ok, but i would rather see spaces..... (i do not have many strong opinions on this)
17:31:52 <warlord> Can we add the magic emacs glue to each file to set the appropriate mode settings?
17:32:53 <jsled> -bbo is pretty good, as you can read the overall structure of the logic straight-away, versus having to pick the '&&' and '||'s out of the ends of lines, maybe visually masked by other line-lengths.
17:33:37 <jsled> I've never really understood the spaced () calls ... I guess, then, it's the same as if (...) and for (...)
17:33:50 <andi5> well, lines need not be longer than 80 characters :-P
17:34:15 * warlord likes if (a; b; c) , while (a), but func(x, y);
17:34:31 * hampton agrees
17:35:11 * chris too.
17:35:24 <warlord> and macros: FOO(a, b)
17:35:39 <warlord> And I've also grown to like:
17:35:44 <warlord> if (foo)
17:35:45 <warlord> {
17:35:46 <warlord> ...
17:35:47 <warlord> }
17:35:48 <warlord> else
17:35:49 <warlord> {
17:35:51 *** twunder has quit IRC
17:35:51 <warlord> ...
17:35:53 <warlord> }
17:36:01 <hampton> ick
17:36:34 <andi5> seems like this will be a non-trivial issue .... 2 vs 3 right now
17:36:44 <jsled> andi5: bottom of http://pastebin.ca/386850
17:37:12 <andi5> jsled: yes, thanks
17:37:43 <andi5> jsled: will -bbo break line 67?
17:38:42 <jsled> No, it won't.
17:39:34 <jsled> Actually, I take it back. `indent` seemed to do {char *bar} (imho, correct), but {SomeReallyLongThingy * foo}.
17:40:52 <jsled> warlord: re: macros, vs what?
17:45:31 <warlord> vs. FOO (a, b)
17:45:56 <jsled> Ahh.
17:50:36 <andi5> do we want rules for comments too?
17:51:26 <jsled> such as?
17:57:12 <andi5> like -sc (Put the ‘*’ character at the left of comments).... too bad indent cannot convert // to /* ... */
17:57:34 <jsled> Oh. Yeah, I do like that one, actually.
18:05:17 <andi5> my last Q: -cs, -ncs or none
18:06:09 <jsled> GNC_PLEASE_TO_BE_CASTING (foo) ?
18:06:16 <jsled> (for -cs?)
18:06:47 <hampton> I think -cs converts "(int)foo" to "(int) foo"
18:07:04 <jsled> Oh, you mean a "cast". ;)
18:07:12 <andi5> yes
18:07:59 <jsled> I like no-space; I don't care too much.
18:08:32 *** egsavage has joined #gnucash
18:35:49 <andi5> Q: do we still need the included copies of libgsf and goffice?
18:37:15 <jsled> I tend toward no. I think we'd discussed this not too long ago.
18:38:52 <andi5> hm... did i read that? ... i cannot remember ;-) ... /me is tired -)
18:39:11 *** ceplma_ has quit IRC
18:43:55 *** ceplma_ has joined #gnucash
18:50:08 <MrN> n8
18:50:13 <andi5> n8
18:50:19 *** andi5 has quit IRC
18:50:20 *** MrN has quit IRC
19:24:46 *** dbr has joined #gnucash
19:30:18 *** jpeac1 has joined #gnucash
19:30:24 *** jpeac1 has left #gnucash
19:34:56 *** ceplma_ has quit IRC
20:18:52 *** twunder has joined #gnucash
20:44:38 *** BlackBsd has quit IRC
20:44:44 *** twunder has quit IRC
20:45:43 *** BlackBsd304 has joined #GnuCash
21:16:45 *** cortana has joined #gnucash
21:30:04 <jsled> http://primates.ximian.com/~federico/docs/summer-of-code-mentoring-howto/index.html
22:31:43 *** BlackBsd304 has quit IRC
22:40:30 *** egsavage has quit IRC
22:52:02 <jsled> rhel4 ~=? fc5
22:52:26 <jsled> No ... fc3, I think.
22:54:29 <hampton> I think its fc3
23:32:45 <hampton> jsled: re -bbo. Doesn't say to force breaking before every && or || operator. Says that if you have to wrap the line containing one of these operators, then do it before the operator instead of after.
23:33:25 <jsled> Yes.
23:42:00 <jsled> And I think that's right. if (foo || bar) shouldn't be broken, of course. But if you need to line break, the operator should be aligned on the left, not dangling-right.
23:42:51 <hampton> You can end up with this type of indentation though:
23:43:10 <hampton> if (is_is_a_reasonable_value(i) && (q < 10 || q < 11)
23:43:15 <hampton> && (i < 10 || i < 11)) {
23:43:21 <hampton> // ...
23:43:27 <hampton> }
23:43:55 <hampton> i.e. indented noticed that it needed to break the line, but it only moved the last clause instead of putting each clause on a separate line
23:44:36 <jsled> Sure. Unfortunate, but we'll clean those up when we come across them.
23:44:56 <jsled> I don't expect indent to do most of the work.
23:45:05 <hampton> np. just wanted to mention it.
23:45:15 <jsled> cool
23:48:04 <jsled> Heh, cuddle. So, you're thinking ... -br -ce and -cdw ?
23:49:39 <hampton> Yup.
23:51:15 <jsled> Ah, /me receives mail...
23:51:16 <jsled> :)
23:53:22 <hampton> thought you were at the top of the mailman distribution list when you asked that question. I had just sent the email.
23:59:59 <jsled> No ... I think I'm fairly late, and then I only fetchmail every 5 minutes or so.