User talk:Kyliotic

Json metadata for the extension

Hey, @Kyliotic I noticed you were experimenting with adding metadata to pages for use with the extension. I just wanted to make sure you are aware that the plan is to add either the Semantic mediawiki or Cargo extension to achieve this easily in the future. Kostas (talk) 23:03, 2 March 2025 (UTC)

Btw, a better way to test would probably be to use your own user page in order to not pollute the history of the page. Kostas (talk) 23:28, 2 March 2025 (UTC)

@Kostas, Hopefully, this works. Captcha is so broken. Oh is that so? I was going to keep it simple and just hide JSON data within the article. And then a browser extension can just getElementsByClassName() and retrieve any needed data. Also, your phrasing implies that there already is a browser extension, is that true? Kyliotic (talk) 23:28, 2 March 2025 (UTC)

@Kyliotic For some reason it did not but I saw your comments on the bug report page. That page should probably be moved as the Consumer Action Taskforce namespace is protected as a whole... Though that is the right place for such page.Anyway:

• Captcha is broken with discussion tools. That is a known bug with the extensions and I am not sure when it will be fixed, hopefully we tune the confirmed/autoconfirmed user process to allow people to skip them as soon as possible.
• The <br> tags are restricted purspose this has to do with what bots like to use to spam on the wiki. This will likely not change.

On the extension conversation:

• Hidden fields like you describe is what the cargo extension would basically provide but it would allow querring by field and using SQL syntax.
• Yes work on the extension has started here https://github.com/WayneKeenan/ClintonCAT coordinating with the devs there would probably be for the best. I am sure they would be open to discussing this.

There is some conversation on a thred on the on-topic chanell on the discord server as well.

The talk page guidelines might help with the captcha issues in the meantime Kostas (talk) 21:16, 3 March 2025 (UTC)

@Kostas, Interesting info. I have to manually edit the discussion source in order to reply. Perhaps that also breaks notifications/pings. Cargo might be helpful with metadata read/write, but I have never used it.

• So far, with simple embedded JSON, my extension can already look up if a website URL has been added to a wiki lookup table/database, and if is, it pulls embedded JSON from the wiki's product/organization articles. I just need to make GUI and then the basic concept is functional.
• I only glanced at the ClintonCAT GitHub and it seems to actually read/search through all the wiki articles. To me, that seems like overkill. Having machine-readable metadata is so much more efficient and conveniant and sure, maybe Cargo may help with that.
• But also, in my opinion, the nice thing about having the metadata embedded in the article source is that anyone may edit and fix metadata (once a proper format is established of course). So when someone creates a new article, someone can easily copy/paste a metadata template, edit it, and have the changes update live to the extension.
• Maybe even some automation to automatically generate extension metadata (likely only possible with a MediaWiki extension)
• Does Cargo achieve those points? I genuinely don't know, but just some things to consider. Kyliotic (talk) 06:04, 4 March 2025 (UTC)

@Kyliotic this time the ping worked. Maybe I missed it last time or I had the page open in another tab and that messed it up. You should take a look at the page I linked, I think it will help (eg ~~~~ automatically signs mesaages for you) I am aware how you are replying, I wrote the FAQ entry on it a month ago. Actually disregard the source editor stuff, let me just confirm your account ao you can just skip the captchas and reply using the ui.

• The metadata used with Cargo/Semantic media wiki is meant to be used like you descibe by the extension. Also they will be user editable as well.
• The clinton cat extension did initially work like you describe(querries to the api for every page) but that has shifted. I would suggest working with them as they have a basic ui (as I understand it), there are a couple of people already working on it and it has been the one Louis has been linking. Work has been stalled for a bit (mainly waiting on the extension to be added) but I think combining resources would help here. Maybe make a PR to share your idea on what the metadata format should contain? Or even the whole flow you currently use.
• Automating this is how we are planning to do it. It should be possible through the API and a scriot but I dont have anything concrete yet.

Also if you join the discrors communication would probably be easier. If you prefer not to I am happy to use talk pages and using github for the extension should also be fine, but it would help get on the same page faster I think. Kostas (talk) 09:04, 4 March 2025 (UTC)