# I am the Watcher. I am your guide through this vast new twtiverse.
#
# Usage:
# https://watcher.sour.is/api/plain/users View list of users and latest twt date.
# https://watcher.sour.is/api/plain/twt View all twts.
# https://watcher.sour.is/api/plain/mentions?uri=:uri View all mentions for uri.
# https://watcher.sour.is/api/plain/conv/:hash View all twts for a conversation subject.
#
# Options:
# uri Filter to show a specific users twts.
# offset Start index for quey.
# limit Count of items to return (going back in time).
#
# twt range = 1 60515
# self = https://watcher.sour.is?uri=https://twtxt.net/user/prologic/twtxt.txt&offset=60291
# next = https://watcher.sour.is?uri=https://twtxt.net/user/prologic/twtxt.txt&offset=60391
# prev = https://watcher.sour.is?uri=https://twtxt.net/user/prologic/twtxt.txt&offset=60191
@bender I don't no. I mostly just use whole themes. And my facourite it AMOLED
@doesnm.p.psf.lt It was always intended to have both Yarn.social and Salty.im integrate together. Yes. This includes having a set of specifications that anyone can write clients to.
@javivf Generally speaking if it has been reviewed, discussed and merged, then we accept it as a standard to the set of specs we support. However we _might_ want to document this process and set some guidelines about this to be clear π€£ We've been fairly lax/lose here and I _think_ that's okay given teh size of our community π
@bender Ahhh ha!!! π€£ I'm too dumb to have thought of that π€£
@bender Do we still want this feature? π§
@bender Would they even accept a server from me? π§π
@eapl.me This is one of my concerns too. The moment you post publicly ciphertext, you open yourself up for future attacks on the ciphertext, which you really want to avoid if you can. If you have a read of the Salty.im Spec you'll note we went to great lengths to protect the user's privacy as well as their identity and make it incredibly hard to guess at inboxes. It's still a WIP, but I'd love to see it progressed even further -- I truly feel strongly about a purely decentralised messaging ecosystem π
I think I would encourage anyone in this community is to care less about supporting "legacy clients" and focus more on value-add whilst balancing the burden of client authors -- which have very precious little "spare time" π€£
I do _think_ integrating things like Salty.im _might_ actually be a good idea. I can also see a future where we integrate other things like todo.txt and calendar.txt. I'd even love to see decentralised forms of "plain text" voting too.
@andros I don't see any "fighting" here. This is just good experimentation. Unfortunately there hasn't really been enough time or effort by other "client authors" yet, me especially as I've been super busy with ya' know my "day job" that pays the bills and refactoring yarnd to use a new and shiny and much better SqliteCache π€£ -- I certainly don't think your efforts are wasted at all. I would however like @doesnm.p.psf.lt encourage you to look at the work we've done as a community (_which was also driven out of the Yarn.social / Twtxt community years back_).
@andros Ahh I see π
> @prologic Yes, it is a security hole. All dm-echo messages are readable. I intend it to be a debugging tool. Maybe I can include a warning message. If many of you see that it is a serious problem, I can remove the links.
@andros your feed is spitting out dupes? π§ 
@andros I _honestly_ think IRC is better anyway π€£
@movq Damn that's a sexy pidgeon π€£
@bender I _think_ mentions are fixed π€£
This is nuts. Lemme eat dinner first (_pizza on the way!_) and I'll fix this utter nonsense π€£
Fascinating read on the emerging Model Context Protocol β a new standard for integrating LLMs with agents and tools.
@andros doesn't this defeat the point of public cryptography?
Dam the search here is sooo good now π
@bender Now you have me thinking, which is better? π€
@bender I will figure this out soonβ’ and fix, it's annoying the fuck out of me π€£
I personally really like the news minimalist (_fuck it mentions are kind of broken atm here in the UI :/_) feed myself, really good quality, very high signal π
@andros It's good advice π
I try to do this myself (_please call me out if you ever catch me NOT doing this π€£_)_
@andros Q: Why would you publish the private key? Hmmm π§
Cool. That's fixed! π₯³ I _believe_ we're now syncing to 6 peers again now. Hopefully with similar behavior as before π€
Oops, I _think_ this pod (twtxt.net) just sync ~1k missing root twts with god only knows which peers π€¦ββοΈ I forgot a couple of important key things:
- Only coverage with a subset of peers
- Only converge with trusted peers
Fuck me π€£ Ooops. Sorry!~
Nice! π
@andros Kind of a cool idea actually! π I'll follow and see what it's like, thanks! π
Broadcast Positioning System, as an alternative to GPS. Very cool π
How's that? Please refresh and see if that's fixed? π
@bender There is only one commit that I can think of that _might_ be the cause here. Shall I revert and redeploy? π€
@bender Seems fine here on my
iPhone? π§ 
Yeah same order of magnitude π No relation mice other than the recent study that precisely measured the number of cells and connections in 1 cubic mm of brain tissue.
I was trying to optimize the SQL query used for the Compact FrontPage (_anonymous view for Discovery when the Admin/Operator chooses "one twt per feed"_).
@movq Pretty cool song I agree βοΈ
@lyse Just needed to update the version of the tool I packaged as an OCI image π€£
There are now two (_recentish_) quotes I really like these days:
> The smartest person in the room is not the one with all the answersβitβs the one whoβs brave enough to ask the dumb questions
and
> The kindest person in the room is often the smartest
Based on a recent study of the brains of mice I estimated the human brain to have 200B cells/neurons and 50,000T connections. We have several orders of magnitude to go before we reach that kind of scale with these fucking stupid Big LLMs π€£ And the best part of all? π§ It is estimated that the human brain only consumes the equivalent of 5 Watts of power !!! π€£π€£π€£
@bender You will be pleased to know that yarnd now only consumes ~60-80MB of memory depending on load π€£ And bugger all CPU π
~
@bender Did you see the Singaporian president's speech on this whole nonsense?! π (trade wards, tariffs, free markets, etc)
@bender Technically no but we'll talk more about this π
@bender It's a bug in the UI for sure. The hash is the primary key.
@bender No, haven't figured out the reproducer yet π§
@bender I hate vendor lock-in too π€£
Ordering issue is fixed π₯³
@bender Let's just optimize/fix those annoyances later on once I've finished pagination. Then I'll merge this branch into main.
@bender Same size they always were.
@bender Hmmm I can't figure out how you did this π€£
Like this? In the post box at the bottom?
@bender Did you mean to mention me? π§
@bendereapl.me@eapl.me A few years ago or decades ago? π€£
I _believe_ the bug has been fixed π₯³
@abucci Apologies, the basic summary is as follows:
- Decided to rewrite the cache backend.
- It will now be a SQLite backend going forward.
- I'm planning on no data migration.
@david yeah @movq and I discovered its a bug in lextwt last night π’ We'll fix it as soon as @xuu can π€£
Anyway. this was a good use for search btw. I couldn't find my Twt, so I just quickly searched for it, snap, bingo I found it in a snap! π«°
@bender Oh fuck you're absolutely right!!! π€£ I only posted this 13h ago π€¦ββοΈ
@aelaraji Yeah I'm in the process of rewriting (_incrementally_) the cache storage backend. It's now been live for at least a week now and pagination and peering are the last things left to do π€
@aelaraji Ahh this is coming back 403 Forbidden here btw π
Yes you were right. It was technically a failure. It is very difficult to reconcile the differences between a distributed network and one that's truly decentralized.
@kate it will if we continue to support the community on the ecosystem π€£
@kate @eldersnake @abucci -- I've already spoken to @xuu on IRC about this, but the new SqliteCache backend I'm working on here, what are your thoughts regarding mgirations from old MemoryCache (_which is now gone in the codebase in this branch_). Do you care to migrate at all, or just let the pod re-fetch all feeds? π€
@lyse I _think_ we found a bug in the lextwt parser actually π
@lyse @movq I mean I have a camper trailer here out the front so we probably have plenty of room for a Yarn party in real life π€£