Switching to Mastodon - see you soon!
Mastodon will soon be supporting ActivityPub . So I'm going to switch there for good now. If everything works as expected I'd be in touch with everybody still sticking with pump.io anyway. But for now I'm not maintaining two timelines anymore.
My new home is https://social.tchncs.de/@mray - follow me there, or wait for ActicityPub here. See you soon!
soloojos shared this.
Waiting for the ursupation
Since it looks like Mastodon is *really* making steps to support ActivityPub, I think I'll wait for the ultimate social ursupation sitting on a hairy elephant more and more. If you haven't tried it out yet - give it a try!
The only thing that bugs me is that there is currently *NO* standalone client for the Linux desktop that I can get to run.
Clacke moved to social.heldscal.la and datamost.com shared this.
Mastodon seems to have attracted a bigger audience than GNU social and pump.io combined. Great times ahead if everybody actualyl embraces ActivityPub soon!
Clacke moved to social.heldscal.la and datamost.com, ostfriesenmärz, Charles Stanhope, Ron K. Jeffries and 2 others likes this.
A little bit of spring-cleaning reveals that the majority of people I follow have parted long time ago...
Then there are people on GNU social / Mastodon that I don't need to follow here.
In the long run I'll probably join them.Show all 7 replies
>> JanKusanagi @i:
“BTW @mray, Dianara has a beautiful (if I may say so) "private follows/unfollows" setting that avoids flooding all your followers with hundreds of unfollows ;)
Also, I see you're unfollowing recently active people, just sayin'.”
Most of them do have a GNU social or Mastodon account.
Sorry for flooding! - I wasn't aware of this, nor of any way that prevents it.
I should mention I'm over at Mastodon, too: https://social.tchncs.de/@mray
Mastodon seems to plan to support ActivityPub, too - just as pump.io.
Maybe I'll "wait" there for the glory of ActivityPub to come true ...
I feel pitty when visiting https://savannah.gnu.org/ .
There probably are many projects hosted there that deserve a better representation.
It is like a website-worm-hole that lets you witness the web from the past with your modern browser and current content.
The idea that this should be the "...central point for development, maintenance and distribution of official GNU software..." makes me shudder.
I know I never spent longer than 20 seconds on that page just because of the walls of text and information overload and can only imagine what it means to really *work* with it. Maybe I'm getting it all wrong? I mean Github feels superior to Gitlab from a UX perspective, but savanna makes me wonder if this is for real. To me this page spells in clear words: "GO AWAY - BIT ROT - CONTAGIOUS - HERE BE DEAD PROJECTS".
What bugs me the most is that I actually care about quite some projects that are "hosted?" there and I wonder what drives the fsf to dogfood this strange thing instead of using something else.
Dana likes this.Show all 5 replies
You are completely right to bring this up in its defense. Because it does not only appear to be – it actually *IS* very old. I also see how it is too much to ask from the 2? people keeping it up and running to also keep updating and actively developing it.
By coincidence I read on reddit that there is a gitlab mirror of savannah: https://gitlab.com/GNU/ ...
“[...] and I wonder what drives the fsf to dogfood this strange thing instead of using something else. [...]”
IIRC they've discussed this themselves, and I think part of the reasoning was that, even if awkward to use and old-looking, it had several functionalities they still couldn't replicate with other systems (yet). Plus it works without JS or any other fancy thing, so compatibility with, let's say "uncommon setups and workflows" is higher.
But don't take my word for it, it's just something I vaguely recall. I might be mixing things up.Non-JS compatibility is a biggie.
Interest in ActivityPub
Looking at GNU-Social, friendica, diaspora mastodon and what not - are there any projects that openly announced they are interested in supporting ActivityPub?
Clacke moved to social.heldscal.la and datamost.com shared this.
2016-11-18T11:51:53Z via Choqok To: Public
Somewhat depressing but interesting fact about "users" https://www.nngroup.com/articles/computer-skill-levels/
uıɐɾ ʞ ʇɐɯɐs likes this.
Debian switching to GUIX
How realistic would that actually be? I mean there is no law that dicates .deb has to stay forever. Just pondering what it would actually take to do such a switch.
Ben Sturmfels likes this.Show all 5 replies@Christopher Allan Webber I agree. That does seem a bit like moving a mountain. So much infrastructure (social, technical, and knowledge) built around a packaging formats. But your own efforts demonstrate that there could be a way forward if a distribution was motivated.
I was referring to moving the project over. I know it sounds grotesque, but it opens the question of: Is Debain bound to grow old and die with this one package manager - or will it renew that part of its OS some day eventually? If so - what would it take, what would it be?
Pumpa crashes - always
Don't know why or how. It just started crashing on me every time I try to post. Even removing chache files and configuration files didn't change a thing. :/
Unless it's a new bug that I don't know about, it should be fixed now in the release I made yesterday: https://pump.saz.im/sazius/note/7Dri4WFlR9evFmtZBKlIfg.
If that's not the case please give me some more info to help debug it, for example using debug mode: http://pumpa.branchable.com/debug_mode/.
trying to click on the "Login" button in pump.io's web interface seems broken for me right now.
Why use the web interface at all? - because Pumpa can't seem to post stuff after follwing a persons thread.
pump.io feels unready :'(Show all 6 replies
It is beyond me what matrix is trying to achieve. Even after reading their "FAQ".
Claes Wallin (韋嘉誠) shared this.Show all 8 repliesTheir rationale visavi XMPP is here:
HTTP-nativeness, identity model, richer mandatory minimum protocol and a model around message history as state, rather than live message passing.
Sounds like a lot of NIH, but given the results they've had, it seems to make sense. I'm happy that people are trying things, and as they've got this far in less than two years, they seem pretty capable.
I read about that rationale there but some things speak for themselves.
They don't explain the difference, they just ramble about shortcomings of XMPP -WTF!
they point out that XMPP vs. matrix seems to bring out the worst in people.... irony?
I wonder if matrix has to offer anything comparable to OMEMO/Signal encryption.
Claes Wallin (韋嘉誠) likes this.