Why in the world is this template so arcane?

Why does me typing {{valid SVG}} return "this is invalid"?! I'm putting in the template because it's valid. These SVG templates are so bizarre and obscure and the documentation is not helpful. Why does this template do precisely the opposite of what you expect and give you no clear instructions on how to do the thing you want it to do? —Justin (koavf)❤T☮C☺M☯ 05:02, 15 February 2024 (UTC)

This template is part of sarang's SVG maze. If you care about your brain, don't enter it… Chealer (talk) 13:15, 2 October 2024 (UTC)

Use nu validator

{{Edit request}}

— Preceding unsigned comment added by Aaron Liu (talk • contribs) 14:38, 22 November 2024 (UTC)

Per Commons:Village pump/Technical/Archive/2024/10#W3C validator has been basically down for months, the W3C validator has uptime issues. Aaron Liu (talk) 13:43, 22 November 2024 (UTC)

•  Oppose w3c validator works again (it has been for 3 weeks now), nu-validator is non-dtd. --Kontributor 2K (talk) 13:52, 22 November 2024 (UTC)

  The fact that it was down for so long means it very well may be down in the future again. Nu validator is also better because it doesn't refuse to validate certain SVG versions.
  How is it non-DTD? Aaron Liu (talk) 14:05, 22 November 2024 (UTC)

It seems to be fixed now. Nu-checker doesn't refuse to validate certain SVG versions because it's non-dtd (i.e. non-xml), it's ok for validating html inline-svg. However, a relevant option here might be to allow the use of the valid/invalid link only to logged users, for bots do follow the no-follow links and many of them don't respect robots.txt directives, in addition to the fact that many people every Earth inhabitant may click on the link for nothing, i.e. without correcting the value if necessary. --Kontributor 2K (talk) 14:18, 22 November 2024 (UTC)

Again, how is it non-DTD? It seems to automatically detect the schema to match against based on the SVG header. Aaron Liu (talk) 14:34, 22 November 2024 (UTC)

On https://validator.w3.org there's a link 'As an alternative you can also try our non-DTD-based validator' (aka Nu Html Checker). I don't know if that answers the 'How?' --Kontributor 2K (talk) 14:42, 22 November 2024 (UTC)

Well, I need to know the "how" to see how that affects the results and thus how it is relevant. Aaron Liu (talk) 14:49, 22 November 2024 (UTC)

closing edit request due to lack of consensus. Please reopen if some consensus is reached. --Jarekt (talk) 23:12, 31 December 2024 (UTC)

Under certain circumstances the DTD validator is required, one example here; as I recall, there are others, at least in topic:heraldry. though not many. --Kontributor 2K (talk) 12:51, 18 January 2025 (UTC)

It is decided, then. Aaron Liu (talk) 14:24, 18 January 2025 (UTC)

Also, sometimes the validator automatically switches to nu checker, like with this file. --Kontributor 2K (talk) 12:17, 19 January 2025 (UTC)

Default display of source code in result page

{{Edit request}}
Currently, only Large SVG files – when parameter < is specified – don't display the file source code on the target page (&doctype=Inline), while any other SVG does (&ss=1); also, the anchor #source included by default makes the page jump to the Source Listing section, which implies having to scroll-up to read the status.
It would be easy to consider that displaying the source code of a file hosted here on a separate website could be moot, but above all, currently every click on the "valid / invalid" link increases the traffic disproportionately, as each validation uses twice the amount of data transfer required, with benefits yet to be demonstrated.
Moreover, there are many weighty files (300kb, 600kb, 1Mb, 6Mb… - examples abound) that don't have the Large SVG file parameter set, so one solution would be for the module not to display by default the source code on w3.org, otherwise large or weighty files should be tracked to prevent this and reduce global traffic load; it's possible to force the source code not to be displayed for not really large files by adding o=O or o~0 (Igen/IgenCoa) (oro=n, or anything), i.e. without categorizing files in Cat:Large SVG files, but this solution would leave many files as is since it's not possible to fix them all.
So, I'm submitting an {{Edit request}} to get support on solving this issue, being aware that what I've done in the sandbox is propably not safe.
--Kontributor 2K (talk) 20:05, 14 January 2025 (UTC)

Are you sure just passing the showsource parameter causes it to request the data twice? I would think that they would only request it once, put that into cache, check the cache, and then display the same cache.

Also, you have to type out {{edit request}} with the brackets to do an actual edit request. The linked version of the template is for deactivating edit requests. I've added the template to the top of this section. Aaron Liu (talk) 14:30, 18 January 2025 (UTC)

Thank you,

As a matter of fact, if the source code is displayed, the issue is that the result page will be as heavy as the file… thus, at least twice the amount of data actually required for validation is globally transferred (u + d), example with this file with which a 1.6Mb result page is received while it's just 16Kb when the source code is not displayed; this could be considered a global waste of bandwidth, in addition to the impact on w3.org's servers and, incidentally, on our browsers' cache. The question of users with reduced bandwidth can also be raised. --Kontributor 2K (talk) 15:26, 18 January 2025 (UTC)

Makes sense. Aaron Liu (talk) 17:27, 19 January 2025 (UTC)

 Done --Kontributor 2K (talk) 12:48, 11 February 2025 (UTC)