<isd>
So I took a whack at updating etherpad. Dear god, the minification pipeline is horrifying. kentonv, I'm sorry you went through that.
<JacobWeisz[m]>
lol
<JacobWeisz[m]>
I suspect the benefits of updating Etherpad are probably worth it right now, they had a really active 2020.
<isd>
I got the basics working, but I did it by starting with a fresh port, and so far I've only cherry-picked the patch that redirects / -> /p/main, so it looks like a one-pad instance.
<isd>
So, it doesn't do any of the other sandstorm integration stuff
<isd>
It's also way slower
<isd>
And not immediately obviously more featureful, though maybe there's stuff I didn't see.
<JacobWeisz[m]>
The usernames for multiple users generally work, but that's most of the extent of the integrations I think.
<isd>
There are a whole bunch of patches on the Sandstorm branch, and I haven't worked out what all of them are about yet. Some of them seem like generic improvements that would make sense to upstream (and I noticed one of them had actually been made upstream), others are fixes & integration
xet7 has joined #sandstorm
<JacobWeisz[m]>
Kenton has a selection of Etherpad plugins included too, since people can't add their own.
<isd>
Which I have not yet attempted.
<JacobWeisz[m]>
I also think there was some hefty data migration stuff because of changes in Etherpad since the first grain.
<isd>
Looks like there's also integration with Sandstorm's notifications system.
<JacobWeisz[m]>
Is there? Wow. Never seen it in practice.
<JacobWeisz[m]>
In the ideal world I imagine these integrations should be Etherpad plugins.
<isd>
Anyway, this is turning int A Project, so I might come back to it later...
<isd>
decern. You really want to be versioning the patches and always have a set of patches that applies cleanly and makes it clear what we've changed, without having to also look at what we used to change.
<isd>
I think it will be more maintainable in the long term to maintain the patches in "unapplied" form; I've noticed with the pattern of forking a repo and adding .sandstorm, it can be a bit fiddly to figure out at a high level what's different from upstream, because you have upstream changes intermingled with our own, and things that are logically updates to the patches we've applied (in accordance to upstream changes) are hard to
<isd>
Thankfully git-format-patch/git-am make this somewhat managable.
<JacobWeisz[m]>
Yeah, especially early ports are a bit messy.
griff_ has joined #sandstorm
<isd>
kentonv: do we control the 'sandstorm' npm account, or is that some rando?
<isd>
Doesn't look like its ever been used.
<isd>
I'm thinking it might make sense to claim an org name so we can publish officially endorsed packages and tell people to install '@sandstorm-io/foo-package'. Might make sense to do the same for capnproto.
<isd>
Well, I created sandstorm-io and sent you an invite. If nothing else it will avoid squatters
xet7 has joined #sandstorm
griff_ has quit [Quit: griff_]
dustyweb has quit [Remote host closed the connection]
dustyweb has joined #sandstorm
<Jassu>
options for richer collaborative formating, like Google Docs among others.
<Jassu>
Ian Denhardt: Yeah, the Etherpad has evolved into more minimalistic direction from where it was earlier. It was trying to be kind of a Rich Text Format -ish thing back in the days. But the main problem with it was that it really broke nearly all copy&paste things. Now it actually has waaay less formatting options, but pasting to it seems to work 100% of the time, which totally is an improvement. Besides, there are plenty of
<Jassu>
The current Etherpad seems to do "just plain text, nothing more" in very well working way.
<isd>
Plugins for video conferencing aside :P
<isd>
That's certainly not the impression I would have gotten from the current website.
<isd>
I didn't immediately notice anything missing either, to be fair
<isd>
all: patches welcome, if anyone feels like helping get the etherpad package updated.
griff_ has joined #sandstorm
<Jassu>
Oh, true... it seems the pad that I've been using usually just dropped some plugins off. Probably by accident. :D
<Jassu>
Yeah, it's now what you get with "basic" config.
griff_ has quit [Ping timeout: 260 seconds]
<JacobWeisz[m]>
I know our Etherpad package has whatever plugins Kenton thought was handy. I think the idea of keeping features in plugins instead of cluttering the main codebase is nice.