Log for on 23rd September 2012:
Times are UTC Toggle Colours
00:01:45  *** FLHerne has quit IRC
04:56:02  *** Eddi|zuHause has quit IRC
04:56:16  *** Eddi|zuHause has joined
07:38:34  *** Supercheese has quit IRC
08:12:46  *** Alberth has joined
08:12:46  *** ChanServ sets mode: +v Alberth
08:15:10  <Alberth> testing.. 1 2 3
08:15:18  <Alberth> \o/ it works!
08:33:20  <Terkhen> hi Alberth  :)
08:33:30  <Alberth> hi
08:36:38  *** Zuu has joined
08:36:38  *** ChanServ sets mode: +v Zuu
08:56:59  *** zooks has joined
10:25:18  *** FLHerne has joined
10:38:35  <Zuu> Hmm, actually the removal of Invalidate window when unpausing in misc_cmd.cpp was correct in the script pause patch. That code now causes the continue button to get disabled when you unpause the game. And now unpausing the game doesn't make the script to continue.
10:42:44  <Alberth> sounds like a fix patch
10:44:31  <Zuu>
10:51:19  <Yexo> fixing by removing code, I like that :)
10:55:38  <Alberth> patch seems ok to me
11:07:54  *** andythenorth has joined
11:07:54  *** ChanServ sets mode: +v andythenorth
12:58:04  *** frosch123 has joined
12:58:04  *** ChanServ sets mode: +v frosch123
12:58:51  <frosch123> wow, lot's of commits :)
12:59:28  <Alberth> :)
12:59:45  <frosch123> hai albert :)
13:01:52  <Alberth> I couldn't really find a nice way to combine the patches, so I decided to just commit them separately
13:09:53  <frosch123> is the right-aligness of the buttons in the group gui intentional?
13:17:56  <Alberth> at the bottom-right?  yes
13:21:43  <frosch123> <- making the non-group vehicle list behave consistent :)
13:22:55  <Alberth> I thought I tested that, but apparently not enough :p    patch is fine
13:29:35  <frosch123> should dorpsgek also put revision numbers into the topic here?
13:30:01  <Zuu> I think it would be useful to have a notice when there is a commit. Updating the topic is one way to do that.
13:30:01  <frosch123> or some other kind of notification?
13:30:09  <frosch123> maybe moving .notice to here
13:30:14  <Alberth> Zuu:  #openttd.notice
13:30:30  <Zuu> Eg. if you discuss a change here, it is useful to see if someone commits it.
13:30:36  <frosch123> i like some kind of separator in the discussion
13:30:49  <frosch123> when it is too late or so :p
13:30:52  <Alberth> Zuu: not really, it always happens after the discussion is finished
13:31:22  <frosch123> though esp. backport commits can be quite spammy
13:32:03  <Alberth> biiiig separator :D
13:32:45  <frosch123> @op
13:32:49  <frosch123> aw :(
13:32:59  <frosch123> @whoami
13:32:59  <DorpsGek> frosch123: frosch
13:33:07  <frosch123> i thought it works in all channels
13:33:52  <Alberth> @op frosch123
13:34:00  <Alberth> @op
13:34:11  <frosch123> dorpsgek is misconfigured :)
13:34:13  <frosch123> TrueBrain: ^^
13:34:32  <frosch123> @topic get 2
13:34:32  <DorpsGek> frosch123: Error: '1' is not a valid topic number.
13:34:41  <frosch123> who said 1 :p
13:35:15  <frosch123> @topic get -1
13:35:15  <DorpsGek> frosch123: OpenTTD Dev Channel | Logs: | Voice (talk-right) upon request; make sure you are registered to NickServ before asking
13:35:29  <Alberth> lol :)
13:36:17  <frosch123> is there something wrong with the separator char?
13:36:19  <frosch123> @topic get 1
13:36:19  <DorpsGek> frosch123: OpenTTD Dev Channel | Logs: | Voice (talk-right) upon request; make sure you are registered to NickServ before asking
13:36:45  <frosch123> @topic set 1 OpenTTD Dev Channel
13:36:45  *** DorpsGek changes topic to "OpenTTD Dev Channel"
13:36:53  <frosch123> @topic add Logs:
13:36:53  *** DorpsGek changes topic to "OpenTTD Dev Channel || Logs: (frosch123)"
13:37:09  <frosch123> @topic add Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking
13:37:09  *** DorpsGek changes topic to "OpenTTD Dev Channel || Logs: (frosch123) || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking (frosch123)"
13:37:24  <frosch123> what, it included my name ? :o
13:37:40  <Alberth> yep :)
13:37:51  <frosch123> @topic set 2 Logs:
13:37:51  *** DorpsGek changes topic to "OpenTTD Dev Channel || Logs: || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking (frosch123)"
13:38:02  <frosch123> @topic set 3  Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking
13:38:02  *** DorpsGek changes topic to "OpenTTD Dev Channel || Logs: || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
13:38:17  <frosch123> so, @topic works, @op does not :p
13:38:22  <Alberth> dorpsgek needs some love :)
13:38:48  <TrueBrain> stop abusing the bot
13:38:51  <TrueBrain> seriously ..
13:39:06  <TrueBrain> what should people think about us? :D
13:39:13  <TrueBrain> and no frosch, permission is per channel :)
13:39:53  <Alberth> not that kind of love :p
13:51:35  <Zuu> The AI Debug window have been missing hotkeys up to now. This patch adds hotkey hooks for most/all buttons and pre-defines some hotkeys:
13:54:04  <frosch123> the switch needs a ' ' after the keyword
13:54:56  <frosch123> doesn't the HEBR_CONFIRM usage conflict with the OSK gui?
13:56:02  <Zuu> In that case there is already a conflict for the sign list window.
13:56:10  <frosch123> :p
13:56:32  <Zuu> That window uses the HEBR_CONFIRM (Return key pressed) to select the top most sign  in the filtered list.
13:57:48  <Zuu> What type of conflict do you exactly fear of?
13:58:10  <frosch123> that it conitues if you hit the ok button
13:58:14  <frosch123> feels weird to me
13:59:12  <Zuu> maybe in this case it is better to leave out the usage of HEBR_CONFIRM as the usage is not clear here?
13:59:49  <frosch123> what's with the sign gui?
13:59:51  *** Eddi|zuHause has left
13:59:55  <frosch123> or does it not matter there?
14:00:15  <Zuu> The sign GUI also handle HEBR_CONFIRM.
14:00:27  <Zuu> in a similar switch construct
14:00:46  <Zuu> and as far as I know, it works fine with the OSK window.
14:02:19  <frosch123> yeah, sign list does not scroll when hitting ok
14:02:29  <Zuu> I'll remove the HEBR_CONFIRM usage in my patch. It doesn't break OSK window as far as I know (and whom will debug a script on a cell phone without keyboard??)
14:03:27  <Zuu> But perhaps it is a bit unexpected behaviour that hitting return in the break string edit box continues the script if it is paused?
14:03:41  <frosch123> just do it the same as the sign list
14:03:47  <frosch123> or remove it from the sign list as well
14:03:52  <frosch123> but, the signlist works for some reason
14:04:13  <Zuu> In the sign list it is there to allow keyboard-only operation of the sign list
14:04:34  <Zuu> eg to first filter for the sign you look for and then hit enter to go to a certain point.
14:04:38  <frosch123> but yes, the enter should only unpause, if the hotkey is enter
14:05:17  <Zuu> That is true when the edit box is unfocused.
14:06:08  <Zuu> When an edit box has global focus, it grabs focus. Apparently some global CTRL+<char> works still.
14:09:30  <Zuu> Put it like this. The reason why I use HEBR_CONFIRM to continue, is not because I have choosen RETURN as the default hotkey and that HEBR_CONFIRM is the consequence of hitting return. Instead the logic is that I connect the CONFIRM action of a text box to continuing a script. This shouldn't be subject to checking with the hotkey definition.
14:10:14  <frosch123> ok, i am just confused why it does not work with the osk
14:11:38  <Zuu> For OSK, there would need to be [cancel], [close] and [confirm] instead of just [cancel] and [ok] if it should support all keyboard operations of an edit box.
14:12:44  <Zuu> However if you use the OSK, you can click on the GUI button for the confirm action without having to worry about first unfocusing the text box.
14:13:52  <Zuu> In the sign window, hitting cancel clears the edit box and there is no hotkey to go to the first sign in the list. So there it is more needed to use the HEBR_CONFIRM to provide a keyboard equivalent of clicking on a sign in the list.
14:19:25  <Zuu> Given the discussion regarding if it is logical or not to continue a script if you hit the confirm button (return) in the edit box or not for the AI debug window, I leave it out for now and could always add it later if it seems to be a missing link.
14:20:32  *** FLHerne has quit IRC
14:20:53  *** FLHerne has joined
14:21:57  <Zuu> New patch with "switch (" and HEBR_CONFRIM removed:
14:25:29  <frosch123> i would expect it needs some "default: break;" to avoid warnings from some compilers
14:25:52  <frosch123> (warnings about missing a few enum values in a switch)
14:26:25  <frosch123> +Window* ShowAIDebugWindow(CompanyID show_company = INVALID_COMPANY); <- s/* / */
14:26:41  <frosch123> looks fine otherwise :)
14:29:49  <Zuu> Thanks, I've now fixed those things (locally)
15:03:23  <TrueBrain> @channel capability add frosch123 op
15:03:23  <DorpsGek> TrueBrain: The operation succeeded.
15:03:27  <TrueBrain> frosch: try it now :)
15:03:32  <TrueBrain> (sorry, bit late, I had stuff to do :))
15:28:42  <frosch123> @op
15:28:42  *** DorpsGek sets mode: +o frosch123
15:28:43  <frosch123> @deop
15:28:43  *** DorpsGek sets mode: -o frosch123
15:28:57  <frosch123> i thought that @channel stuff was nickserv stuff
15:28:59  <frosch123> not bot stuff
15:29:49  <TrueBrain> you thought wrong :)
15:30:30  <TrueBrain> you can, ofc, query chanserv and also get ops :)
15:30:55  <frosch123> nah, i prefer the @op and @kick stuff
15:31:03  <TrueBrain> me too
15:31:04  <frosch123> i don't want to do the dirty work myself :p
15:31:29  <frosch123> but with voiceonly there are likely hardly uses for kick
15:32:39  <TrueBrain> lets hope so :D
16:36:45  <Zuu> @get 1
16:36:45  <DorpsGek> Zuu: OpenTTD Dev Channel
16:37:05  <Zuu> In my client it says "or OpenTTD Dev Channel"
16:37:23  <Zuu> But maybe it's my silly client that translates || to "or"?
16:37:38  <frosch123> it does not start with a ||
16:39:20  <Zuu> Hmm, when hovering it says the right topic. And apparently it let me to edit the topic shown otherwise so I might have edited it before by mistake then.
16:58:13  *** andythenorth has quit IRC
17:01:03  *** LordAro has joined
17:05:37  *** andythenorth has joined
17:05:37  *** ChanServ sets mode: +v andythenorth
17:21:52  *** andythenorth has quit IRC
17:22:11  *** andythenorth has joined
17:22:11  *** ChanServ sets mode: +v andythenorth
18:57:25  <Alberth> \o/ industry and industry tile scope resolvers actually compile!
18:59:52  <frosch123> :o
19:00:16  <frosch123> night though
19:00:23  <frosch123> need to get up way too early tomorrow :s
19:00:30  *** frosch123 has quit IRC
19:03:44  <Zuu> Alberth: In the new scenario format?
19:04:18  <Alberth> lol, no, in the newgrf_* source code files  :)
19:04:48  <Alberth> I am trying to do code changes there to improve the structure
19:05:56  <Zuu> Ok. The closest I have come to NewGRFs is when extending the NoAI API to work with rotaded airports.
19:06:28  <Yexo> is that supported?
19:06:32  <Alberth> well, I don't know much of it either :)
19:07:45  <Zuu> Yexo: I think I stopped with that work due to the uncertainty on where the airport work would go. Parts of it may be possible to commit already but IIRC it does break the current API a bit to create a better structure.
19:08:02  <Yexo> ah, ok :)
19:08:06  <Zuu> So it might be wise to only break it once.
19:09:26  *** LordAro has left
19:09:35  *** LordAro has joined
19:09:35  *** ChanServ sets mode: +v LordAro
19:09:45  <LordAro> that's better :)
19:10:05  <Terkhen> use your rights wisely ;)
19:10:31  <LordAro> dsfASEFHf.
19:10:33  <LordAro> :P
19:10:58  <Zuu> Duno if it is better to break it just after or before a new stable. If it is done just before a new stable, AIs that update to the new API will not need to maintain two versions for more than 3-4 months. But it will on the other hand give a shorter window for AIs to update before the stable is out.
19:11:31  <Yexo> as long as we have a compatibility script to fix things there shouldn't be a problem, right?
19:11:50  <Yexo> the 'nightly' api has always been considered unstable
19:12:19  <Yexo> if you write an AI based on the api in a nightly you have to consider it might get broken without warning
19:12:44  <LordAro> i did that... and it did :L
19:13:00  <Zuu> Yexo: True, we got the the compatibility layer.
19:13:03  <LordAro> silly HasNext() -> !IsEnd() changes
19:13:48  <Zuu> As long as people specify a correct API version it works.
19:15:16  *** andythenorth has quit IRC
20:04:39  <Alberth> good night
20:06:18  *** Alberth has left
20:29:19  *** Supercheese has joined
20:38:18  *** LordAro has quit IRC
22:33:22  *** Zuu has quit IRC
22:38:15  *** FLHerne has left

Powered by YARRSTE version: svn-trunk