IRC logs for #openttd on OFTC at 2020-11-08
⏴ go to previous day
07:19:10 *** sla_ro|master has joined #openttd
08:27:19 *** andythenorth has joined #openttd
08:44:19 *** jottyfan has joined #openttd
09:07:25 *** andythenorth has joined #openttd
09:45:12 *** andythenorth has joined #openttd
09:52:18 *** andythenorth_ has joined #openttd
09:52:34 *** innocenat_ has joined #openttd
09:59:38 *** mindlesstux has joined #openttd
10:00:28 *** reldred has joined #openttd
10:18:22 <TrueBrain> glad we had this conversation :D
10:19:34 *** frosch123 has joined #openttd
10:20:25 <TrueBrain> lol ... I found a nice bug in TrueWiki ... \r\n -> \n is missing :D
10:20:50 <TrueBrain> frosch123: request for migration: add a newline at end of file for all files :D
10:21:16 <frosch123> what does \r break?
10:21:45 <TrueBrain> the only reason I found out, honestly :P
10:21:55 <frosch123> to templates change when they gain a \n at the end?
10:22:21 <TrueBrain> newlines are pretty much ignored, but there can be some odd cases ..
10:22:26 <TrueBrain> all templates should be in <includeonly> blocks
10:22:31 <TrueBrain> but fair point, nevermind my request
10:22:43 <frosch123> i can do it for non-templates :)
10:22:52 <TrueBrain> nah; either for all or for none, honestly
10:23:05 <frosch123> ok, so just remove \r
10:23:19 <TrueBrain> replace("\r\n", "\n").replace("\r", "\n")
10:23:23 <TrueBrain> is what I now do on editing
10:23:36 <TrueBrain> \r\n is Windows, only \r is old Mac
10:23:52 <TrueBrain> well, removing \r is better
10:24:02 <TrueBrain> content.replace("\r", "")
10:24:10 <TrueBrain> now I don't have a diff anymore if I hit "save" :D
10:25:35 <TrueBrain> git-operations are not quick :D Saving takes a bit more time now :P
10:26:13 <frosch123> yep, when i migrate the whole history, it fill take hours :)
10:26:26 <Eddi|zuHause> content.replace("\r", "") <-- can't git do this for you with eol-style?
10:26:49 <TrueBrain> sure; but git-backend is optional
10:28:47 <TrueBrain> hmm .. how do I see the commiter of a git commit ..
10:33:30 <TrueBrain> how/where can I improve on commit messages?
10:33:39 <TrueBrain> (the logged in user was "test")
10:39:08 *** WormnestAndroid has quit IRC
10:39:21 *** WormnestAndroid has joined #openttd
10:47:39 *** HerzogDeXtEr has joined #openttd
10:55:10 *** gelignite has joined #openttd
11:00:28 *** mindlesstux has joined #openttd
11:01:00 <TrueBrain> I don't like spaces in page names, it makes commit messages look weird :P
11:01:10 <TrueBrain> but quoting them is even weirderderder
11:01:42 *** Progman has joined #openttd
11:02:18 <TrueBrain> meh; hopefully frosch123 has more inspiration than I have :P But "git commit" part is done .. now the "git push" part :D
11:14:59 <andythenorth_> I would have quoted them, but eh, YMMV
11:15:15 * andythenorth_ learning about boats
11:36:06 *** Progman has joined #openttd
11:37:21 <andythenorth_> ships are inconvenient
11:37:26 <andythenorth_> IRL is not arranged how I think it should be
11:37:40 <TrueBrain> frosch123: I was talking about the commit message itself
11:37:44 <TrueBrain> that is a single line
11:37:48 <TrueBrain> or .. hmm, no, that works
11:37:58 <TrueBrain> not sure what the "git show" is showing?
11:38:10 <frosch123> yeah, ignore that, it's weird
11:38:45 <frosch123> i guess git status is mostly the same as what you already have
11:38:58 <TrueBrain> yup, but a bit better, I would say
11:39:02 <TrueBrain> we just cannot use our normal Add: etc
11:39:03 <frosch123> just "->" instead of "to"
11:39:48 <TrueBrain> a change is just "modified", right?
11:40:26 <TrueBrain> I only made it "new page", instead of "new file" :P
11:42:18 <TrueBrain> okay .. so that leaves 1 item on my TODO for v1 :)
11:46:11 <frosch123> hmm, there are no \r in my conversion
11:47:58 <frosch123> yeah, i have to reject your request as "not reproducible" :)
11:49:57 <TrueBrain> I did not request you to remove \r, for the careful reader ;)
11:50:21 <TrueBrain> My request was about end of file newline :p
11:50:36 <TrueBrain> My code on the other hand did need a \r fix :D
11:51:51 <TrueBrain> It is however good to know mediawiki does something right :D
11:53:48 <frosch123> ok, "User:Pystro/Template:Link Translation" is way too complicated to migrate, but luckily there are only 13 usages :)
11:54:23 <TrueBrain> hmm .. I think it is easiest if I disallow ":" in page names btw
11:54:32 <TrueBrain> I can add support for it, but it makes the code complex ..
11:55:36 <frosch123> forbidden: ":|", non-printables, leading ".", leading/trailing ws ?
11:56:32 <TrueBrain> aa..bb is also not allowed
11:56:33 <frosch123> ok, for parts separated by /
11:56:42 <TrueBrain> it doesn't split the /
11:56:47 <TrueBrain> it just scans the whole path
11:57:05 <frosch123> but "/" should also not allow neighbouring whitespace
11:57:42 <TrueBrain> so many exceptions! :P
12:00:57 <TrueBrain> okay, so the current rules:
12:01:11 <TrueBrain> in FULLPAGENAME, no "..", "|", or ":"
12:01:25 <TrueBrain> in every part of the FULLPAGENAME (so between the /), no starting or ending with " " or "."
12:01:31 <TrueBrain> FULLPAGENAME cannot end with /
12:01:46 <TrueBrain> pretty sure people will find other annoying things to put in names :P
12:02:52 <frosch123> there is probably some js injection for ie6
12:03:16 <TrueBrain> many more I can blacklist, it seems :P
12:03:47 <TrueBrain> # < > [ ] | { } _ will be blacklisted too :)
12:05:13 <frosch123> allow "_", we have a proper " "
12:06:11 <TrueBrain> but what use-case would _ be a valid use?
12:06:21 <TrueBrain> sounds that the user meant " " :P
12:06:57 <Eddi|zuHause> a page about an identifier?
12:07:07 <frosch123> also, textareas are really broken in my ff version
12:08:15 <TrueBrain> hmm ... a title with the name "&" is .. not doing what I expect it to do :D
12:08:46 <TrueBrain> I think I am missing some url escaping ..
12:08:51 <TrueBrain> and/or html escaping
12:08:53 <TrueBrain> that will be fun :D
12:09:31 <TrueBrain> same with %20 in names, also fun :D
12:09:38 <frosch123> i disabled spell checking now, it blocked the gui thread and disrupted editing
12:09:41 <TrueBrain> I would like to ban & and %, but there are legit uses for it ...
12:09:51 <TrueBrain> frosch123: ... lol :D
12:10:33 <Eddi|zuHause> for example, if someone was writing a wiki page about the setting "newgrf_developer", why would you assume the page title should be "newgrf developer"?
12:12:20 <Eddi|zuHause> on the other hand, if you allow "_", then you get ambiguity in the lookup
12:12:40 <TrueBrain> "A title can normally contain the character %. However it cannot contain % followed by two hexadecimal digits (which would cause it to be converted to a single character, by percent-encoding)." <- this .... wuth?
12:13:10 <TrueBrain> Eddi|zuHause: I am only willing to allow "_", if we refuse a page named "A_B" and "A B" to both exist
12:13:18 <TrueBrain> but it is just easier to ban "_", honestly
12:13:51 <frosch123> TrueBrain: only because mediawiki authors did not know how to escape, you do not need to copy that
12:14:02 <TrueBrain> frosch123: I fully agree :)
12:14:15 <TrueBrain> well, I don't know how to escape either, clearly, but that is something I can hopefully fix :D
12:14:38 <TrueBrain> are there any characters not allowed by git?
12:15:16 <TrueBrain> * and " are not allowed on NTFS
12:15:20 <frosch123> depends whether you want to allow checkout on windows
12:15:31 <frosch123> backslash is difficult on windows
12:15:46 <TrueBrain> should we just ban those 3 chars?
12:15:49 <frosch123> LPT and CON are difficult
12:15:59 <Eddi|zuHause> i think windows bans a few more characters
12:16:19 <Eddi|zuHause> but i don't remember which ones
12:16:36 <frosch123> TrueBrain: \ did not work on mediawiki either. some user actually created a page containing \, but when trying to view it something converts it to /, and then reports 404
12:16:40 <TrueBrain> that is why I looked it up and came with those missing 3 Eddi|zuHause ..
12:16:57 <TrueBrain> okay, so I banned \\, * and " too
12:17:18 <Eddi|zuHause> i'm pretty sure ":" was also banned last time i tried
12:17:35 <TrueBrain> Eddi|zuHause: yes ........ which we talked about 5 minutes earlier to already ban
12:17:57 <TrueBrain> frosch123: hmm ... both NTFS and mediawiki doesn't mention it
12:18:19 <TrueBrain> it does need url-escaping, ofc
12:19:07 <TrueBrain> FAT32 does ban ?, it seems
12:19:53 <Eddi|zuHause> i think FAT32 used ? to denote deleted files
12:20:16 <Eddi|zuHause> i could misremember that, though
12:20:30 <TrueBrain> ah, no, it is also reserved on NTFS
12:21:01 <TrueBrain> frosch123: I think I will escape ?, honestly
12:21:34 <michi_cc> I think NTFS itself allows more characters than what userspace windows API will let through.
12:24:15 <frosch123> <> without & is weird, but okay
12:24:39 <TrueBrain> we can also just escape < > honestly
12:24:58 <TrueBrain> ah, no, <> is not allowed by NTFS
12:25:01 <TrueBrain> I just change the comment :D
12:25:04 <frosch123> i still don't get why you put such a focus on "..", when the problem applies to "." after "/"
12:25:32 <TrueBrain> frosch123: because any URL you try to access with ".." anywhere, will give you a 404
12:25:35 <frosch123> do you want to allow hidden pages? :)
12:25:58 <TrueBrain> and again, there is ALSO a filter on starting/ending with " " and "." per split on /
12:26:11 <TrueBrain> so a page cannot be named ".bla"
12:26:52 <Eddi|zuHause> so no wiki page on "~/.openttd" :p
12:28:01 <TrueBrain> for the "..", code just becomes a lot easier if you ban access to any URL containing ".."
12:30:32 <TrueBrain> let me know if I missed something or if you know a cleaner way to write that function :D
12:33:59 <frosch123> i think there were cases with a trailing "."
12:34:24 <frosch123> ah, yes, the "T.I.M." engine
12:34:37 <TrueBrain> the others are typos, I guess
12:34:47 <TrueBrain> The parameters of the xUSSR set.
12:34:54 <TrueBrain> that makes little sense to have a dot there in the title :)
12:35:04 <frosch123> yes, they are just translations of whole sentences :)
12:35:14 <frosch123> but "CS-Inc." is the same as "T.I.M."
12:35:17 <TrueBrain> but okay, only lstrip on dot it is :)
12:35:44 <frosch123> again, same evolution in the conversion scripts :p
12:38:09 <frosch123> TrueBrain: i think leading / and // are also invalid, so the split loop could check for part.strip() == ""
12:40:30 <TrueBrain> now I really need to rewrite this function, but okay
12:40:33 <TrueBrain> pushed a new version for now
12:41:02 <TrueBrain> I split off the "ending with /" and "starting with /", as that requires a different message to make clear to the user what is going on
12:44:20 <TrueBrain> and now it is a lot more readable :P
12:45:24 <TrueBrain> right, now lunch, after lunch ... htmlencode and urlescape ... that will be FUN!
12:46:55 <frosch123> (" ",) <- what emoticon is that? :)
12:47:47 <TrueBrain> why doesn't GitHub cancel actions of an earlier PR, I wonder ..
12:47:53 <TrueBrain> or why is that not configurable
12:48:03 <TrueBrain> it is now running 10+ CodeQL checks on outdated commits
12:49:28 <frosch123> maybe they think they charge for each one
12:58:30 *** Progman has joined #openttd
13:08:04 <andythenorth_> achieved that already
13:10:00 * andythenorth_ deletes and adds ships
13:10:04 <frosch123> some daily quets are pretty grindy
13:10:25 <andythenorth_> I have some annoying reminders about missions
13:10:36 <andythenorth_> and at the end there's no level up, it's just a silly badge
13:10:42 <andythenorth_> then do it all again tomorrow
13:21:04 <TrueBrain> bah, it is difficult to detect if the user made no change, yet hit the "save page" button :P
13:40:27 <TrueBrain> oef, supporting % in page names is a lot more tricky than it sounds ...
13:40:34 <TrueBrain> I get why mediawiki just bans %20 etc
13:40:50 <TrueBrain> it depends on where a pagename is used what you have to do with it
13:41:02 <TrueBrain> as in a href, you have to do different things,than anywhere else
14:00:13 <frosch123> is it not putting all hrefs through urlencode?
14:01:31 <TrueBrain> but finding all the places to do so, that is the issue :D
14:03:54 <frosch123> 7 in wikitexthtml, 7 in truewiki. what am i missing?
14:05:28 <TrueBrain> check templates/ in truewiki :)
14:06:01 <frosch123> is there a urlencode parser function?
14:07:13 <frosch123> i mean for wikitext. {{urlencode|/edit/{{FULLPAGENAME}}}}
14:07:29 <frosch123> something to use in templates/
14:08:38 <TrueBrain> so you now understand why I say it is tricky :)
14:08:56 <TrueBrain> you have to deal with HTML escaping and URL quoting
14:09:07 <TrueBrain> but I am slowly working my way through it, don't worry
14:09:11 <TrueBrain> I just get why mediawiki bans it
14:09:26 <frosch123> ok, may python be with you :)
14:17:40 <TrueBrain> a page named "test&%20test and & this % with" seem to be the best test case out there :P
14:24:04 <frosch123> add " and ', unless you banned both
14:35:09 <TrueBrain> hmm .. "used on page" only detects templates
14:35:13 <TrueBrain> not pages linking to you
14:35:19 <TrueBrain> so you can rename a page, even so other pages link to you
14:35:46 <TrueBrain> not even indexed, which pages link to you
14:36:00 <frosch123> mediawiki add a "(linked)" or "(transcluded)" after every what-links-here item
14:36:32 <TrueBrain> so you can now rename pages that we most likely don't want to have renamable :)
14:37:12 <frosch123> adding redirects is easier, but messier :)
14:38:02 <TrueBrain> I would prefer if we can avoid any redirect system, honestly
14:39:01 <TrueBrain> new page: en/test&%20test and & this % with ' and the end
14:39:53 <TrueBrain> okay, now to test images ..
14:55:37 <TrueBrain> right ... longest I took to fix any bug in TrueWiki so far :P
15:03:45 <TrueBrain> pretty sure I forgot some src/href somewhere :)
15:04:06 <TrueBrain> but even the GitHub history URLs should work :)
15:06:56 <frosch123> haha, let's hope there is no githubencode :)
15:07:00 <FLHerne> Thoughts on a "Tooling: " or so commit prefix?
15:07:19 <FLHerne> For changes to actions, flake8 rules, blah
15:07:58 <FLHerne> It seems a bit odd to be writing Add/Fix/Codechange etc. for things that aren't actually part of the deployed codebase
15:08:06 <TrueBrain> what we do in other repos, is still tell if it is an Add or Fix or what-ever, and add a [Actions] after the :
15:08:22 <TrueBrain> because you are fixing the Actions, for example
15:08:28 <FLHerne> NML is the same, I was just wondering if we should keep doing that :p
15:08:39 <TrueBrain> well, in your reasoning, the commits should not show up, basically :)
15:08:54 <TrueBrain> but git log has little to do with the deployed codebase tbh
15:09:27 <TrueBrain> but adapting a more modern commit messages would also solve it .. but we had that debate a few months ago :P
15:09:31 <TrueBrain> were a lot of NAYs :)
15:09:48 <TrueBrain> as I would call it a chore() commit :P
15:11:09 <TrueBrain> lot of projects switched to a commit message format which allows tooling to create changelogs etc
15:11:35 <TrueBrain> they use the "type(scope): subject" format
15:11:58 <TrueBrain> one of the types is "build", or "ci", or "docs"
15:12:16 <TrueBrain> it allows you to make commits as feat(), meaning they should be highlighted in the changelog
15:13:24 <TrueBrain> basically, the type splits the commits in "development" and "production" changes
15:13:43 <TrueBrain> what you referred to as "deployed codebase", I guess
15:14:17 <FLHerne> Nope, I don't like that format :p
15:14:27 <TrueBrain> I am in no way surprised ;)
15:14:37 <TrueBrain> but this is what many many projects are currently using :)
15:15:28 <TrueBrain> getting to change #openttd in any of this is like trying to move a building ....... :P
15:18:34 <FLHerne> TrueBrain: What you need is a set of those modular transporter thingies with hundreds of little wheels, then
15:18:45 <FLHerne> Perhaps andy can draw some for you
15:20:17 <frosch123> what that filesize thingie always there?
15:22:19 <TrueBrain> well, "always" is such a weird word to use :P
15:22:27 <TrueBrain> but I added it when I created the File pages
15:22:57 <TrueBrain> you don't like it, or? :)
15:22:58 <frosch123> didn't notice yet :)
15:23:39 <TrueBrain> mediawiki had it, and I liked it enough to add it
15:24:03 <TrueBrain> although I am using the correct SI units
15:24:09 <TrueBrain> mediawiki has a weird opinion about those :P
15:24:45 <TrueBrain> although ... hmm, no, I pretty sure I divided by 1024 ... still wrong :P
15:24:59 <TrueBrain> well, mediawiki is more wrong :P
15:25:08 <TrueBrain> no clue what KB is ..
15:25:11 <TrueBrain> it is not SI at least :)
15:25:26 <TrueBrain> isn't it KiB btw? Let me check
15:26:07 <frosch123> so it's not just "insert 'i' where it looks funny"
15:26:38 <frosch123> TrueBrain: mediawiki uses JEDEC obviously :p
15:27:10 <TrueBrain> hahaha, yeah .. right :P
15:27:22 <TrueBrain> owh,it is not even SI, it is IEC
15:28:40 <TrueBrain> right, fix is being deployed :)
15:28:43 <TrueBrain> that is how we roll :)
15:29:05 <TrueBrain> ignoring the "git push", I think this is ready to be tested
15:29:10 <TrueBrain> see if people can break it :P
15:57:12 *** longtomjr has joined #openttd
16:20:53 <frosch123> editing templates without proper preview is so frustrating on mediawiki... i have such high hopes for truewiki :)
16:38:42 <Eddi|zuHause> wait... it's not called "nowiki"? :p
16:45:49 <FLHerne> WIKI Interactive Knowledge Interface?
16:48:57 <FLHerne> OpenMW? (wait, that's taken)
17:04:12 <andythenorth_> FLHerne can you find any dedicated ships for carrying plastic pellets, grain, etc? :P
17:04:30 <andythenorth_> I can only find pneumatic-discharge cement carriers
17:07:07 <andythenorth_> is that a regular bulker?
17:07:21 <longtomjr> dunno, reading the wikipedia page now
17:07:43 <longtomjr> Owned by the Canadian ~Weed~ Wheat board
17:09:24 <longtomjr> I find plastic pellets linked to container ships
17:12:38 <longtomjr> This process may be carried far enough to cause the growth of moulds and bacteria and, in the case of grain, to cause sprouting. Ship malted wiskey
17:12:53 <longtomjr> (that sentance were meant to be in quotes)
17:13:27 <longtomjr> andythenorth_, here is a infographic that you can use to see a 3d model of the Equinox class
17:13:38 <longtomjr> it will take some time to load since webarchive
17:14:23 <longtomjr> Looks like the Equinox is used for Grain and Iron Ore from some articles
17:16:00 <longtomjr> Seems like it does grain one way, and the ore on the return
17:18:19 <Eddi|zuHause> longtomjr: sounds like a regular bulker, then
17:19:53 <longtomjr> Hmm, seems like big bulk ships does have hatchcovers pretty commonly
17:20:35 <andythenorth_> mini-bulkers sometimes don't
17:20:39 <longtomjr> it might just be the barges that does not.
17:20:40 <andythenorth_> if used on rivers and lakes
17:21:51 <longtomjr> Are you looking to split up bulk carriers between cargo lines?
17:22:43 <andythenorth_> might do covered bulker for sheltered cargos
17:23:29 <longtomjr> Maybe do a set of small bulkers, where there are covered and uncovered veriaties, and then have the big ones all be the same?
17:25:37 <longtomjr> Then you have a split on the lower end of the capacities, but a merged line for the big ones?
17:25:42 <longtomjr> Or is that too confusing?
17:44:22 *** Wormnest has joined #openttd
17:45:08 <andythenorth_> very definitely covered
17:49:22 <frosch123> what... 556, did not expect that
17:50:08 <frosch123> changing 2 templates halfed the number of errors :p
17:53:33 <frosch123> it's so random. sometimes i change something and think "this will reduce a lot", and then it's like -50 errors. and then you change something because it annoys you, and it changes a lot
18:04:27 <TrueBrain> The shitshow we call wikitext :p
18:10:35 *** glx is now known as Guest5307
18:10:35 *** ChanServ sets mode: +v glx_
18:31:45 *** arikover has joined #openttd
18:51:39 <DorpsGek_III> - Update: Translations from eints (by translators)
19:30:07 *** andythenorth has joined #openttd
19:43:49 <DorpsGek_III> [OpenTTD/master-server-web] TrueBrain opened pull request #16: Fix: make sorting of servers a bit more robust https://git.io/JkfQ7
19:44:20 <TrueBrain> to proof people don't notify us when things break, turns out there were over 150 error events triggered by a server of which we failed to parse the version correctly .... and sentry didn't have alerts setup on that project, for what-ever reason :P
19:44:25 <TrueBrain> but okay, this should fix that :D
19:45:26 <TrueBrain> all other repos do have alerts setup
19:45:32 <TrueBrain> Sentry can be a bit annoying in configuring projects :D
19:48:10 *** glx is now known as Guest5316
19:52:52 <frosch123> TrueBrain: there are some File: pages that link to LICENSE
19:52:56 <frosch123> what to do with that?
19:53:07 <frosch123> is there a url that works? or should i move the license?
19:54:39 <TrueBrain> well, we should have a license in the root for GitHub. But that should be in Markdown, as that means GitHub can render it
19:54:56 <TrueBrain> and either we make an exception for License to be on /License, or we move it into Page/en/
19:55:01 <TrueBrain> but I am afraid people will try to translate it
19:56:09 <frosch123> (i got there via githubs "license" link)
19:56:27 <TrueBrain> wuth, GitHub can render wikitext?
19:56:42 <frosch123> same as github wiki
19:56:48 <frosch123> but it's the same broken wikicloth
19:56:56 <TrueBrain> so I can make this a special URL, sure
20:01:41 *** lastmikoi has joined #openttd
20:05:18 <frosch123> i like this andy game
20:07:01 <andythenorth> it's a good game
20:09:08 <andythenorth> I don't think we've played 'newgrf spec' enough recently :P
20:10:02 <TrueBrain> hmm .. you can now only not link to it with [[License]] or something
20:10:57 <TrueBrain> well, you somewhat can, but it is not seen as a valid page, I guess .. I can fix that
20:13:21 <TrueBrain> okay, frosch123 , I assumed it would be [[License]] . If so, that will work :)
20:13:36 <frosch123> yes, i already pushed that
20:14:02 <frosch123> TrueBrain: isnumeric() looks really wrong, but isdigit() is not much better
20:14:12 <frosch123> possibly isdigit() && isascii() may work
20:14:52 <TrueBrain> it is a shitshow left or right
20:14:55 <TrueBrain> does it matter, I wondered
20:15:21 <TrueBrain> isdecimal() might be better, but .. meh
20:17:16 <DorpsGek_III> [OpenTTD/master-server-web] TrueBrain updated pull request #16: Fix: make sorting of servers a bit more robust https://git.io/JkfQ7
20:17:59 <TrueBrain> isdecimal() works fine for this usecase :)
20:18:27 <frosch123> i think the last return is missing an item
20:18:29 <TrueBrain> (as int() on that works :P)
20:18:36 <frosch123> it return 6 items now, right?
20:18:58 <TrueBrain> frosch123: you are correct!
20:18:59 <DorpsGek_III> [OpenTTD/master-server-web] TrueBrain updated pull request #16: Fix: make sorting of servers a bit more robust https://git.io/JkfQ7
20:20:22 <frosch123> be careful, you may be identified as spambot
20:20:34 <TrueBrain> frosch123: pushed /License to TrueWiki
20:20:53 <TrueBrain> should solve a few more on your list :D
20:22:00 <DorpsGek_III> [OpenTTD/master-server-web] frosch123 approved pull request #16: Fix: make sorting of servers a bit more robust https://git.io/Jkfdc
20:23:32 *** argoneus has joined #openttd
20:30:04 <DorpsGek_III> [OpenTTD/master-server-web] TrueBrain merged pull request #16: Fix: make sorting of servers a bit more robust https://git.io/JkfQ7
20:34:33 <TrueBrain> right, that should prevent the server list from crashing :D
20:48:15 *** DasPoseidon has joined #openttd
20:52:46 <FLHerne> (should be 'done' for review now)
21:04:53 <LordAro> have to do actual changes now :p
21:06:04 <FLHerne> Yeah, but now I can do that without fear of rebase hell :F
21:06:38 *** sla_ro|master has joined #openttd
21:07:40 <FLHerne> Oh, the one more mega-change I wanted was to use f-strings
21:08:15 <frosch123> f-strings are such a mis-feature :/
21:09:23 <frosch123> but tb also uses them, so i have to wait for the next cycle of "mixing code and formatting is bad"
21:10:43 <frosch123> old-print, c-printf, c++-iostream, python-print, python-f-strings, what's next?
21:11:45 <Xaroth> fstrings improves readability a lot
21:13:37 <LordAro> though i will occasionally use format strings for the more complicated ones
21:13:45 *** jottyfan has joined #openttd
21:13:49 <LordAro> traditional format strings, that is
21:13:51 <frosch123> i am just amazed that c++20 finally gained a pythonic print method. and now python goes to make the same mistakes as c++iostreams did before :)
21:14:31 <frosch123> they switched places :p
21:17:22 <LordAro> i'm not sure they're really equivalent
21:18:07 *** tokai|noir has joined #openttd
21:18:07 *** ChanServ sets mode: +v tokai|noir
21:21:17 <frosch123> it's more obvious when you do not take raw strings, but something like sqalchemy, where you need to preprocess/escape parameters
21:21:59 <frosch123> c++ and java are full of sql injections, because they just do "WHERE a='" + b + "'..."
21:22:14 <frosch123> sqalchemy is amazing by providing print-like methods to compose them
21:22:33 <LordAro> i'm not sure .format() vs fstring changes that much
21:22:33 <frosch123> f-strings are the opposite. i expect more code-injection issues with them
21:29:11 <dwfreed> You should be using any kind of formatting with SQL
21:29:59 <dwfreed> but f-strings are so much nicer to work with when composing strings that don't go to SQL
21:36:46 *** WormnestAndroid has quit IRC
21:37:00 *** WormnestAndroid has joined #openttd
22:13:39 *** WormnestAndroid has quit IRC
22:13:48 *** WormnestAndroid has joined #openttd
22:19:09 *** jottyfan has joined #openttd
22:50:59 <TrueBrain> SQL is the problem, not however you replace strings tbfh
22:51:40 <TrueBrain> SQL language is by concept an insecure language.. lucky enough, more and more alternatives are in the make :)
22:54:35 <andythenorth> silly SQL designers
22:54:50 <andythenorth> they should have put 'unsecured user input from the global population' into the spec
22:55:15 <andythenorth> clearly incompetent
22:55:16 <TrueBrain> Happens with old languages :)
22:55:31 <andythenorth> it's only 46 years old
22:55:40 <andythenorth> they should have done more scoping and requirements
22:55:40 <TrueBrain> Alternatives are only a bit slow getting there ....
22:59:39 <frosch123> next round is editing the wiki again :/
23:00:16 <frosch123> sounds like a thing for tomorrow
23:03:47 *** glx is now known as Guest5323
23:09:44 *** cyberjunkie[m] has quit IRC
23:10:20 *** TinoDidriksen has joined #openttd
23:10:51 *** TinoDidriksen is now known as Guest5335
23:16:07 *** robert[m] has joined #openttd
23:27:45 *** Guest5335 is now known as TinoDidriksen
23:36:19 *** grossing has joined #openttd
23:45:27 <andythenorth> 'pseudo-realistic' :P
23:51:29 *** cyberjunkie[m] has joined #openttd
23:52:43 *** m1cr0man has joined #openttd
continue to next day ⏵
