IRC logs for #openttd on OFTC at 2020-10-30
⏴ go to previous day
07:45:43 *** sla_ro|master has joined #openttd
08:25:44 *** andythenorth has joined #openttd
08:37:10 *** sla_ro|master has joined #openttd
08:59:48 *** WormnestAndroid has quit IRC
09:00:02 *** WormnestAndroid has joined #openttd
09:43:38 *** Laedek_ has joined #openttd
09:52:34 *** jottyfan has joined #openttd
10:01:46 *** Progman has joined #openttd
10:20:06 *** sla_ro|master has joined #openttd
10:27:46 *** longtomjr has joined #openttd
10:42:44 <TrueBrain> turns out a lot of people were using URLs into the OpenTTD infrastructure that they shouldn't have
10:42:53 <TrueBrain> URLs that no longer exist, even :P
10:43:09 <TrueBrain> for example, the debian watch file was using an URL it really should never have used :P
10:43:23 <TrueBrain> sadly, retiring URLs on the web is .. difficult
10:46:01 <TrueBrain> I did a GitHub code search on old URLs ... I closed it :P
10:47:02 <LordAro> TrueBrain: how many results? :p
10:48:16 <TrueBrain> 176 hits on one URL, 58 on another .. all various of projects, like debian-watcher, auto-downloads, dockers, etc
10:48:33 <LordAro> well, they'll get the idea soon enough
10:48:54 <TrueBrain> that is the thing .. it turns out it takes months for someone to report it :P
10:49:01 <TrueBrain> (for upstream to fix it)
10:49:08 <TrueBrain> says something about the popularity, possibly
10:51:21 <longtomjr> Old links seems to be popular yep
10:51:56 <LordAro> tbf, it was never particularly clear which links you weren't "supposed" to use
10:52:52 <TrueBrain> the ones you cannot find easily :P
10:53:05 <TrueBrain> those always had links we support
10:53:14 <TrueBrain> but somehow people find these weird links from the weirdest places
10:53:43 <TrueBrain> and sometimes I have no clue what-so-ever how they found a link
10:54:03 <TrueBrain> example, master.binaries.openttd.org was the primary server from the mirror-network; from there, it was distributed to all the other mirror servers
10:54:12 <TrueBrain> it was only meant to be used by the other mirrors to fetch their downloads
10:54:21 <TrueBrain> but somehow .. people found that URL and used it
10:54:58 <TrueBrain> one thing I learnt from OpenTTD: you can try to hide an URL how-ever you like, people will find it, and will use it, and will complain if it no longer works :P
10:55:26 <LordAro> mm, probably just taking the URL it resolves to, or something
10:55:45 <LordAro> i imagine people will start using the direct s3 links this time round :p
10:55:55 <TrueBrain> yeah, I made sure they really cannot :)
10:56:05 <longtomjr> going through the cdn?
10:56:07 <TrueBrain> longtomjr: yup, anything on cdn.openttd.org is
10:56:36 <longtomjr> is there a CDN link for the checksums as well?
10:56:37 <TrueBrain> LordAro: the S3 bucket is not available from any URL outside of the AWS VPC :P
10:56:59 <TrueBrain> okay, that is silly
10:57:33 <longtomjr> If I were a machine, how do I get the checksum without scraping?
10:58:25 <LordAro> is there a checksum for the manifest file? :p
10:58:47 <TrueBrain> LordAro: and the checksum of the checksum?
10:58:50 <TrueBrain> and of that checksum? :P
10:59:10 <TrueBrain> longtomjr: it is how the website knows what to use btw .. :D
10:59:29 <longtomjr> We might need a new EC2 instance to just hash everything recursively
10:59:40 <TrueBrain> but who is hashing the EC2?
10:59:51 <longtomjr> Time for Trusted Execution Environments
11:00:00 <longtomjr> Anyone have credits on Azure
11:01:06 <longtomjr> Or maybe using replicatable builds. (there is a better term for it, but builds that will have the same checksum (and contents) when built on another machine)
11:03:39 <TrueBrain> reproducible builds :)
11:04:17 <LordAro> debian's very keen on it these days
11:05:45 <longtomjr> How easy is it to do it. I have not really looked into it much at all, and I only rember it from a couple of years ago
11:06:32 <LordAro> mostly it's eliminating timestamps and absolute paths from everything
11:07:02 <longtomjr> and using the same compiler and build dep versions
11:07:50 <TrueBrain> it is not as trivial as it sounds, most of the time
11:08:01 <TrueBrain> often you need an LD_PRELOAD to eliminate timestamps, basically
11:08:25 <TrueBrain> tools do start to get options to do these things for you
11:08:29 <TrueBrain> but .. it is a slow process :P
11:09:14 <TrueBrain> andythenorth: you do know this is #openttd, right? :P
11:09:21 <TrueBrain> wow, that was random as fuck ... :P
11:09:24 <longtomjr> That is really interesting
11:09:32 <andythenorth> from lordaro's link
11:10:06 <TrueBrain> I like their title .. it says absolutely nothing :D
11:10:09 <TrueBrain> just a bunch of words, basically :P
11:11:32 <TrueBrain> yeah .. it is there :P
11:11:52 <TrueBrain> I love how people are proud enough to use PHP to make a badge for it :D
11:12:06 <TrueBrain> longtomjr: in some ways, aren't we all? :)
11:12:07 <longtomjr> ^ my wife asking for translator access
11:12:22 <TrueBrain> ah, finally a reason to deny one :P
11:12:26 <longtomjr> plan get her into train game is going along quite nicely
11:12:48 * andythenorth tries to envisage a world where my wife applies for translator access
11:13:36 <longtomjr> She is currently doing her Honors in Translation and Editing, so thought it would be fun for her to do more work. :P
11:13:47 <TrueBrain> and you are right about that! :P
11:18:07 <longtomjr> Getting a 404 I think
11:18:40 <TrueBrain> because you are not part of that team :)
11:18:45 <TrueBrain> and she needs to accept the org-invite first
11:18:54 <TrueBrain> (as mentioned in the bot's reply too :D)
11:19:13 <TrueBrain> "If you are not a member of OpenTTD yet, you will receive an invitation shortly (or click here).
11:19:13 <TrueBrain> You will have to accept this invitation before your membership is updated."
11:19:26 <longtomjr> It is because she did not accept the invite before yes, I thought she did already.
11:19:50 <TrueBrain> 8 people failed to accept the invitation ..
11:19:56 <TrueBrain> (out of the 50 or so)
11:20:00 <TrueBrain> it is not a good track-record ...
11:20:08 <TrueBrain> but we have no idea how to make it more obvious
11:21:24 <longtomjr> what is the timeout?
11:22:00 <TrueBrain> LordAro: but they ALSO get an email from GitHub inviting them
11:22:15 <TrueBrain> anyway, more important, wiki
11:23:24 <longtomjr> I think having an icon or something inbetween the crumbs will help, currently it looks a bitl ike tabs
11:23:42 <TrueBrain> that is a nice idea
11:24:12 <TrueBrain> something like this or so (but prettier, ofc)
11:24:13 <longtomjr> Maybe the first crumb can also be ommitted, since the title links to that page as well
11:24:28 <longtomjr> Yep, something like that
11:24:48 <TrueBrain> the first one lets you easily see how to get back
11:24:53 <longtomjr> Maybe having the root there makes sense yep
11:24:55 <TrueBrain> that sounds like one of the more useful ones to have? :)
11:25:16 <TrueBrain> okay, so how am I going to make this pretty .. hmm ..
11:25:25 <longtomjr> what is the deepest nested page that you have?
11:26:03 <TrueBrain> the one I linked, I think
11:26:16 <longtomjr> Ah ok, then that should be fine.
11:26:54 <longtomjr> You can use pseudo-elements to do the crumbs
11:28:12 <TrueBrain> yeah, I just need to find a nice UTF-8 char for it
11:30:14 <TrueBrain> #navigation-bar li.crumb:not(:first-child)::after
11:30:18 <TrueBrain> I do love CSS selectors :)
11:31:01 <TrueBrain> hmm, I like ❱ more :)
11:34:11 <TrueBrain> well, this will have to do for now :)
11:36:07 <TrueBrain> right, now something to do folder listing .. hmm
12:16:52 <TrueBrain> guess that is sufficient for now
12:17:53 <TrueBrain> now only where to link it from :D
12:18:05 <longtomjr> Still needs the breadcrumbs
12:18:56 <longtomjr> Also, the breadcrumb links does not link to the folders yet
12:20:20 <TrueBrain> they are never going to
12:20:28 <TrueBrain> I do however need a place to link to folders, not sure where ..
12:21:56 <TrueBrain> the breadcrumbs normally point to an "index.html" (in this case, "Main Page"). That is most likely what a user wants to visit anyway
12:22:05 <TrueBrain> visiting the folder structure is a more rare event
12:34:26 <TrueBrain> okay, I made something at the bottom .. hopefully that is enough :)
12:47:59 <longtomjr> True, is there a history of translations pre-eints?
12:48:18 <longtomjr> or at least the current eints history
12:48:46 <longtomjr> Some of the af_ZA translations look a lot like machine translations
12:52:25 <andythenorth> does Horse need GT3 locomotive?
12:52:31 * andythenorth awaits crowd answers
12:52:32 <longtomjr> ok, will have a look, sorry for bothering with this :)
12:52:59 <TrueBrain> except for some bugs, it should most of the time mention who made the change
12:53:09 <longtomjr> Yep, blame is probably the easiest
12:53:27 <TrueBrain> longtomjr: and you are not bothering me with these kind of questions; they are always welcome to be asked :)
12:53:31 <longtomjr> What is WebTranslator v3.0?
12:53:44 <TrueBrain> you will find v2 if you go back even further
12:53:47 <TrueBrain> and ... v1 before that :P
12:53:56 <longtomjr> Is there history of that somewhere?
12:53:57 <LordAro> looks like it's been a long time since there were any meaningful Afrikaans translations
12:54:15 <TrueBrain> longtomjr: not written down; also not really interesting for most people
12:55:10 <longtomjr> My wife does not want to just change things around without knowing the context, but there are some translations that is very literal and does not sound or read like native Afrikaans
12:55:49 <TrueBrain> the context is ingame, tbh
12:55:57 <TrueBrain> what past people did, is not that relevant
12:56:09 <TrueBrain> what we want to avoid, is one translator doing A, and the other doing B, ofc
12:56:15 <TrueBrain> we had that with some .. they kept changing it back and forth
12:56:21 <longtomjr> I meant more the history. Ok so it is fine if she updates some of the more literal translations.
12:56:25 <TrueBrain> but given Afrikaans hasn't had a change in ... YEARS
12:56:28 <TrueBrain> that shouldn't be an issue here :)
12:56:52 <longtomjr> Ok, but with the new teams setup when a new Afrikaans translator signs up they should be able to communicate.
12:56:57 <TrueBrain> what we HOPE that happens, when there are more than 1 translator, they start on GitHub a discussion in their language section, to set some ground rules
12:57:13 <longtomjr> Yep, will pass that on to her.
12:57:21 <LordAro> of course, none of that has happened yet
12:57:45 <TrueBrain> and she is the only one for af_ZA :)
12:58:00 <FLHerne> andythenorth: tbh, I've never really used the UKRS2+ one
12:58:09 <TrueBrain> right, lunch, I really should do lunch ..
12:58:43 <FLHerne> More than one is unrealistic, and just one is pointless micromanagement
12:59:48 <TrueBrain> LordAro: and we didn't even get email about it! :P
13:00:09 <TrueBrain> LordAro: honestly, I am as shocked as you are that it is used :D
13:00:18 <LordAro> i get so many github emails i probably wouldn't have noticed it
13:00:29 <TrueBrain> my settings are pretty good these days
13:07:21 <longtomjr> When did you move over to the github integrated eints?
13:11:30 <TrueBrain> Few months ago .. why?
13:13:36 <longtomjr> There is some recent afr translations that were in this year still, so just wondered.
13:19:13 <andythenorth> FLHerne correct answer :)
13:38:55 *** snail_UES_ has joined #openttd
14:30:20 *** jottyfan1 has joined #openttd
14:36:19 *** jottyfan1 is now known as jottyfan
15:13:45 <TrueBrain> it reports all errors found during rendering of all pages
15:14:09 <TrueBrain> 3505 pages report something :)
15:17:37 <LordAro> out of how many pages?
15:18:15 <TrueBrain> so over 50% contains some form of error
15:18:22 <TrueBrain> most of the cases it links to a page that doesn't exist
15:19:54 <longtomjr> Is redlinks an error?
15:19:57 <TrueBrain> and, 0 exceptions happened during rendering \o/
15:20:46 <TrueBrain> because someone should fix that
15:21:19 <longtomjr> Fair enough, Is there only 1 severity for the errors?
15:21:43 <TrueBrain> errors is a severity, isn't it? :D
15:21:43 <LordAro> mm, definitely a low severity error if any
15:22:02 <TrueBrain> everything is "low" if you look at it like that
15:22:16 <TrueBrain> label it how-ever you like; it requires attention :)
15:22:58 <longtomjr> I am just thinking if there is a dashboard that shows a list of pages that need attention, it might be nice to have redlinks grouped and seperate from other issues
15:23:20 <TrueBrain> problem is that a lot of those broken links are not intentional
15:24:23 <TrueBrain> but okay, the list above is for frosch123 specifically, as many broken links are due to the migration, and requires attention on another level
15:25:04 <TrueBrain> but take for example this page:
15:25:08 <longtomjr> Ah some of them are duplicates, as in if an error gets fixed with the migration, the several of the reported issues will be fixed?
15:25:12 <TrueBrain> it is really bad those links are broken
15:25:36 <TrueBrain> "duplicates"? No, there are no duplicated errors
15:25:46 <TrueBrain> an error can happen in a template, and by fixing 1 many pages can be fixed
15:25:50 <TrueBrain> that can be considered "duplicate", but no
15:25:51 <longtomjr> as in it refers to one bug
15:26:20 <longtomjr> Those are all external links
15:26:30 <longtomjr> or am I missing something?
15:26:38 <TrueBrain> I do not validate external links, so no :)
15:26:41 <TrueBrain> these are all innerlinks :)
15:26:57 <TrueBrain> but again, that gist is not for you guys; frosch will know what to do with it :)
15:27:17 <longtomjr> sorry, were just curious
15:27:35 <TrueBrain> it is okay; it just brings us in a weird route :P
15:27:55 <TrueBrain> hmm .. that figure-of-speech does not translate well to English :D
15:28:09 <TrueBrain> well, to give an example of where the migration fails:
15:28:28 <TrueBrain> the links on top are migrated wrong; this can either be that it cannot be automated, or that there is a boo-boo
15:29:10 <TrueBrain> in this case it is because {{en/Release|prev=OpenTTD 0.5.0|next=OpenTTD 0.5.2|date=2007-04-20}} doesn't have its "prev" and "next" fixed
15:29:16 <TrueBrain> which is difficult to do for any script, to know it should fix those
15:32:10 <TrueBrain> hmm .. this is an interesting case .. I wonder what frosch123 wants here ... should "[[abc]]" when you are in "en/Archive/Bla/def" point to "abc" or to "en/Archive/bla/abc"
15:32:35 <TrueBrain> in other words, should it always be absolute, or relative, or try-absolute-first-relative-after?
15:35:58 <TrueBrain> either do not link it, or who-ever started this should have finished it :P
15:36:32 <TrueBrain> the only thing this causes, that someone makes a page somewhere else, totally unrelated, with the same name, and they now link ..
15:36:54 <TrueBrain> so yeah, when talking about "severity" of errors ... I rather say: this is not right and requires some form of attention :)
15:37:21 <longtomjr> Yep, I agree with you there for sure
15:37:37 <TrueBrain> so possibly I should rename "errors" with something else
15:37:48 <TrueBrain> "messages during rendering" :P
15:39:33 <longtomjr> even the links on that page will need some cleanup
15:39:58 <TrueBrain> I would expect so ;) As my given scenario most likely became reality :D
15:40:00 <longtomjr> Maybe linked with the settings page
15:40:15 <TrueBrain> but okay ... migrate first, fix after :D
15:40:31 <TrueBrain> many many pages need attention ..
15:40:39 <TrueBrain> for example, there is a template that renders an image with a caption
15:40:45 <longtomjr> Accounts cannot be created for the wiki atm right?
15:40:46 <TrueBrain> like ... yes ... wikitext has that already
15:41:39 <TrueBrain> like ... I doubt even I can provision new accounts atm within a reasonable amount of time :P
15:42:25 <longtomjr> Goodness, Is the source mass editable on the migration wiki (without creating conflict nightmares?). I can maybe start a PR to clean some of this up;.
15:43:27 <TrueBrain> with truewiki (still love the name :P) you can mass-edit via GitHub, yes
15:43:35 <longtomjr> Anyways, pinned it for the future.
15:43:41 <TrueBrain> but don't expect it to be online before the end of November :)
15:44:19 <TrueBrain> hopefully editing somewhat works by the end of this weekend, so we can start testing it .. and we will find tons of edge-cases and bugs
15:44:40 <TrueBrain> which normally takes anywhere from 2 to 4 weeks to iron out :)
15:45:39 <longtomjr> Would be happy to help test it if you need more hands.
15:46:16 <TrueBrain> we tend to try drag people that idle here in our "now test this shit" phase, so yeah, please :)
15:46:37 <TrueBrain> (except for the known stuff, like lots of broken links and images)
15:46:58 <TrueBrain> anyway, I am off for a few hours; talk to you guys tonight!
16:01:53 <longtomjr> What is the difference between the Archive pages and other pages?
16:04:30 <longtomjr> some things I picked up
16:09:57 *** tokai|noir has joined #openttd
16:09:57 *** ChanServ sets mode: +v tokai|noir
17:00:17 *** Wormnest has joined #openttd
17:03:18 *** frosch123 has joined #openttd
17:43:57 <FLHerne> Hm, grfcodec doesn't compile :p
17:45:03 <FLHerne> Oh, someone else already found that
17:46:04 <FLHerne> Why is grfcodec code formatted so horribly?
17:46:16 <FLHerne> Hadn't they invented spaces back then?
17:48:22 <andythenorth> it has taken me a week to draw 1 ship
17:48:31 <andythenorth> and I have about 60 planned
17:48:41 <andythenorth> this is not desirable
17:51:49 <frosch123> FLHerne: you are looking at nforenum
17:53:48 <frosch123> wait, dalestan is now gone for > 10 years?
18:03:42 <LordAro> FLHerne: grfcodec compiles now :p
18:04:04 <LordAro> i wanted to set up all the relevant dorpsgek & github actions gubbins, but never got around to it
18:04:09 <FLHerne> There are some other compile warnings with gcc 10
18:04:17 <LordAro> this does not surprise me
18:04:19 <FLHerne> I was going to try and fix them, but then I saw the code :p
18:04:21 *** HerzogDeXtEr has joined #openttd
18:05:06 <FLHerne> I swear half of inlines.h is reimplementing bits of the stdlib for no reason
18:05:55 <frosch123> you are probably better off reimlpementing it in python :)
18:06:24 <LordAro> FLHerne: oh god, don't spend time looking at it
18:06:29 <LordAro> you'll make your eyes bleed
18:07:24 <FLHerne> frosch123: I got into this because I want to make NML tests compile the resulting NFO using grfcodec
18:07:36 <FLHerne> To check that it produces the same as direct grf output
18:09:05 <frosch123> there are some parts if grf, that can be reordered without changing the meaning
18:09:16 <frosch123> pseudo sprites are obviously the same, or everything would break
18:09:27 <frosch123> realsprites are the same because they literally use the same code for compression
18:09:36 <frosch123> FLHerne: okay, have fun :)
18:17:03 <FLHerne> frosch123: Is there a benefit from not using Unicode in grf strings?
18:17:20 <FLHerne> nmlc goes to considerable effort to use only ascii when it can
18:17:22 <frosch123> smaller file size :p
18:17:40 <frosch123> otherwise i am not aware of any difference
18:17:48 <frosch123> ottd transcodes them to the same on loading the grf
18:17:54 <FLHerne> How is unicode encoded in the grf?
18:18:08 <frosch123> there is a starting þ
18:18:20 <FLHerne> UTF-8, or some fixed-width encoding?
18:18:20 <frosch123> the control characters are mapped into the private area
18:18:46 <FLHerne> Then...the file size should be virtually identical anyway?
18:19:10 <FLHerne> Except two bytes per string for the thorn...
18:19:31 <LordAro> not a bad fallback, tbh
18:19:34 <FLHerne> Yeah, I was reading that last night
18:19:49 <frosch123> it's proably to simplify using the control codes
18:19:57 <frosch123> so grf people do not need to learn the private range
18:22:57 *** gelignite has joined #openttd
18:26:14 <frosch123> FLHerne: maybe action8 does not support utf-8
18:26:43 <FLHerne> frosch123: I think the "parse broken stuff as old bytes" thing covers that anyway?
18:27:05 <frosch123> ah no, it does as well
18:37:45 <frosch123> so, factorio is copying the ttd title screen :p even the orange game title
18:43:07 <frosch123> i guess V got homesick
18:44:00 <longtomjr> (he is still watching)
18:46:03 <Eddi|zuHause> or it's like The Big Bang Theory, who remade advertisment pictures like from "Dexters Laboratory"
18:51:58 <DorpsGek_III> - Update: Translations from eints (by translators)
19:08:34 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #8331: Fix: [NoAI] don't notify caught exceptions https://git.io/JTSNd
19:32:22 *** otetede has joined #openttd
20:02:46 *** andythenorth has joined #openttd
20:09:54 <frosch123> i am not sure whether that link to the main page makes sense at that point. maybe there could be a "/" folder that lists P,age Category, File, Template instead
20:22:01 *** andythenorth has joined #openttd
20:37:24 *** Gustavo6046 has joined #openttd
21:03:17 *** andythenorth has joined #openttd
21:21:15 *** virtualrandomnumber has joined #openttd
21:21:34 *** virtualrandomnumber has quit IRC
21:28:32 *** jottyfan has joined #openttd
21:40:34 <TrueBrain> frosch123: nice catch :D I changed the first entry in breadcrumbs a bit too often :P
21:41:00 <TrueBrain> frosch123: I wanted the most left button to always bring you back to the main page, to give some consistency
21:41:41 <TrueBrain> but I can add another page in between there that allows browsing Files etc
21:41:53 <TrueBrain> just the current structure in Files is not ideal for that ..
21:42:13 <TrueBrain> but I like the idea of at least being able to browse Templates and Categories
21:42:39 <TrueBrain> hihi, and I should hide the View Source / History buttons :P
21:42:58 <TrueBrain> but otherwise, this a bit what you had in mind?
21:51:16 *** WormnestAndroid has quit IRC
21:51:35 *** WormnestAndroid has joined #openttd
22:14:49 <TrueBrain> frosch123: I tweaked some things about /Folder/. It is still not really how I would like it, especially code-wise .. lot of exceptions etc, never a good sign. But I think this works sufficiently for v1. Let me know what you think!
22:15:07 <TrueBrain> (and all URLs below :P)
22:18:22 <frosch123> oh... now i see what you meant with the problems about File:
22:19:45 <frosch123> btw. i think my next export will put more folders into Template and File
22:20:15 <frosch123> i'll probably give those TOC templates a consistent name and put them into directories matching the pages they contain
22:20:56 <TrueBrain> frosch123: that would be lovely :)
22:21:03 <TrueBrain> and File is empty because I filter for ".mediawiki" :D
22:21:25 <TrueBrain> I really have to rethink how truewiki is build up tbh
22:21:25 <frosch123> should i rename LICENSE to .md, so it does not show up?
22:21:41 <TrueBrain> it is now a bit of a mess with endless exceptions for the namespaces ..
22:21:55 <TrueBrain> hmm .. I am fine with the LICENSE to work; I was more thinking of fixing it so it does work :P
22:22:08 <frosch123> well, if we have to start over, we can name it OpinionatedWiki :)
22:22:34 <TrueBrain> I now understand more how I can logically put namespaces in there honestly
22:22:48 <TrueBrain> but I first want to build the edit part tbh
22:22:57 <TrueBrain> otherwise it will never get to a "workable" state :D
22:24:16 <TrueBrain> or should we just make them absolute paths?
22:25:13 <frosch123> the migration will make everything absolute paths
22:25:34 <frosch123> if it doesn't work, i missed something :)
22:25:40 <TrueBrain> what is nice about allowing relative, is that if we move folders, everything keeps working
22:25:51 <TrueBrain> but I am also absolutely fine fixing pages like that to use absolute
22:26:09 <TrueBrain> being explicit is useful in most cases, honestly
22:26:12 <frosch123> but how relative should they be? do you also want to support ".." ?
22:26:25 <TrueBrain> "current folder" or bust :P
22:26:36 <TrueBrain> okay, well, in that case we have some manual fixing to do after migration :)
22:26:41 <frosch123> so you want detect absolute paths by checking whether they start with a language code?
22:26:55 <frosch123> and forbid people to create pages/templates named like a language code?
22:26:57 <TrueBrain> I was more thinking: test absolute first (if file exists), otherwise try relative
22:27:13 <frosch123> that does not work if the page does not yet exist
22:27:15 <TrueBrain> but okay okay, too many edge-cases, fiinnnneeeeeeeee
22:27:33 <frosch123> when i write a wiki page, i always add the links i want. later i click on the dead links to add the pages
22:28:11 <frosch123> i consider that easier than creating a page in a wild, and then figuring out how to link to it :)
22:28:12 <TrueBrain> how do you plan on doing the migration: just migrate with broken pages like this, and we fix them after manually?
22:28:17 <TrueBrain> or do you want to fix them during migration already?
22:28:51 <frosch123> in this specific case, i have code to convert those prev/next links
22:28:58 <frosch123> it just does not work apparently :)
22:29:11 <TrueBrain> the /Fr template is a bit weird in that it adds /Fr :P
22:29:38 <frosch123> yes, those need manual fixing :)
22:30:06 <frosch123> the templates have like 4 methods to link to pages/images given by parameters
22:30:12 <TrueBrain> k; so you just do as much as you can, and we go through it after fixing the rest; sounds like a plan :)
22:30:14 <frosch123> i consider unifying them a bit
22:30:24 <TrueBrain> ppllleeeeaaasssseeeee do :)
22:30:36 <TrueBrain> I dislike some have "Image:" implied, others do not, for example
22:31:27 <frosch123> for example: {{#if:{{{image|}}}|[[File:{{{image}}}]]|[[File:default.png]]}} is really complicated
22:31:42 <frosch123> compared to [[File:{{image|default.png}}]]
22:33:22 <TrueBrain> or: [[{{{image|File:default.png}}}]]
22:33:27 <frosch123> and then there is one template that composed a image name from two or three parameters
22:33:31 <TrueBrain> depends a bit if you want "File" to be out or inside the template :D
22:33:44 <TrueBrain> yeah, people tried insane stuff ....
22:34:21 *** Wormnest has joined #openttd
22:36:12 <frosch123> oh, lol, about the relative paths.... i just remembered there were some legacy build tools at my previous company that used "./" to distingiush relative from absolute paths
22:36:24 <TrueBrain> I know these kind of tools :)
22:36:32 <TrueBrain> haven't seen them in YEARS
22:36:47 <frosch123> really? i thought that was a unique idea. they kind of told people, relative paths have to start with ./
22:37:06 <frosch123> maybe there is some historic reason then, and it was needed in the 70s
22:37:31 <TrueBrain> well, starting apps from local folder has to be done with ./
22:37:33 <frosch123> i used to delete #ifdef ULTRIX :)
22:37:35 <TrueBrain> so it is not that an unique idea ;)
22:38:20 <TrueBrain> right, pushed the latest changed to the repos .. if you like, you can run --validate-all yourself now :)
22:38:34 <TrueBrain> tomorrow I will add GitHub login, and hopefully some basic edit functionality
22:38:41 <TrueBrain> just so we can get a bit of feeling for that part
22:39:06 <TrueBrain> I always assume that v1 needs to be redone because you learn during v1 what you want and don't want :)
22:39:17 <TrueBrain> so I love touching all parts a bit, to get a feel for it
22:39:26 <TrueBrain> that sounds REALLY wrong
22:39:56 <TrueBrain> I am not sure yet how we want to do editing tbh .. I think the URL will be something like /edit/ as prefix
22:40:06 <TrueBrain> the ?action=edit stuff is bad for caching
22:40:20 <frosch123> i kind of assumed it would be /edit/... :)
22:40:41 <frosch123> gollum did the same. everything that was not a page started with /gollum/
22:40:54 <frosch123> but since we have language codes at the top, we have a lot of freedom
22:41:09 <TrueBrain> I am still not sure making "Page" the implicit namespace for the URLs
22:41:20 <TrueBrain> but so far it seems fine
22:42:11 <TrueBrain> what-ever, it is late
22:42:30 <TrueBrain> btw, I am going to need a lookup table for en -> English, etc
22:42:34 <TrueBrain> to make some pages more pretty :)
continue to next day ⏵