I almost feel like it’s October last year, when I pled for improvement on all fronts regarding the Kbin development strategy. Now it seems development has ceased once again and there hasn’t been chat on the matrix channels for over a week. Update: that’s two weeks now (including his blog) and over a month of no visible Development.
Update: According to https://fedidb.org/software/kbin there are a grand total of 29 active kbin servers. Of 61,489 users on all instances, 59,962 of these are on kbin.social. To those users I would like to say that should kbin.social fail, there will always be Mbin servers to fall back to.
What is going on?
We can only speculate, based on what has happened in the past. Several mind bending theories can come to mind.
Perhaps Ernest (and whatever team still exists) is continuing development in the background, not publicly sharing his work on codeberg. He may have had enough of all the criticism and wants to do it his way without interference. This may sound a bit far fetched, but he’s admitted in the past that he prefers concentrating on coding over communicating with his community.
Other theories could involve something bad involving Ernest personally, let us hope that is not the case. Ernest is a great person, nobody would wish anything bad upon him.
But in reality, speculation doesn’t change anything, we can only deal with the situation at hand.
I believe waiting for a new release comes with too many risks in the current setting. Nobody can monitor the code during progress or do any testing for themselves while development is ongoing (assuming it is). If anything, what’s coming next? Your guess is as good as mine.
A new version may well break compatibility between Kbin and Mbin. Mbin is trying to stay compatible with Kbin for as long as reasonably possible. However, staying compatible with something that is out of sight and out of one’s control is challenging, if not impossible. A break in compatibility would mean there will be no easy way to migrate to Mbin after an upgrade for Kbin users who have patiently waited for one, should they want to.
I would not want the future of my instance to be dependent of such a level of uncertainty, now or in the future.
Ernest is still posting devlogs:
It’s also not a huge mystery that what is essentially a one-man hobby project will slow down development in some periods, and speed up in others. I’m sticking with Kbin.social because I’m happy with the experience here despite some rough edges, but of course there are other platforms better suited for people who are unhappy about being beta testers of slowly developing software.
(For anyone interested, Fedia.io is the largest mbin instance!)
At least we know he’s doing alright then, that’s a relief
Well, as you stated, nobody knows.
I am more hopeful with Mbin than Kbin for now. I just hope that Ernest is doing fine.
I’m not too familiar with the drama but any time I visit kbin.social there’s some error or outage going on. Also the documentation is pretty lackluster, developing 3rd party tools for kbin is pretty much impossible.
Mbin looks way more promising, if anything because of the better docs, new features and community-driven direction it’s taking. I hope most kbin users jump ship.
There was a bunch of downtime over the holidays. I assumed it was due to the Lemmy update withholding posts for several days and then flooding kbin. It doesn’t bother me in the slightest and I have multiple fediverse accounts.
If Mbin wants to keep compatibility, Maybe another fork should be considered. The new fork could go on to faster development while Mbin stays behind.
Kbin was a fantastic project and its developer deserves all our respect.
Unfortunately Kbin suffered a perfect storm that overwhelmed him: -he was born at the wrong time; indeed: he was born at a very right time! That is, just before migration from Reddit reached its peak
- it was too immature and too good at the same time: users saw how good it was and dived into it en masse
- as if that wasn’t enough, even more users flocked to Kbin due to a well-known Fediverse popularizer, blinded by ideological hatred towards the Lemmy developers, who recommended it to everyone as the best alternative to Reddit, saying something lie: immature software like Kbin CAN NEVER be a better alternative than mature software like Lemmy (oh yes, congratulations FediTips: you actively contributed to sinking the Kbin boat!)
- the developer was alone, a very normal circumstance for a project in development; but this led to an absurd overload and a totally destabilizing overexposure!
- finally, the usual dynamic of free software has struck again: herds and herds of ignorant, selfish, resentful people lacking any empathy began to DEMAND improvements, modifications and corrections to a software that (I repeat once again) was still immature, without giving ANYTHING, and when I say NOTHING I mean without financing the project.
I completely understand that the developer went into burnout! The meltdown of lone developers is a very frequent phenomenon in free software… These pressures are almost impossible to bear: to bear them you need to have balls of steel, decent financial resources, many true friends and a pathological level of enthusiasm and self-esteem to resist such strong emotional pressure.
If it helps, you can just use something else like Lemmy and still access all kbin content, that’s what I’m currently doing.
right, to his point; if users or admins are having issues with kbin, mbin is in active community development for accessing lemmy instance content.
if users or admins are having issue, they can jump to an mbin instance for a near identical experience.
it just sucks because we care about the source of k/mbin in Ernest, and he deserves our consideration.
Then don’t use it.
My point exactly, to instance operators, instead switch to Mbin, now that it’s still possible
I would not want the future of my instance to be dependent of such a level of uncertainty, now or in the future.
You can always go to a big provider like Facebook. They’ll surely won’t have issues with your documented whining on Kbin.
he has already propose a solution in the fork to mbin, which is active and prospering. your comment is unhelpful.
