Esolang talk:Community portal

From Esolang
Jump to navigation Jump to search

Archives:

Translate to other languages

I think we can't only English in this site. Can I Translate some content to other languages?

Yes, but i think we should join a Cateory, some site aren't english, we should Translate them to english. Z Ling (talk) 01:56, 12 June 2022 (UTC)
some creater hate english, so they use them mother language to write site. Z Ling (talk) 01:56, 12 June 2022 (UTC)
but we should copy a sand box to Translate to english and create hook to it,
creater can edit in the first page and push first page to sand box.

Shall we clarify how often we archive this talk page?

IMO I think we should archive when the number of topics reaches to 20. Any other ideas? Also when we got a conclusion, we should put the policy on the top of this page. --YamTokTpaFa (talk) 02:27, 12 May 2019 (UTC) i think we should save by the lines, not topics reachesZ Ling (talk) 01:48, 12 June 2022 (UTC)

Categorization policy?

Have you ever cared it? Seeing Special:UncategorizedPages, there are exactly 715 uncategorized pages at this point. For example, some pages are categorized into Category:Joke languages but not to Category:Languages. Seeing Esolang:Categorization#Languages,

Category:Languages (this should be present in all languages, as opposed to e.g. Category:Computational models)

and isn't the case above against this policy?

Also some pages are categorized to no categories yet. Therefore we should clarify categorization policy more, e.g. categorize on at least a category, as an example. --YamTokTpaFa (talk) 02:27, 12 May 2019 (UTC)

Template usages

There are many templates with no usage descriptions. We should add some. --YamTokTpaFa (talk) 02:27, 12 May 2019 (UTC)

yes, i think is a good idea. it's can make creators laarn to them.

Usage of this talk page

Shall we clarify it? --YamTokTpaFa (talk)

Existence of Special:WithoutInterwiki

Why do we still have this although we have no other language versions yet? --YamTokTpaFa (talk) 02:27, 12 May 2019 (UTC)

(idea)Introducing SyntaxHighlight?

Though this is nothing but an idea, but shall we introduce SyntaxHighlight extension? Also is it possible to add available languages by ourselves? --YamTokTpaFa (talk) 02:48, 12 May 2019 (UTC)

I'm in favour, but another perspective is that it privileges 'proper' languages. Code in Python and COBOL will look colourful, but our own brainfuck, binary combinatory logic and deadfish would not. IFcoltransG (talk) 08:38, 9 April 2020 (UTC)
SyntaxHighlight doesn't support most esolangs, but it does support brainfuck and Befunge (since it is powered by pygments). --None1 (Nope.) 00:18, 1 October 2024 (UTC)
I'm in favour. Some languages are need it like Gino. Z Ling (talk) 01:43, 12 June 2022 (UTC)
Maybe we should highlight manually. --None1 (talk) 06:01, 12 July 2024 (UTC)

Admin list?

Is there any definite list of people with special permissions? --Emerald (talk) 12:42, 2 March 2021 (UTC)

Special:ListAdmins gives a list of administrators. Changing the Group lets you view other lists of people, such as bots and bureaucrats. —User:PythonshellDebugwindow (talk) ~~~~ 13:22, 2 March 2021 (UTC)

Freenode and the future

As you may have heard, there's been happenings™ over at Freenode organizationally. I'm not qualified to explain, but you can look at some of the ex-staffer resignation letters (mniip, kline, Fuchs, many more) for at least one perspective. What it boils down to is, we might find it proper to migrate the "official" (FSVO) IRC presence somewhere else; as of this writing, based on a handful of channels I hang out on, the new libera.chat platform looks like the majority choice.

Assuming things don't settle down, here's a provisional proposal:

  • Set up shop on ##esoteric at libera. Since topic channels are first-come-first-served, I've gone ahead and registered it, out of an abundance of caution.
  • Before moving "services" (bots and so on), wait a couple of days to see where the freenode user community in general decides to go.
  • Potentially: Consider doing a "registered community" for the #esolangs (and #esolangs-*) channel namespaces. I'm of two minds about this: the kind of ad-hoc, informal community we have here, and the (low) amount of activity there is on IRC, doesn't necessarily warrant it. On the other hand, since they've bothered to set up this route for communities with no official claim to name, we might as well avail of the opportunity.

Please feel free to discuss. If there's support for the community registration, I would be using this thread as "evidence", since it's not like we as a community have anything like a governing structure or whatnot. Also, *if* we go that way, I think I'd prefer to wait a bit first, since I imagine the people involved in reviewing that kind of applications will be quite busy with more respectable communities.

--fizzie (talk) 13:27, 19 May 2021 (UTC)

I think we definitely have a good claim to a registered community. Personally, I feel that that's the right thing to do, too, for peace-of-mind if nothing else. --Taneb (talk) 16:08, 19 May 2021 (UTC)
I say if you want to leave Freenode, we should go to OFTC, they've been running an IRC network with culture similar to Freenode, I've known them for almost a decade, several open source software projects has a community there. Irc hostname is irc.oftc.net . – b_jonas 23:02, 19 May 2021 (UTC)
As we're not specifically about free and open source software, I think we're likely off-topic on OFTC according to their network policy, so we'd need to set the channel mode to +s and it would be harder for us to get help from network staff in the case of spammers or whatever. --Taneb (talk) 15:38, 20 May 2021 (UTC)
Regarding the two "obvious" networks, I don't have an unambiguous opinion. In very practical terms, I don't think the OFTC policy is likely to cause any issues, and I doubt being +s (if required) would really make much of a difference. On the other hand, it might still be nice to be an officially designated community; Libera is feeling like a closer "spiritual successor" of the freenode project; and the number of people betting on it (e.g., Wikimedia, possibly Ubuntu) makes me think it's likely to survive. On the third (gripping?) hand, it is undeniably very new, and OFTC's governance structure is a lot more transparent (which I like) than an unspecified "Swedish non-profit" (but they literally just launched, so). On balance, I'm probably still leaning towards the new thing. Not sure if strongly enough to look into doing some groundwork about it. --fizzie (talk) 21:49, 20 May 2021 (UTC)
On the topic of community registration, I had a discussion with a staffer (rationale: even if the new network doesn't pan out, a registration there won't hurt anyone), and they were positive about us claiming the #esolangs channel (and the about/esolangs/... cloak namespace, which I — can't lie — sounds pretty nice), but also keeping #esoteric set up as a forwarder for "quite some time", which I expect to mean "until a group of people come around with a plausible claim why they're more closely associated with the word". I mailed out the e-paperwork for that. --fizzie (talk) 12:33, 22 May 2021 (UTC)
Given the widely publicized (1, 2, 3, 4) freenode move to take over any #foo (primary channel) that dared to mention the alternative in its topic, by force-forwarding them to ##foo (topic channel) and clearing all registration details, I think we might as well just make this official. While we've never had any valid claim on the name #esoteric (other than by adverse possession), it feels like renaming to ##esoteric to stay there at this point would be a bit pointless. In practice all the activity seems to have already moved to #esolangs on Libera.Chat, and some of it has even been (gasp) on-topic. Unless anyone objects by EOW (or unilaterally does it before that), I'll update the community portal page to reflect that. (At the moment we've bridged the networks with a bot, as one does; will have to see how long it makes sense to keep that up.) --fizzie (talk) 11:54, 26 May 2021 (UTC)

LifeWiki links

The link to LifeWiki doesn't seem to be dead. Why is it marked "it may be dead"? -- New Army (talk) 03:54, 20 May 2021 (UTC)

IIRC LifeWiki didn't show any actual posts without being logged in for some time. But the link works now, great! --Int-e (talk) 08:45, 20 May 2021 (UTC)
On my computer it sometimes give 403 errors and sometimes the page is simply blank. ColorfulGalaxy (talk) 09:22, 20 May 2021 (UTC)
403 means forbidden. But that is your own forum page. It must be your computer's problem. --New Army (talk) 03:26, 27 May 2021 (UTC)
After my LifeWiki forum account got hacked, they deleted all my posts except for one I didn't actually post. I had to create a new account on the LifeWiki forums. ColorfulGalaxy (disambiguation) (talk) 09:15, 27 May 2021 (UTC)

Libera.Chat community and cloaks

We have just finished with the community registration for #esolangs over on Libera.Chat:

-ChanServ(ChanServ@services.libera.chat)- The #esolangs namespace is registered to the esolangs project
-ChanServ(ChanServ@services.libera.chat)- Public contacts: int-e, fizzie

Along with the community comes the about/esolangs/* cloak namespace. Most (admittedly generally more "serious") projects/communities have some sort of a policy (e.g., Ubuntu, Fedora, Wikimedia) on how to manage their namespace. What should ours be?

Here's are some potential thoughts — please discuss below:

  • Allow anyone registered on the wiki to request a cloak that's either about/esolangs/<IRC nickname> or about/esolangs/<wiki username>. Should there be any edit requirements, Wikimedia style? Possibly not; or if so, something very modest — something like, at least one nontrivial and on-topic contribution into the main namespace.
  • AIUI, cloak registration works by having one of our community contacts (at the moment: fizzie or int-e) talk to a Libera.Chat staffer and ask them to set the cloaks. To avoid causing any extra toil to people who're undoubtedly quite busy, especially now, I'd suggest collecting at least the initial chunk of interested people into a batch, rather than having one of us ask about them one by one.
  • Keep track of registration requests over the wiki somehow? Maybe just set up a (hypothetical) Esolang:IRC cloaks page, sorted into "pending", "requested" and "active" lists.
  • Dedicate about/esolangs/bot/* to IRC bots. It's not as short as some of our bots might like, but it's not too terrible in the grand scheme of thing.

--fizzie (talk) 09:47, 2 June 2021 (UTC)

I think using the wiki username is better (since if the cloak is used at all it is linked with clearly the IRC with esolangs wiki, and the IRC name is already present in the IRC anyways) (although there is the consideration, what if the wiki username is not a valid cloak?), and perhaps at least one edit should be required (but doesn't matter what it is, whether or not it is nontrivial, what namespace, etc; it just has to use the wiki account; I suppose an edit to the request page would satisfy this criteria). Perhaps another question is, is some kind of verification required? --Zzo38 (talk) 21:24, 18 June 2021 (UTC)

Add gemini:// to the recognised URI schemes

Hello, I made myself a user page, and tried to make a link to my Gemini “capsule” (as they are customarily called, basically just means “website”), and a web search and a couple of MediaWiki help pages later, I found $wgUrlProtocols.

For those unfamiliar with Gemini, you can find it’s website at https://gemini.circumlunar.space/. It’s a protocol and accompanying markup format for simpler websites than the usual HTML+CSS+JS+HTTP(S)+whatever blob; but more capable than a similar protocol, Gopher (which is in that variable by default).

It seems that MediaWiki doesn’t yet include gemini:// in this variable, which means one can’t link to resources served via Gemini, which is a shame. I, for one, would greatly appreciate if this was done, so that I could have a link directly to it; and if any resources for any esolangs are served via Gemini, it would be possible to link to them.

--OrichalcumCosmonaut (talk) (it/its or they/them please) 09:49, 14 July 2021 (UTC)

Sure, makes sense. Added it to the configuration; seems to work on your user page. --fizzie (talk) 10:20, 14 July 2021 (UTC)
Thanks! --OrichalcumCosmonaut (talk) (it/its or they/them please) 10:45, 14 July 2021 (UTC)

Wiki quality and "List of ideas" section

The existence of the page List_of_ideas always gave the Wiki an unprofessional and childish vibe to me. The same is true for unimplemented esolangs (excluding Category:Computational_models or other conceptual pages which have actual content); or at least the implemented ones could be made the default list, i.e. the one linked in the homepage. Related to this last point, there is no quality check, and I mean a very basic one. See e.g. the "content" of a page like A_very_long_language_name_that_is_very_weird_and_yeah,_this_is_your_but_its_this.

Notice that I don't have anything against joke languages. My point is that if the user doesn't bother coding the interpreter, there is no value in cluttering the Wiki with his idea.

Just my humble opinion.

Please discuss. --Andrea Calligaris (talk) 19:46, 21 November 2021 (UTC)

How would you suggest this be implemented? Ginger Industries (talk) 18:40, 29 April 2022 (UTC)

Suggestion: Bot to auto-update the language list

One of the things that language creators should do is, of course, add their language to the Language list or the Joke language list, as well as their respective categories. However, sometimes editors forget to do one of those but do do the other (i.e. add a lang to the list but not add the category). Therefore, I propose a bot which scans both lists and both categories, and keeps them synchronized. Thoughts? Ginger Industries (talk) 18:24, 29 April 2022 (UTC)

Translate to other languages

I think we can't only English in this site. Can I Translate some content to other languages?

Yes, but i think we should join a Cateory, some site aren't english, we should Translate them to english. Z Ling (talk) 01:56, 12 June 2022 (UTC)
some creater hate english, so they use them mother language to write site. Z Ling (talk) 01:56, 12 June 2022 (UTC)
but we should copy a sand box to Translate to english and create hook to it,
creater can edit in the first page and push first page to sand box.

IDK how to report bugs, nor where known wiki bugs could be already listed, so

When you visit the https://esolangs.org/wiki/Category:Usable there are no languages displayed there. If you log out and refresh the page, there are three. --(this comment by Nakilon at 9:49, 15 October 2022‎ UTC; please sign your comments with ~~~~)

Checking the three languages shown when logged out (Boner++, Golunar and LOLGraphics), none of them actually have "Category:Usable"; they all did once, but it was removed. So the logged-in empty page is the correct one. The reason for the logged-out stale page is almost certainly a longstanding known limitation of the MediaWiki file cache: it doesn't support automatically updating category pages (https://phabricator.wikimedia.org/T26575) when the member pages change. The particular situation will likely resolve itself the next time I update the MediaWiki version, because that involves refreshing all the cached pages. I should probably schedule some sort of periodic forced cache refresh for category pages, or something, but have never gotten around to it. --fizzie (talk) 10:35, 15 October 2022 (UTC)

Clarification for "Unusable for programming" category

As it stands, the current Category:Unusable for programming is a mix of languages which cannot perform computations (i.e. less than combinational logic, like AnnoyStack), languages with an Unknown computational class (like Ftack), as well as languages with the ability to perform computations but are either Turing tarpits (like ShaFuck) or are less than Turing complete (like Bipoint).

My question is, which of these types of languages belong in the category? I think that the category should only include languages which are strictly less powerful than a bounded combinational logic machine. That is, they can't do at least one of: conditional execution, arbitrary halting (if there is not conditional execution), or data manipulation. Thus, Ftack and other unknown languages do not belong. Turing complete languages of any kind should not either, nor languages which are FSMs, PDAs, or combinational. But languages like AnnoyStack, Deadfish, HQ9+, etc. should belong. stkptr (talk) 05:56, 16 April 2023 (UTC)

A good working definition is probably "the language can't write more than a finite number of essentially different programs", although there might be some debate about what an essential difference is. For example, AnnoyStack basically becomes a set of linearly increasing counters that can't interact with each other and that can't have any effect on control flow (or indeed change anything other than the program's output). --ais523 02:44, 29 July 2024 (UTC)

Abuse filter 10 unfriendly to those who want to post large programs

Recently, I wanted to post A large program, but was blocked by the abuse filter rule. That rule blocks non-admins from making large edits, but some normal users have to post large programs. --None1 (talk) 12:35, 18 December 2023 (UTC)

One of the primary purposes of the filter is to prevent people posting very large programs on the wiki, because that's not a good use of space – you can host them offsite and link to them, or come up with a compressed representation. (In particular, if a language writes mostly in unary, there isn't much purpose to actually writing programs out by repeating the same string again and again – just say what the string is and the number of repeats, which is easier both to write and to read.) --ais523 12:20, 6 June 2024 (UTC)

Who hosts (& owns) this wiki now

--PaxtonPenguin (talk) 14:02, 3 May 2024 (UTC)

User:fizzie is the current owner. --ais523 23:23, 3 May 2024 (UTC)

429 Too Many Requests

When I access this wiki recently, many scripts used by the wiki failed to load with error 429 Too Many Requests. Is it the problem with my network or the wiki itself? --None1 (talk) 15:27, 31 July 2024 (UTC)

Recently, there have been some bots making a huge number of very rapid accesses to the wiki, placing a large load on the server and triggering rate limits for the number of requests allowed from an IP range. If your IP happens to be within the range of addresses that the bots are using, then the server won't be able to distinguish your legitimate accesses from the bots' illegitimate accesses and both will end up being blocked when the bots are active (because it doesn't have enough resources to check whether each connection is legitimate or not – there are too many illegitimate connections to do to that). So the fault lies neither with your local network nor with the wiki, but with other Internet users near you. --ais523 17:00, 31 July 2024 (UTC)
To add some color to above, yes, for the last week or two there's been a pretty much constant (and increasing) stream of requests (chart) that appear to be attempting to crawl the entire wiki, including fetching all possible diffs between old page revisions, despite the wiki having a robots.txt file that disallows those URLs. While it's not a lot of traffic in an absolute sense, we also don't have a lot of capacity either, and the diff requests are relatively expensive to serve. At times, this was severely degrading the serving of normal traffic.
These crawlers do not set a distinguishable user agent (but rather use a collection of common "normal" user agents), and they do not use a single IP address range either (but a large set of regular domestic Internet service IP addresses, mostly in China and Singapore, which makes me suspect it is done through a botnet of some sort), so I was forced to rate-limit rather large chunks of address space. Since almost every request is from a different address, in order to be effective the rate-limiting has to merge all these addresses into a single pool, meaning any real users from the same ranges also suffer from it.
There's a few things I could potentially do to mitigate this, that I'll try to investigate later today. Specifically, I can try narrowing down the rate limiting just to the history and diff pages the crawlers are accessing (currently it applies to the entire /w/ prefix, which includes those scripts you mention). Since all the crawler traffic is logged-out, I might also be able to put in place some sort of heuristic to detect and exempt logged-in requests, though this is not entirely straight-forward because the rate limiting is done "outside" the wiki software. I suspect if I manage to do this, it would have to be done through inspecting the (presence of) cookies, meaning you would have to use the "keep me logged in" option to benefit from it.
(If you have the option of configuring IPv6 service, that will also avoid the rate-limiting, because I have exclusively listed IPv4 address ranges, but I understand this isn't really an option for most people.)
--fizzie (talk) 19:16, 31 July 2024 (UTC)
Assuming I configured it right, you are now exempt from rate-limiting if your requests contain the cookie esolang_wikiUserName. This should normally be set when you log in. Unfortunately this does mean you have to have cookies enabled and browse the wiki while logged in (or figure out how to manually set that cookie to any non-empty value), but it's the best I can do under the circumstances. --fizzie (talk) 22:26, 31 July 2024 (UTC)
The problem is, when I enter the link esolangs.org in the browser, it accesses the wiki over HTTP. Then the wiki redirects to HTTPS, before it can log in, which causes 429 too many requests because the wiki thinks I'm not logged in in the HTTP page and thus fails to load the HTTPS page.
I could have entered https:// before the link but that would be too annoying because who wants to type 8 more characters (Including 1 shift character)?
--None1 (Nope.) 23:07, 24 September 2024 (UTC)