Log for #openttdcoop.devzone on 22nd January 2012:
Times are UTC Toggle Colours
00:11:42  <Brot6> OpenGFX - Bug #3563: Road depot variants for arctic and tropic (athanasios) @
00:12:03  *** frosch123 has quit IRC
00:54:16  <Brot6> NewGRF Meta Language - Feature #1348: Bridge support (Hirundo) @
04:05:11  *** MinchinWeb has joined #openttdcoop.devzone
04:05:53  <MinchinWeb> how do I get nmlc to run (on Windows)?
04:06:32  <MinchinWeb> I need getting an error "Error: only a single nml file can be read per run"
04:06:43  <MinchinWeb> I kept getting that error, rather
04:14:29  <MinchinWeb> or nmlc seems to go into a infinite loop (it never finishes)
04:28:09  <MinchinWeb> (I'm running r1788 on Windows...)
05:11:24  *** MinchinWeb has quit IRC
08:10:33  *** andythenorth has joined #openttdcoop.devzone
08:40:15  *** andythenorth is now known as Guest146
08:40:16  *** andythenorth_ has joined #openttdcoop.devzone
08:40:16  *** andythenorth_ is now known as andythenorth
08:43:59  <Brot6> DACH Trains - Feature #3525: ÖBB City Shuttle double deck control car (officercrockey) @
08:45:45  <Brot6> DACH Trains - Feature #3535: ÖBB City Shuttle control car (officercrockey) @
09:10:24  <Brot6> DACH Trains - Feature #3535: ÖBB City Shuttle control car (Yoshi) @
09:11:41  *** JVassie has joined #openttdcoop.devzone
09:12:07  *** Zuu has joined #openttdcoop.devzone
09:20:24  *** ODM has joined #openttdcoop.devzone
09:23:33  <Brot6> DACH Trains - Revision 41:9a38071c2501: added SBB IC2000 (officercrockey) @
09:23:33  <Brot6> DACH Trains - Feature #3532 (Closed): SBB IC2000 coaches (officercrockey) @
09:24:44  <Brot6> DACH Trains - Feature #3535: ÖBB City Shuttle control car (officercrockey) @
09:34:15  <Brot6> DACH Trains - Revision 42:c194e92cb87d: added SBB IC 2000 control car (officercrockey) @
09:34:15  <Brot6> DACH Trains - Feature #3533 (Closed): SBB IC 2000 control car (officercrockey) @
10:01:52  <Brot6> DACH Trains - Revision 43:531ebc1261d6: push-pull for IC 2000 (with Re 460) (officercrockey) @
10:02:20  <Brot6> DACH Trains - Feature #3550 (Closed): push-pull for SBB IC2000 (officercrockey) @
10:07:07  <Brot6> DACH Trains - Revision 44:f6b12a1b4922: updated additional text of TRAXX (officercrockey) @
10:25:30  *** ODM has quit IRC
10:36:06  <Brot6> OpenGFX BigGUI - Feature Request #3568 (New): x1.5 sized GUI (Yoshi) @
10:58:54  <Brot6> BANDIT - Feature #3567 (Closed): Set fixed IDs (andythenorth) @
10:58:54  <Brot6> BANDIT - Revision 123:22a5eb7a4945: Change: set numeric IDs for vehicles (andythenorth) @
10:58:54  <Brot6> BANDIT - Feature #3567 (Closed): Set fixed IDs (andythenorth) @
11:00:10  <Brot6> BANDIT - Feature #3400 (Closed): Consider templating spritegroups / spritesets (andythenorth) @
11:00:46  <Brot6> BANDIT - Feature #3569 (New): CMS template needs to handle truck_type (andythenorth) @
11:03:28  <Brot6> Dutch Road Furniture - Feature Request #3570 (New): highway and speed signs in combination with a... (foobar) @
11:07:36  <Brot6> NewGRF Meta Language - Feature #3571 (New): default cargo (Hirundo) @
11:43:17  *** JVassie has quit IRC
11:43:38  *** JVassie has joined #openttdcoop.devzone
11:48:20  <Brot6> DACH Trains - Revision 45:4b5ee9d9cb7b: added BLS RABe 535 (officercrockey) @
11:55:05  *** JVassie_ has joined #openttdcoop.devzone
11:58:08  <Brot6> DACH Trains - Revision 46:89813f943798: fixed BLS 535 offset (officercrockey) @
12:00:46  *** frosch123 has joined #openttdcoop.devzone
12:00:52  <Brot6> BANDIT - Revision 124:009c5a3624e4: Change: set intro dates (andythenorth) @
12:02:00  *** JVassie has quit IRC
12:03:52  *** JVassie_ has quit IRC
12:04:17  <Brot6> NewGRF Meta Language - Feature #1348: Bridge support (foobar) @
12:04:55  <Brot6> OpenGFX BigGUI - Revision 3:7cde98b9618e: Fix: Order of rail tool icons was partially erronous (planetmaker) @
12:04:55  <Brot6> OpenGFX BigGUI - Revision 4:92c0618f470e: Add: Parameter and sprites to allow choosing 1.5x zoom ... (planetmaker) @
12:12:02  <Brot6> Dutch Road Furniture - Feature Request #3572 (New): median with light posts (foobar) @
12:25:05  <Brot6> NewGRF Meta Language - Feature #1348: Bridge support (Hirundo) @
12:25:10  <Brot6> OpenGFX BigGUI - Revision 5:ac5678b0167b: Fix: Glyphs must have palette index 0 and 1 only. Colou... (planetmaker) @
12:53:26  <Brot6> Dutch Road Furniture - Feature Request #3079 (Rejected): Medians with no central tile (foobar) @
13:13:26  <Brot6> Dutch Road Furniture - Bug #3573 (New): overbuildable object are also overbuilt by town owned hou... (foobar) @
13:13:26  <Brot6> Dutch Road Furniture - Bug #3574 (New): shoulder ends don't recognize all relevant other objects ... (foobar) @
13:14:33  <Brot6> BANDIT - Revision 125:1044ee84c752: Change: set various vehicle properties (andythenorth) @
13:16:29  *** andythenorth has quit IRC
13:46:21  *** andythenorth has joined #openttdcoop.devzone
13:51:57  *** ODM has joined #openttdcoop.devzone
14:06:26  <Brot6> OpenGFX+ Landscape - Feature Request #3575 (New): Wind turbine orientations (PaulC) @
14:09:23  <Brot6> OpenGFX+ Landscape - Feature Request #3575: Wind turbine orientations (PaulC) @
14:19:34  <Brot6> Tutorial - Revision 21:43bedcadf035: Updates to english.txt by Pingaware (Zuu) @
14:19:34  <Brot6> Tutorial - Revision 22:eee49e9af527: Fix: wrong comment (Zuu) @
14:29:03  * Hirundo installs Debian VM
14:29:44  <planetmaker> do we get back a dev'ing Hirundo? :-)
14:35:14  <Hirundo> perhaps ... ;-)
14:38:57  <Ammler> why else should he announce it here? :-P
14:39:14  <Brot6> DACH Trains - Feature #3521 (Closed): BLS Rabe 535 Lötschberger (officercrockey) @
14:46:26  <Hirundo> I'm not sure I'll have much time to develop large features, but bug fixes and small features should work
15:16:17  <Ammler> planetmaker: the mail is a bit confusing, but there is a extra ttdp thread to discuss such things
15:16:58  <Ammler>
15:16:59  <Webster> Title: Transport Tycoon Forums View topic - [TTDP] OpenGFX 0.4 for TTDPatch (at
15:22:14  <Brot6> BANDIT - Revision 126:6be28b5b605d: Fix: correct capacity refit for fifth wheel trucks (andythenorth) @
15:35:41  <Brot6> Tutorial - Revision 23:149def4d7467: Add: ConditionalStep (Zuu) @
15:35:41  <Brot6> Tutorial - Revision 24:39fd1e9c791f: Add: If quick_goto is off, tell the user to click on the GoT... (Zuu) @
15:48:55  <frosch123> <- better?
15:49:08  <frosch123> ( )
15:49:24  <andythenorth> frosch123: less spaghetti
15:49:37  <andythenorth> you don't like all that spaghetti? :P
15:49:44  <andythenorth> it makes creating vehicles such fun too :P
15:49:54  <frosch123> he, i could put them in one graph
15:50:25  <frosch123> i could also add grfv8 :p
15:51:11  <frosch123> replaced cargoslot with cargotype, so it applies to both v7 and v8 :p
15:51:26  <andythenorth> I have no idea if it's accurate :D
15:55:17  <frosch123> why is eddi not in this channel, it is also meant for cets
16:02:06  *** JVassie_ has joined #openttdcoop.devzone
16:03:34  <Brot6> Tutorial - Revision 25:eaf08f388fd0: Version 5 (Zuu) @
16:15:25  <Brot6> BANDIT - Revision 127:f34ca7c6d7ef: Change: split capacity of fifth wheel trucks over truck and f... (andythenorth) @
16:16:11  <Brot6> BANDIT - Feature #3569 (Closed): CMS template needs to handle truck_type (andythenorth) @
16:16:41  <Brot6> BANDIT - Feature #3576 (New): Don't offer subtype refits for fifth wheel trucks with 1 trailer (andythenorth) @
16:20:17  <Brot6> BANDIT - Feature #3576 (Closed): Don't offer subtype refits for fifth wheel trucks with 1 trailer (andythenorth) @
16:20:18  <Brot6> BANDIT - Revision 128:53feb65017dd: Change: don't offer subtype refits for fifth wheel trucks wit... (andythenorth) @
16:20:18  <Brot6> BANDIT - Feature #3576 (Closed): Don't offer subtype refits for fifth wheel trucks with 1 trailer (andythenorth) @
16:29:36  <Hirundo> frosch123: looks sane
16:30:17  <frosch123> i hope it does not break everything :p
16:31:27  <Hirundo> In NML, it would be preferable if that flag always applied w/o user intervention
16:31:42  <Hirundo> Since users tend to mess up such things, then complain when it doesn't work
16:31:52  <frosch123> for vehicles carrying a single cargo, it is more useful if it is not set
16:32:20  <frosch123> e.g. armored vehicles
16:32:34  <frosch123> that's why i did not include it in grfv8 :)
16:32:49  <frosch123> nml might set it, if a vehicle is refittable
16:32:59  <frosch123> but it would break compatiblity
16:33:53  <Hirundo> currently I have to tell users to always set the 'cargo_capacity' callback to avoid capacity multiplier mess
16:34:30  * andythenorth ponders
16:34:35  <andythenorth> should I be setting the flag?
16:34:49  <Hirundo> multipliers are 2 for mail/goods, 4 for pax and 1 for others, right?
16:35:02  <Hirundo> or 2 for gold also?
16:35:08  <frosch123> yes, the definition is kind of "capacity is always in tons of coal"
16:35:23  <frosch123> 2 for mail/goods, 4 for pax
16:35:46  <frosch123> i would imagine that industry sets would use other multipliers for other cargos
16:35:53  <frosch123> so i want to fix it on "coal" :)
16:35:59  <andythenorth> multiplier for mail is 2?
16:36:11  <frosch123> depends on the vehicle type :p
16:36:16  * andythenorth thinks not for RVs :P
16:36:17  <Hirundo> e.g. 1/4 or so on FIRS vehicles, right?
16:36:27  <andythenorth> 18/13
16:36:47  <andythenorth> 1.4 for RVs
16:36:53  <frosch123> andythenorth: a rv can transport 2 bags of mail if it can transport 1 ton of coal
16:36:58  <andythenorth> bizarre
16:37:03  <andythenorth> frosch123: that's not true :)
16:37:13  <frosch123> then you used cb 15
16:37:23  <andythenorth> not intentionally :P
16:37:27  <andythenorth> cb36 afaik
16:37:36  <andythenorth> whatever nml uses
16:38:05  <andythenorth> hmm
16:38:07  <Hirundo> From an NML perspective, I'd like if a user could say "apply capacity multipliers" and "don't apply capacity multipliers"
16:38:13  <andythenorth> varies for articulated / non articulated
16:38:20  <Hirundo> With proposed patch, I can do a), but b) not quite
16:38:41  <frosch123> b) would be use cb15 unconditionally
16:38:48  * andythenorth wonders if this is a bug in BANDIT
16:38:56  <andythenorth> but I do nothing with cargo types
16:38:59  <andythenorth> maybe that's the bug
16:39:00  <Hirundo> I'd like to be able to do that w/o callbacks
16:39:16  <frosch123> imo that makes no sense :p
16:39:31  <frosch123> a vehicle is never refittable to the same amount for all cargos
16:39:50  <andythenorth> eddi had a proposal for this
16:39:53  <andythenorth> but he's not here :P
16:40:24  <frosch123> the intention is, that control freaks can use cb15 to set specific capacities for specific cargos, while everyone else can just set the capacity for coal and not care about cargos
16:41:02  <frosch123> i.e. capacity is then controlled by the cargo definition
16:41:04  <andythenorth> frosch123: I should be setting this new flag then?
16:41:17  <andythenorth> currently a lead part of RV consist has mail = 2 x coal
16:41:24  <andythenorth> but trailing parts have mail = 1x coal
16:42:41  <Hirundo> frosch123: Let's say I have basic a vehicle refittable to gold and goods and I'd like to have an equal capacity for both
16:43:28  <Hirundo> IMO that should be possible without using callbacks
16:43:58  *** JVassie has joined #openttdcoop.devzone
16:45:07  <frosch123> if you set the refittable cargos so specific, you can as well use specific callbacks
16:45:40  <frosch123> if you are control freak enough to set two explicit cargos instead of cargos, you must control capacities as well
16:45:42  <Hirundo> or gold and mail, whatever, that'd be bitmask(CC_ARMOURED, CC_MAIL)
16:46:11  <frosch123> if you use cargo classes, you already claim that the cargo defining grf shall define the refiittablity
16:46:19  <frosch123> so, the cargo grf shall also define the capacity
16:47:39  <Hirundo> That's annoying for stuff like pax
16:47:46  <frosch123> i do not understand the intention of giving the same capacity to mail and gold. it seems like the author would not be happy with the multipliers of the default cargos. but then he shall write a cargo grf to change the cargos
16:48:01  <Hirundo> So for a bus, you'd not set that flag
16:48:20  <frosch123> yes, not for a bus
16:48:44  <frosch123> so, the pax multiplier applies to the default capacity
16:48:51  <frosch123> but tourists would still be affected
16:49:15  <Hirundo> Tourists are a bad invention... did I mention that yet ;-)
16:49:58  <andythenorth> capacity should be set by the cargo imho
16:50:01  <frosch123> they are express, so meant for maglev or air cargo :p
16:50:16  <andythenorth> if a vehicle author wants to vary capacity, use cb36 checking cargo label from ctt
16:50:27  *** JVassie_ has quit IRC
16:50:28  <frosch123> andythenorth: s/cb36/cb15/
16:50:38  <andythenorth> could do maths inline :P
16:50:41  <andythenorth> you should see my code :P
16:50:57  <andythenorth> anyway, the only thing that bothers me is that the multipliers are pot luck :P
16:51:30  <andythenorth> also...I should fix the mail problem with trailing parts in BANDIT, or wait for some upstream fix?
16:51:53  <Hirundo> Currently the 'cargo_capacity' callback in NML is used for both cb36 and cb 15
16:52:08  <Hirundo> So if a user uses the callback, he always gets the exact capacity
16:52:09  <frosch123> andythenorth: i am not aware of any code distinguishing front/trailing parts :p
16:52:25  <Hirundo> If he doesn't capacity multipliers apply
16:52:27  <frosch123> Hirundo: both? what do you mean?
16:53:18  <Hirundo> The callback is internally mapped to both cb36->capacity and cb15
16:53:29  <Hirundo> so both callbacks end up calling the same switch chain
16:55:56  <Hirundo> That'd have to be changed, else your proposal makes no difference (except for ships)
16:56:44  <frosch123> ideally you would be able to return either "15" or "15 * multiplier"
16:57:02  <frosch123> i.e. something like adding a unit to the capacity
16:59:52  <Hirundo> perhaps add some variable that checks (cb == 15)
17:00:01  <Hirundo> or split the callback back into two callbacks
17:01:55  <Hirundo> Still, I'd like to suggest an additional flag 'NO_CAPACITY_MULTIPLIERS' that disables all usage of multipliers w/o resorting to cb 15
17:02:30  <Hirundo> <away for 10 minutes or so
17:02:56  <andythenorth> frosch123: front part uses action o prop, trailing part uses [nml] capacity cb
17:03:11  <andythenorth> I could switch front part to use capacity cb also
17:04:29  <frosch123> i still thing that using cb15 without multipliers should be a rare case
17:04:40  <frosch123> vehicles no using capacity multipleirs is mo the wrong way
17:04:46  <frosch123> which should not be supported
17:06:35  <frosch123> so nml should probably also add two callbacks, like "general_capacity" and "explicit_cargo_capacity"
17:21:04  <Brot6> ogfx-trains: update from r292 to r295 done -
17:22:37  <Brot6> bandit: update from r119 to r128 done (248 warnings) -
17:24:37  <Hirundo> Important consequence of flag bit 5 would be, that cap. multipliers are based on coal instead of default cargo
17:24:42  <Brot6> OpenGFX BigGUI - Bug #3577 (New): DevZone compile failed (compiler) @
17:25:03  <Hirundo> IMO a good thing, as default cargo is messy enough already w/o other things depending on it
17:25:17  <Hirundo> PITA to document, though
17:25:31  <Brot6> dach: update from r39 to r46 done -
17:28:44  <michi_cc> frosch123: Is it even possible to specify the capacity multiplier for custom cargoes? Source very much looks like hardcoded to me.
17:29:24  <frosch123> michi_cc: that's 50_cargoprop.diff
17:29:59  <michi_cc> Ah, didn't see that.
17:30:16  <frosch123> (10 - 40 were already committed months ago btw)
17:31:05  <Brot6> OpenGFX BigGUI - Revision 6:ab2479cd83bc: Change: Notify user that zoom level 1x has no effect (planetmaker) @
17:31:06  <Brot6> OpenGFX BigGUI - Revision 7:23ded9b21794: Fix: Also add the source file for the 1.5x zoom (planetmaker) @
17:31:06  <Brot6> OpenGFX BigGUI - Revision 8:cc51998127ed: Fix: Action colours in the wrong places (planetmaker) @
17:31:07  <Brot6> OpenGFX BigGUI - Revision 9:af337512cea7: Add: 1.5x and 2x zoom for the eight newest OpenTTD GUI ... (planetmaker) @
17:31:11  <Brot6> OpenGFX BigGUI - Bug #3577 (Closed): DevZone compile failed (compiler) @
17:31:14  <Brot6> OpenGFX BigGUI - Bug #3577: DevZone compile failed (planetmaker) @
17:32:02  <Brot6> OpenGFX BigGUI - Bug #3577 (Closed): DevZone compile failed (planetmaker) @
17:33:30  <Brot6> ogfx-biggui: update from r2 to r9 done -
17:51:47  <Hirundo> OK, I've got my debian VM up and running and NML installed
17:52:25  <Hirundo> Now when I run regression test, it works as root, but I get warnings about .hg/hgrc being from an untrusted user (myself)
17:53:20  <Hirundo> when I run regression as a normal user, I get an error in python regarding missing permissions
17:53:44  <Hirundo> it can't create (parser output)
17:53:55  <Hirundo> How do I fix that?
17:54:17  <andythenorth> Hirundo: I had a similar issue when I installed nml
17:54:28  <andythenorth> for some reason, I recall having to fiddle a shebang
17:54:34  <andythenorth> but that might be misremember
17:56:36  <Ammler> Hirundo: never use root
17:57:10  <Hirundo> andythenorth: line 1 of nmlc is #! usr/bin/env python, did you change that?
17:57:30  <andythenorth> don't think so
17:57:33  <andythenorth> let's see
17:57:44  <Ammler> why should he, it would be wrong
17:58:19  <andythenorth> the shebang is set to specific python version, but that's a local thing to me
17:58:23  <Ammler> isn't there a "/"?
17:58:36  <Ammler> #!/usr/bin...
17:58:38  <andythenorth> yes mine has a leading /
17:59:34  <Hirundo> Yes there is, I mistyped
18:00:07  <Ammler> Hirundo: you should checkout and build (not needed) as user
18:00:15  <Yexo> Hirundo: in which directory did you install nml?
18:00:19  <Yexo> did you install it as root?
18:00:24  <Hirundo> ~/nml
18:00:47  <Hirundo> yes I did, NMLTutorial told me :-)
18:02:01  <planetmaker> Hirundo: I have a symlink from ~/ottd/grfdev/nmlc to /usr/local/bin/nmlc
18:02:08  <Yexo> what is the exact error message you get?
18:02:28  <Hirundo> I redownloaded as user, now it works
18:02:58  <Hirundo> at least, regression tests 1-14 have compiled so far
18:03:03  <Yexo> that's basically what the tutorial says, although it's not very clear
18:03:20  <andythenorth> took me ages to install nml :)
18:03:29  <Hirundo> "Also this time you need the terminal as superuser"
18:03:29  <andythenorth> worth it though
18:03:44  <Ammler> planetmaker: on linux, ~/bin is also part of $PATH
18:03:49  <Ammler> so no need to use root
18:03:54  <Yexo> Then change directory to the (extracted) NML source <- you need to have the nml source extracted before you change to superuser
18:04:11  <Brot6> OpenGFX BigGUI - Revision 10:f683f0092597: Add: Window layout change and debugging buttons for 1.... (planetmaker) @
18:04:22  <Yexo> Ammler: not for me
18:05:49  <Ammler> well, then you can still add it ot ~/.bashrc or whatever
18:07:30  *** Hirundo_vm has joined #openttdcoop.devzone
18:08:01  <Hirundo_vm> this should save some real machine<>VM switching :-)
18:11:25  * andythenorth tests something
18:13:18  <andythenorth> do any of the nml devs try the 'clean sheet of paper' nml install occasionally?
18:13:29  <andythenorth> it's the most important thing for making nml easy to use now
18:14:17  <Yexo> I've used it not too long ago: no trouble at all
18:15:28  <andythenorth> I don't have a vm to test it
18:15:41  <andythenorth> and I don't much fancy blitzing whatever it is that causes mine to work :P
18:15:47  <planetmaker> I guess we now have all GUI icons in 2x and 1.5x zoom levels :-)
18:16:03  <andythenorth> Ammler: you should have a 'fresh state' vm build nml every now and then ;)
18:16:58  <Yexo> andythenorth: Ammler already builds suse packages from nml, and the stable release will be packaged in debian too
18:17:13  <Yexo> that will make it really easy to use on various linux platforms
18:17:21  <Yexo> for windows we already have the zip with an exe inside
18:17:59  <andythenorth> hmm
18:18:01  <Ammler> the build environment is every day made from clean env
18:18:18  <andythenorth> I just tried a buildout on a 'clean' OS X user
18:18:22  <andythenorth> not the same as a vm though
18:18:32  <andythenorth> it worked first time, and all tests passed :o
18:18:52  <Yexo> great :)
18:18:52  <andythenorth> but I'm never sure if that's because my system already has awkward deps like PIL
18:18:58  <Yexo> the buildout part was actually untested :p
18:19:01  <andythenorth> afaik buildout is supposed to be a sandbox, but I'm not sure
18:19:09  <Yexo> it probably is, you'll always need to install those dependencies first
18:19:18  <andythenorth> nah, buildout will do it for you
18:19:25  <Yexo> oh, right
18:19:30  <andythenorth> with a set of eggs local to the dir
18:19:37  <andythenorth> so you don't mess with the rest of your system
18:19:45  <Ammler> imo, dependency of buildout is not really nice
18:19:45  <andythenorth> similar idea to virtual-env
18:20:00  <andythenorth> Ammler: it's only not nice because none of us understand it :P
18:20:05  <Ammler> since it could be done with only
18:20:12  <andythenorth> good luck with that :)
18:20:21  <Ammler> well, you have 2 dependencies
18:20:30  <Ammler> how much luck would that need?
18:20:43  <Ammler> but now, alone doesn't work
18:20:58  <Ammler> and buildout is too complicated for everyone
18:21:02  <andythenorth> +1
18:21:15  <andythenorth> buildout is complicated for maintainers, of which we have none
18:21:19  <andythenorth> for users it's easy
18:21:31  <Ammler> it is useless for packagers
18:21:36  <andythenorth> +1
18:21:40  <Ammler> but a good would be nice
18:21:53  <Yexo> just doesn't install dependencies, but that's something a package manager should do anyway
18:22:20  <andythenorth> will also create a virtual-env?
18:22:22  <Ammler> but could define require
18:22:34  <Ammler> then pip or easy_install would do it
18:22:57  <Yexo> I'm not too familiar with either of those, and I'm not going to spent time figuring it out
18:23:21  <andythenorth> will pip also run the tests?
18:23:28  <Ammler> e.g. there are scripts which converts scripts to rpm specs
18:23:37  <Ammler> which are useless for nml
18:24:34  <andythenorth> pip seems well liked
18:24:45  <andythenorth> it appears to be able to create virtual-env
18:25:13  <andythenorth> here's an interesting blog post about pip vs. other others like buildout etc
18:25:14  <Webster> Title: Why I like pip (at
18:25:20  <Hirundo_vm> hmm... It seems if I modify nml .py files, the updated files are not used
18:25:32  <andythenorth> make?
18:25:36  <Ammler> pip install nml pil ply should already work
18:25:52  <andythenorth> Ammler: you don't want it to go into a virtual-env?
18:26:11  <andythenorth> most of the python world I know either uses virtual-env or buildout to isolate projects
18:26:14  <Ammler> as linux user, you usually do not need
18:26:16  <andythenorth> otherwise madness happens
18:26:29  <Ammler> as you have binaries available for your distro
18:26:58  <Ammler> but it might make sense for people like you who run different versions
18:27:21  <Ammler> no, clue if such things works on windows at all
18:27:31  <andythenorth> it's pretty common to isolate
18:27:42  <andythenorth> you might need to try out a development version of a package for example
18:27:48  <Yexo> Hirundo_vm: if you installed nml, indeed, that's what happens
18:27:48  <Yexo> the installed files are used
18:27:48  <Yexo> deinstall, create a simlink from a directory in your path to nmlc and it should work
18:27:52  <Ammler> at least windows isn't able to create hard links so it would need to waste space for every fresh environment
18:28:09  <Yexo> or put NMLC=/path/to/nmlc in Makefile.local for the makefile-based newgrf projects
18:28:38  <planetmaker> that's much more tedious, though ;-)
18:28:41  <andythenorth> Ammler: out of interest, do you have a clean vm you could test the buildout on
18:28:46  <andythenorth> with only python 2.6 installed
18:28:48  <andythenorth> no deps
18:29:00  <Ammler> which os?
18:29:01  <Hirundo_vm> where are installed files put?
18:29:24  <Yexo> I'd guess something like /usr/lib/python/sites-packages/
18:29:42  <Ammler> /usr/local
18:29:53  <andythenorth> Ammler: any OS ;)
18:29:54  <Ammler> else you have already screwed your system :-)
18:30:17  <Yexo> hence the "something like" :p
18:32:06  <Ammler> andythenorth: the devzone does everytime setup a clean environemnt to install nml
18:32:37  <Ammler> you can check the logs, how it is setup, which packages are installed etc.
18:33:07  <Ammler> based on the spec in .devzone/build
18:38:50  <andythenorth> Ammler: how does that build?
18:39:00  <andythenorth> you've packaged it?
18:39:59  <Brot6> OpenGFX BigGUI - Revision 11:575d3bcd50c9: Fix: one row of pixels was missing in the 2x bug icon (planetmaker) @
18:44:14  <andythenorth> hmm
18:44:20  <andythenorth> buildout fails on my wife's laptop
18:46:55  <Yexo> what error messages do you get?
18:47:23  <andythenorth> bootstrap needs a newer setuptools
18:47:44  <andythenorth> if I easy_install -U setuptools, I don't have sufficient permissions
18:47:50  <andythenorth> which is a dead end for most users
18:47:55  <andythenorth> :/
18:48:22  <andythenorth> installing an app shouldn't require use of sudo
18:48:28  <andythenorth> and many users don't know how to sudo
18:50:14  <andythenorth> for example...if you don't know your wife's password, in this case :P
18:50:41  <Yexo> but that's quite unrelated to nml
18:51:08  <andythenorth> indeed
18:51:22  <andythenorth> but it's an out-of-the-box issue for new users
18:51:56  <Yexo> majority of users will use windows, and they have a zip with an exe inside that "just works"
18:52:58  <andythenorth> \o/
18:56:16  <Brot6> NewGRF Meta Language - Revision 1790:5ccc2602ac20: Fix #3450: Add NML output for replacenew-block (Hirundo) @
18:56:16  <Brot6> NewGRF Meta Language - Bug #3450 (Closed): nml output for replacenew (Hirundo) @
18:56:16  <Brot6> NewGRF Meta Language - Revision 1791:42938761281f: Add: regression test for replacenew-block (Hirundo) @
18:58:26  <Hirundo_vm> Last commit was 26 aug, acc. to log
19:06:22  <planetmaker> ogfx+biggui is a big regression test for that ;-)
19:09:37  <Hirundo> I used the one-way roads, that's only 6 sprites
19:12:10  <andythenorth> ok
19:12:12  <Brot6> Tutorial - Revision 26:149f7b72071c: Add: parameter for if script should allow loading older vers... (Zuu) @
19:12:12  <Brot6> Tutorial - Revision 27:a853712b7232: Add: Script to update GS version reference in the scenario t... (Zuu) @
19:12:27  <andythenorth> I found my wife's password, found the bootstrap fails with python 2.5
19:12:32  <andythenorth> installed python 2.6
19:12:41  <andythenorth> bootstrap is fine, but there's no gcc on the box :P
19:12:46  <andythenorth> so PIL won't build
19:14:47  <andythenorth> no solution to that one
19:15:28  * andythenorth blitzes PIL from his own system to see what happens
19:18:02  <andythenorth> bad
19:18:44  <Zuu> sorry planetmaker I worte a python script instead of making a Makefile :-)
19:19:18  <planetmaker> hm?
19:19:18  <Zuu> You even need to edit it manually to set your own path to openttd(.exe) before using it.
19:19:29  <Zuu> @r27 of the turorial
19:19:32  <planetmaker> ah
19:20:04  <planetmaker> Oh, I don't mind.
19:20:15  <planetmaker> Make is also just means to chain commands ;-)
19:21:42  <planetmaker> and important is that it works
19:21:55  <Zuu> indeed
19:22:05  <planetmaker> but granted, a hardcoded path to openttd is... bad ;-)
19:22:11  <andythenorth> hmm
19:22:21  <andythenorth> PIL 1.1.7 throws epic amount of build errors
19:22:33  <Zuu> though, I don't think that many people will need it unless they plan to do their own relaess :-)
19:23:00  <Zuu>  /releass/releases/
19:23:05  <planetmaker> uh. it hardcodes the .exe :-(
19:23:20  <planetmaker> well, indeed
19:23:56  <Zuu> And it still needs some glue on the OpenTTD side that doesn't yet exist. :-)
19:24:05  <Zuu> (namely save_scenario)
19:25:14  <andythenorth> ho
19:25:16  <andythenorth> I broke ply
19:30:36  <andythenorth> ok
19:30:43  <andythenorth> so I'm now a good test case for installing nml
19:30:48  <andythenorth> as I have broken it
19:30:51  <andythenorth> and can't fix it
19:31:38  <andythenorth> ImportError: No module named ply.lex
19:33:50  <andythenorth> so..
19:33:57  <andythenorth> what do I do?
19:37:36  <andythenorth> nvm
19:37:39  <andythenorth> installed the deps manually
19:37:52  <andythenorth> note: macports does *not* work for these deps
19:37:58  <andythenorth> despite reporting they installed ok
19:45:52  <andythenorth> ok
19:45:59  <andythenorth> so for OS X:
19:46:05  <andythenorth> > python 2.6.x
19:46:10  <andythenorth> > setuptools
19:46:38  <andythenorth> > manually install PIL following build instructions in PIL readme
19:47:06  <andythenorth> > manually install ply-lex, using python install
19:47:19  <andythenorth> > make nml
19:47:29  <andythenorth> > cd /opt/local/bin
19:47:49  <andythenorth> > ln -s PATH-TO-NML/nmlc nmlc
19:49:06  <andythenorth> worth a wiki note?
19:51:44  <Brot6> OpenGFX - Bug #3578 (New): Font update (PaulC) @
19:53:15  <Hirundo> To what extent are the recommendations stated here ( actually used in practice?
19:53:50  <andythenorth> currently?
19:53:54  <andythenorth> or in an ideal world?
19:54:08  <Hirundo> currently
19:54:29  <andythenorth> we recently reworked FIRS cargos in light of those
19:54:29  <Hirundo> ideal worlds exist only in the imagination and even those are often flawed
19:54:39  <andythenorth> we found we couldn't comply with them in all cases
19:55:03  <Hirundo> for hysterical raisins?
19:55:24  <andythenorth> yup
19:55:40  <andythenorth> or because we couldn't agree with them...or each other :)
19:55:49  <andythenorth> planetmaker: ^^^ your opinion?
20:26:16  <Brot6> Japanese Trains - Revision 51:72c2f4337d25: Fix some more trains weights (Toni) (dandan) @
20:28:02  <Brot6> Japanese Trains Addon - Revision 11:f7effa12467c: Fix some more trains weights (Toni) (dandan) @
20:37:41  <Brot6> Tutorial - Revision 28:1cfcb0fb7057: Add: tutorial scenario (Zuu) @
20:37:41  <Brot6> Tutorial - Revision 29:1b0fb2b49072: Add: Progress tutorial when the user adds an goto station or... (Zuu) @
20:37:41  <Brot6> Tutorial - Revision 30:2cbd6513b4a3: Change: Be more specific about the station window (Zuu) @
20:37:42  <Brot6> Tutorial - Revision 31:ab273277ac77: A change to STR_AIRPLANES_1_3_5_ORDERS was lost when the cha... (Zuu) @
21:12:46  <Hirundo_vm> NowI want so symlink nmlc so I can use it from any directory
21:12:58  <Hirundo_vm> I should place the symplink in /usr/local/bin, right?
21:13:08  <frosch123> mine are there :)
21:13:43  <Hirundo_vm> Now it won't let me do so, should I log in as root to fix that (which, I was told, is bad) ?
21:14:16  <planetmaker> Hirundo_vm: don't login as root. Use sudo
21:14:27  <Hirundo_vm> sudo won't let me
21:14:33  <planetmaker> hu?
21:14:51  <planetmaker> you might need to allow yourself to use sudo
21:15:03  <Hirundo_vm> " not in the sudoers file"
21:15:10  <Hirundo_vm> now how do I do that?
21:15:19  * Hirundo_vm feels dumb
21:15:21  <planetmaker> yeah. then you need to allow yourself to access sudo for that. How... hm..
21:15:48  <Hirundo_vm> "sudo allow-access sudo" sounds like catch-22 to me :-)
21:16:19  <planetmaker> <-- does that help?
21:16:20  <Webster> Title: Allow A Normal User To Run Commands As root Under Linux / UNIX Operating Systems (at
21:16:28  <planetmaker> and yes, you might need to login as root at least once ;-)
21:19:06  <^Spike^> root is where the real men are separated from the boys ;)
21:19:10  <^Spike^> the boys use sudo ;)
21:21:04  <andythenorth> Hirundo: installing nml made me feel dumb too
21:21:37  * andythenorth wonders if the buildout should be removed
21:21:39  <andythenorth> it doesn't work
21:22:49  <andythenorth> I think it's failing to call step for install
21:23:19  <Hirundo_vm> worked :-)
21:24:20  <Hirundo_vm> Thanks to the 'vi in 10 minutes' tutorial for telling me how to save and close a file ;-)
21:24:44  <andythenorth> hmm
21:24:52  * andythenorth just tried following the instructions in nml readme
21:25:05  <andythenorth> "NML uses buildout for packaging / installation. To install NML run:
21:25:05  <andythenorth> python install"
21:25:14  <andythenorth> that's not related to buildout
21:25:27  <andythenorth> and that also fails on the mac
21:27:06  <andythenorth> does anyone know if toni babeloni got his nml working?
21:27:17  <Hirundo_vm> Is there any reason to keep the xor refit-mask property?
21:27:30  <andythenorth> in nml?
21:27:39  <Hirundo_vm> Except hysterical reasons
21:27:40  <andythenorth> I'd recommend binning it
21:27:41  <Hirundo_vm> andythenorth: yes
21:27:43  <andythenorth> it sucks
21:27:53  <andythenorth> it's flawed by design
21:28:30  <andythenorth> a remarkably bad piece of design with a crippling logical flaw
21:28:35  <andythenorth> but we all do that :)
21:29:29  <Hirundo_vm> Probably it needs some grace period to allow grf authors to switch over
21:30:37  <planetmaker> nml 0.3 might deprecate it, 0.4 might remove it ;-)
21:32:28  <Rubidium> 0.3 is still "trunk" right?
21:32:40  <Rubidium> so just add the deprecation to 0.2.x and bin it even earlier ;)
21:32:41  * andythenorth reads about PILlow
21:32:41  <andythenorth>
21:32:42  <Webster> Title: Pillow 1.7.6 : Python Package Index (at
21:34:49  <Ammler> Hirundo_vm: you could symlink to ~/bin which doesn't need root
21:34:55  <Hirundo_vm> climate variable is useless now right? Climate cheat got banned, so we can read the value via action6/D
21:35:18  <planetmaker> yes, Rubidium. Feasible
21:35:46  <Ammler> you might need to add ~/bin to your $PATH
21:35:50  <Hirundo_vm> Ammler: My system didn't come with a ~/bin, though I could add that
21:35:53  <Brot6> BANDIT - Feature #3579 (New): CMS needs to insert trailer graphics files (andythenorth) @
21:36:12  <^Spike^> you can create that folder freely
21:36:25  <^Spike^> just need to log out and back in again to get it active if i'm right
21:37:12  <Ammler> well, if there is no ~/bin, it isn't part of $PATH either, most probably
21:37:31  <Hirundo_vm> it's indeed not
21:37:40  <Hirundo_vm> though the discussion is moot now
21:37:41  <Ammler> so you need to add that via .bashrc or however you intialize your shell
21:39:10  <Ammler> not using root does not mean using sudo :-P
21:39:26  <Hirundo_vm> I don't know how my shell is initialized
21:39:47  <andythenorth> me neither
21:39:52  <Ammler> check your dotfiles
21:39:59  <andythenorth> how can we make this better?
21:40:08  <Ammler> andythenorth: this is for devs only, anyway
21:40:20  <andythenorth> installing nml is for devs only?
21:40:25  <Brot6> BANDIT - Feature #3580 (New): CMS needs to insert truck and trailer lengths (andythenorth) @
21:40:25  <Ammler> this way, yes
21:40:38  <Ammler> others use the installer
21:40:41  <Hirundo_vm> This has not much to do with installing nml
21:40:52  <Ammler> exactly
21:40:52  <andythenorth> which installer?
21:41:08  <Ammler> or whatever :-)
21:41:12  * andythenorth contemplates breaking nml again to test the installer
21:41:49  <Ammler> symlinking to the repo is that you can use the script directly from repo
21:41:52  <Hirundo_vm> Basically I needed a linux VM to work on NML as cygwin sucks monkey testicles, but I suck monkey testicles at using linux
21:42:58  <planetmaker> mingw maybe?
21:43:01  <andythenorth> we should remove the buildout
21:43:05  <andythenorth> it's broken and adds cruft
21:43:12  <Hirundo_vm> I have grown up with buttons to click on instead of commands to type, nothing can fix that
21:43:32  <planetmaker> that surprises me a bit to hear, Hirundo_vm :-)
21:43:47  <Rubidium> ghehe...
21:43:47  <andythenorth> the installer is
21:43:51  <Ammler> sorry, but who the hell told you to start using linux with debian?
21:44:00  <Rubidium> guess I'm then just old enough and started young enough with computers
21:44:19  <Hirundo_vm> Ammler: I'm surprised at how well it works actually
21:44:22  <Rubidium> though back when I started with Linux it was even more a hell to start with
21:44:34  <Hirundo_vm> I just dug up some tutorial and worked through it
21:44:53  <andythenorth> anyone mind if I remove the buildout and commit it?
21:45:07  <Ammler> andythenorth: it isn't just that file
21:45:09  <Hirundo_vm> Actually when I take the time to study cli stuff, I can work with itquite well, hg for example
21:45:16  <Ammler> afaik, there is also other parts related to it
21:45:20  <andythenorth> Ammler: I know which files to remove
21:46:45  <andythenorth> buildout.cfg docs/zpl-2.1.txt
21:46:54  <andythenorth> and update readme to remove (incorrect) references to buildout
21:47:16  <Brot6> Tutorial - Revision 32:d0889c0b673d: Add: Detect if OpenTTD is too old (Zuu) @
21:47:16  <Brot6> Tutorial - Revision 33:61f1fdbb5adc: Add: Script to make tar file (Zuu) @
21:47:16  <Brot6> Tutorial - Revision 34:0475c584077d: Version 6 (Zuu) @
21:47:17  <Ammler> maybe add pil and ply as requires to
21:47:29  <andythenorth> that won't work
21:47:36  <andythenorth> PIL is not setuptools compatible
21:47:46  <andythenorth> PIL is barely-anything compatible
21:47:51  <Ammler> hmm, true
21:48:06  <andythenorth> if we want to automate installing the deps, we need buildout or pil
21:48:11  <andythenorth> pil / pip /s
21:48:18  <andythenorth> pip seems more widely liked
21:48:36  <Ammler> yes, pip reads
21:48:39  <andythenorth> if all the deps are python there is no gain from buildout
21:48:52  <Ammler> at least, it could alert, if there is no pil
21:50:25  <andythenorth> I really wanted to remove the need to screw your system
21:50:35  <andythenorth> pip coud do that with a virtual-env
21:50:44  <andythenorth> but virtual-env is yet one more dependency
21:51:22  <Ammler> andythenorth: the issue with pil is that it needs building
21:51:30  <Ammler> with gcc and such stuff
21:51:31  <andythenorth> buildout builds it
21:51:40  <andythenorth> yes it needs gcc as I found earlier
21:51:45  <Ammler> it's not a simply python script
21:51:51  <andythenorth> gcc is a significant dependency
21:52:09  <Ammler> so you can't expect someone able to build it
21:52:17  <Ammler> people need to have it on their system already
21:52:38  <andythenorth> PIL or gcc?
21:52:43  <Ammler> PIL
21:53:11  <andythenorth> I can't see a way around that no
21:53:18  <andythenorth> it's annoying though
21:53:39  <Ammler> well, it should not be a issue of nml
21:53:43  <Ammler> nml simply requires it
21:55:06  <andythenorth> it's an issue for adoption of nml
21:56:16  * andythenorth wonders if the buildout works, but has path issues
21:56:20  <Ammler> for linux, installing nml is no issue
21:56:26  <andythenorth> that's lovely
21:56:32  <andythenorth> let's all switch to linux :P
21:56:40  <Ammler> maybe for osx, it would be an idea to use maxports or how that is called
21:56:48  <andythenorth> macports fails for this case
21:57:04  <andythenorth> it doesn't install ply-lex or pil correctly
21:57:06  <andythenorth> I tested
21:57:10  <Ammler> and windows use the exe from nightlies
21:59:53  <Brot6> NewGRF Meta Language - Revision 1792:901b872dfd98: Cleanup: Remove now-redundant climate action2 ... (Hirundo) @
22:00:09  <Ammler> hmm, what about building e mac exe too?
22:00:36  <Ammler> at least including the binary for pil?
22:02:07  <Ammler> or couldn't the openttd compile farm build nml?
22:08:58  *** andythenorth is now known as Guest187
22:08:59  *** andythenorth has joined #openttdcoop.devzone
22:09:01  *** andythenorth has left #openttdcoop.devzone
22:09:12  <Hirundo_vm> planetmaker: Yexo: Do you agree with deprecating the xor refit mask? And removing it in a month or so
22:09:17  *** andythenorth has joined #openttdcoop.devzone
22:09:27  *** Guest187 has quit IRC
22:09:29  <andythenorth> does the .exe eliminate all deps?
22:09:54  <planetmaker> I'd not be that fast with it, Hirundo_vm
22:09:57  <planetmaker> but yes
22:10:00  * andythenorth reads docs for answer
22:10:26  <andythenorth> so windows is sorted
22:10:40  <andythenorth> linux is trivial
22:10:40  <andythenorth> just OS X is problematic
22:10:44  <andythenorth> it seems
22:11:19  <Hirundo_vm> planetmaker: For now, I just wanted to add a generic.print_warning upon usage
22:12:06  <planetmaker> yes, that's fine with me. But... Yexo will need to agree, too :-)
22:12:14  <andythenorth> do we have OS X users for nml?
22:14:21  <Hirundo_vm> Do we have OS X users for openttd?
22:14:55  <andythenorth> yes
22:15:54  <Rubidium> hmm... andy isn't using OSX anymore?
22:16:32  * andythenorth wonders if OS X users should just buy Crossover and use the nmlc.exe
22:16:48  <andythenorth> Crossover is a packaged Wine
22:16:48  <Rubidium> might be better for OpenTTD as well
22:17:06  <andythenorth> I tested OpenTTD with it when OS X support was deprecated
22:17:23  <andythenorth> it works adequately.  It noticeably lags only occasionally
22:17:32  <andythenorth> CPU use is higher than for the OS X binary
22:17:49  <andythenorth> I used crossover for grfcodec and renum for a long time
22:17:50  <Rubidium> well... OS X support lags noticeably ;)
22:18:39  <andythenorth> crossover is £26
22:18:39  <andythenorth> they should just buy that and use the .exe
22:18:52  <andythenorth> the alternative is:
22:19:02  <andythenorth> - install Apple Developer Tools (or build your own gcc)
22:19:06  <andythenorth> - install python
22:19:09  <andythenorth> - install PIL
22:19:13  <andythenorth> - install ply
22:19:25  <andythenorth> - make a symbolic link
22:19:37  <andythenorth> all the time typing 'sudo' which you are conditioned to never do
22:19:42  <andythenorth> it's just not going to work
22:21:00  <Rubidium> that's pretty easy isn't it?
22:21:28  <andythenorth> it is if I write down all the instructions in steps
22:21:34  <Rubidium> just get genius from the app(le) store
22:25:11  <Brot6> NewGRF Meta Language - Feature #3581 (New): Consider removing buildout (andythenorth) @
22:28:57  <Brot6> NewGRF Meta Language - Feature #3582 (New): Write instructions for installing on OS X (andythenorth) @
22:31:52  *** andythenorth has left #openttdcoop.devzone
22:32:13  <Brot6> NewGRF Meta Language - Feature #3581: Consider removing buildout (andythenorth) @
22:32:43  <Brot6> Tutorial - Revision 35:765cb2c0d384: Update: string updates by Pingaware (Zuu) @
22:33:01  *** frosch123 has quit IRC
22:38:43  *** JVassie_ has joined #openttdcoop.devzone
22:42:45  <Brot6> NewGRF Meta Language - Feature #3583 (New): deprecate / remove refittable_cargo_types (Hirundo) @
22:43:11  <Hirundo> Close inspection of openttd code taught me that 'just' emitting a warning has issues
22:43:43  <Hirundo> Leaving properties unset can cause bizarre issues in some cases, this is one of them (see #3583)
22:43:43  <Brot6> Hirundo: #3583 is "NewGRF Meta Language - Feature #3583: deprecate / remove refittable_cargo_types - #openttdcoop Development Zone"
22:44:27  *** JVassie has quit IRC
22:49:36  *** Hirundo_vm has quit IRC
22:50:50  *** JVassie has joined #openttdcoop.devzone
22:55:57  *** JVassie_ has quit IRC
22:59:01  *** ODM has quit IRC
23:14:24  *** JVassie has quit IRC
23:29:34  *** KenjiE20 has quit IRC
23:29:55  *** KenjiE20 has joined #openttdcoop.devzone
23:29:57  *** KenjiE20 has quit IRC
23:36:13  *** KenjiE20 has joined #openttdcoop.devzone
23:57:04  *** Zuu has quit IRC

Powered by YARRSTE version: svn-trunk