Some years ago we used to post weekly development updates to let the community know what we are working on. For some reason we stopped posting these updates, but now we want to continue giving you information every two weeks about the recent development progress. This should allow average users to keep up with development, without reading Github comments or knowing how to program.
We’ve been working towards a v0.19.0
release of Lemmy, which will include several breaking API changes. Once this is ready, we’ll post the these changes in dev spaces, and give app developers several weeks to support the new changes.
This week @nutomic finished implementing the block instance feature for users. It allows users to block entire instances, so that all communities from those instances will be hidden on the frontpage. Posts or comments from users of blocked instances in other communities are unaffected. He also reworked the 2-Factor-Authentication implementation, with a two-step process to enable 2FA which prevents locking yourself out. Additionally he is reworking the API authentication to be more ergonomic by using headers and cookies. Finally he is adding a feature for users to import/export community follows, bocklists and profile settings.
@dessalines is currently implementing a redesign of the join-lemmy.org website. He is also keeping the lemmy-js-client updated with the latest backend changes 1 2 3.
@phiresky optimized the way pagination is implemented. He is also fixing problems with federation workers which are causing test failures and performance problems in the development branch. These problems were introduced during a complex rewrite of the federation queue which was recently finished, and is thought to allow Lemmy federation to scale to the size of Reddit.
@SleeplessOne1917 is implementing remote follow functionality, which makes it easy to follow communities from your home instance while browsing other instances. He is also fixing problems with the way deleted and removed comments are handled .
@codyro and @ticoombs have been making improvements to lemmy-ansible, including externalizing the pict-rs configuration, adding support for AlmaLinux/RHEL, cleaning up the configuration, as well as versioning the deploys. These will make deploying and installing Lemmy much easier.
Support development
@dessalines and @nutomic are working full-time on Lemmy to integrate community contributions, fix bugs, optimize performance and much more. This work is funded exclusively through donations.
If you like using Lemmy, and want to make sure that we will always be available to work full time building it, consider donating to support its development. Recurring donations are ideal because they allow for long-term planning. But also one-time donations of any amount help us.
- Liberapay (preferred option)
- Open Collective
- Patreon
- Cryptocurrency (scroll to bottom of page)
Thank you, all of that sounds like very good news.
Thanks to all the people working on this.
Looking at the Liberapay, it’s shocking Lemmy only gets $332 a month to fund this entire social media. Compare that to the totally “grass roots” normal NFT growth where they immediately had millions poured into them from venture capitalist. As it turns out, actual grass roots social media without profit incentive isn’t profitable! And that’s how you know we’re on the right path.
Great work. Thank you for the update Nutomic.
We’ve been working towards a v0.19.0 release of Lemmy, which will include several breaking API changes. Once this is ready, we’ll post the these changes in dev spaces, and give app developers several weeks to support the new changes.
Thank you for the update and the heads up.
How will this be announced, and what specifically does several weeks mean? Since Lemmy goes beyond Mobile Apps to all kinds of systems including moderation tools, auto-purgers, bots, CSAM, auto-subscribers, searchers, etc, breaking changes to the API can have far-reaching impacts.
Could something be set up specifically for breaking-change announcements where participants could be alerted? Even just a Breaking Changes issue that could be followed would work nicely.
Thank you again.
When we are ready to publish the first release candidate, we will make a post that lists all the breaking changes. You can follow !announcements@lemmy.ml via rss reader to get notified about it. We will also share it in different Matrix chats, and I’m sure it will get upvoted to the frontpage as well.
Do you have any sort of timeline for when 0.19 or the release candidates will become available? I only wonder because I’m eager to check out some of the new features that have been mentioned here and on Github
See here: https://lemmy.ml/post/5711722
For the truly breaking changes like API auth, pagination, and TOTP, is there a reason you don’t roll the deprecation like most software?
I.E. 0.19 supports both methods, and 0.20 deprecates the old one? This way developers aren’t caught off guard if they’re not following (which will get worse as time goes on), and allows development using official releases vs RCs.
For instance, if I want to update my app now, I have to release it with an RC library. If there was a version between deprecation, I could update at any point during the official 0.19 lifespan.
In case of pagination both old and new variants are supported in 0.19 (see my reply to fmstrat above). TOTP is currently broken so it wouldnt make sense to keep supporting the old version. In case of auth it would be possible to keep backwards compat, but keep in mind that we are only two fulltime devs with tons of other things to work on. If we spend a lot of time on this, it means less time for other important tasks. Besides you can support both 0.19 auth and 0.20 auth at the same time by sending auth as param and header/cookie.
Thanks for the response, I actually put another comment in after I started diving in and saw that the pagination/auth were overlapping, which was great news, it just didn’t come across clearly to me in the write-up for some reason. Thank you for structuring things this way.
Actually, as I start to do updates, it looks like the JS library still has
page
along withpage_cursor
(and the auth can be run in parallel) which seems to be great news! Does this mean I can continue to usepage
for pagination until .19 officially releases and rolls out, and then switch over?0.19 supports both
page
andpage_cursor
. We will probably get rid ofpage
in 0.20.
This week @nutomic finished implementing the block instance feature for users.
Thank youuuuuu!
Amazing job as always!
Thanks for the update! I think it’s a good idea to get the word out there more often on what’s being worked on and how progress is going at a regular interval.
If these are API-breaking changes, shouldn’t you bump the major version? https://semver.org/
I think you’re forgetting this:
- Major version zero (0.y.z) is for initial development. Anything MAY change at any time. The public API SHOULD NOT be considered stable.
Ha. Well yea, devs need to learn to commit at some point I guess.
Fair enough.
From an end user perspective, it feels like it’s operationally stable, though I don’t know about developmentally stable. Maybe it’s worth a 1.0 release soon. Lots of people are running it in production now.
Before 1.0 we definitely need to do an API cleanup, the paths are a real mess. However that will require lots of breaking changes so Im not sure when we can do it.
see also: perpetual beta
Yea. I feel like once it can scale pretty well and has been for a bit of time that would be a good opportunity to release 1.0. But another major factor here is that the backing and sustainability of the project is still up in the air, so the flexibility of breaking changes is maybe rather valuable for a while.
deleted by creator
Excellent work! Keep it up!
Posts or comments from users of blocked instances in other communities are unaffected
Hate to say it but this makes that feature a little insufficient
Great news, thanks for the update!
Donated! Keep up the good work
I spend enough time on Lemmy, enjoying the spoils of your hard work that I couldn’t not donate. Like others have noticed, the weekly amount at it’s current level ($350/w) is embarrassing.
IF YOU ARE READING THIS, GO DONATE.
did you forget to mention the new Scaled sort? or is it not gonna make it into v0.19.0?
The post only includes pull requests which were active last week. Scaled sort was already completed earlier. It will definitely go into the next release, like every merged PR.
Thank you for your work. I love seeing that my money actually goes to something great