There's been some confusion about archive.matrix.org over the last few weeks - we've written https://matrix.org/blog/2023/07/what-happened-with-the-archive/ to try to clear things up.
@matrix I think that blog post is disingenuous, but please correct me if I am misunderstanding something: Until a week ago the archive did list also rooms with public join rules and world-readable history. That is what many people were upset about. You don't seem to address that in any way. That behavior was changed by an outside contribution in https://github.com/matrix-org/matrix-public-archive/commit/1d3e930fbd57788ed60494a64835127ff97dbd53
Instead of blaming this on a misunderstanding, please be honest in your communication.
@matrix It is okay to make a mistake, even at that scale. It is okay to apologize for not having understood the feelings of the community. However, not being honest about your mistake and instead blaming others for being upset and silently changing stuff and acting as if it never happened is NOT okay (at this scale at least). It seriously damages the trust I have in the people behind the project and I have invested a LOT into Matrix, which is why this hurts even more.
@deepbluev7 "public join rules and world-readable history" is precisely what archive *should* be archiving.
There are two upset factions here: IRC communities who didn't realise a chanop had set their room world-readable, and separately Matrix communities who (correctly) picked up on the thinko that *shared* history visibility was visible (but excluded from being indexed) via the interface. We acknowledged and even credited tulir for this in the blog post if you look.
@matrix Right, I missed that sentence. I am pretty sure I searched for "shared" in the blog post, but I guess I have a history of missing parts like that. Sorry about that.
I still think that should have been more front and center, but at least you do acknowledge that the application has been behaving contrary to expectations at some point. My apologies!
@deepbluev7 the blog post was written primarily for the IRC audience, fwiw.
@matrix Which I am not, but I had to deal with the archive behind the scenes quite a lot, which is why I have a very different perspective than IRC peeps, I guess...
@matrix Hm you are contradicting yourself on who hosts the irc bridges. This post says "It is important to note that the Libera Chat and OFTC bridges hosted by the Matrix.org Foundation" but the https://matrix.org/blog/2023/07/deportalling-libera-chat/ one says it is hosted by EMS. Which one is correct?
@deepbluev7 @matrix Well thats true. Still this is somewhat confusing.
@MTRNord @matrix Can't disagree, but also most non-profit + commercial couples are affected by that. See Mozilla as an example. Who pays for what and who is in control of what without clear documentation is very hard to figure out in such cases and the Matrix Foundation doesn't really do that good of a job on that. Maybe the new website could have a section on that though!
@deepbluev7 @MTRNord The Matrix.org Foundation contracts Element to operate the bridge.
@deepbluev7 @matrix Especially since libera also says it is EMS and not matrix foundation in their posts
@matrix Am I the only one being confused here? At the top you say that archive.matrix.org is "just a client" and not an archiver or indexer, and then later on you say that some rooms weren't indexed—implying that all the others thus *were* indexed.
I think that you should rework the TL;DR section to be more clear on this.
@piegames archive.matrix.org is just a client; it doesn't archive or index data. however, given it's a webapp, it could get indexed by search engines. Originally it explicitly set <meta name="robots" content="noindex, nofollow"> on shared-history (but not world-readable) rooms so that they could be read by the interface but would not be indexed by search engines. That was a thinko, and the ability to view shared-history rooms was removed entirely.
@[email protected] looks like you need an ethics board and a pr team.