As it is now, Mlem also supports the ancient v1 of the API. Pretty much the only instance that still uses this version is Hexbear. All other instances use v3.
Dropping support for v1 would mean much less maintenance on my part, and would mean that updates would come out faster and would be less buggy.
On the other hand, Hexbear is probably the largest single instance.
Even now, Mlem doesn’t have the same level of support for both API versions; in v1, for example, Mlem doesn’t show user avatars (or any avatars for that matter). So overall, v1 of the API is much more limited, and not properly tested.
Is there anyone who uses Hexbear? Should Mlem have so much more code to support just one instance? Please share your thoughts!
You must log in or register to comment.
Drop v1 but leave a single alert that says sorry but the server you’re on hasn’t updated to v3 yet.
this sounds like a sensible approach
I agree that v1 should be dropped, even if some instances haven’t upgraded, they should try and stay current with the standard, and dropping v1 does provide an incentive. I agree with the other comment that an alert about how your server doesn’t support v3 is good.
My vote is to drop v3. At some stage Hexbear will upgrade to v3 and all the effort to maintain v1 will be wasted.
I’m using Lenny.ML, for what it matters.
Luckily lemmy.ml uses v3, like the vast majority of instances
Has v1 gone out of development for years? I’m down for dropping it, maintenance is a burden
Hello, Hexbear dev here, we’re going to move back to upstream Lemmy and stop using the v1 API soon, so no problems on our end with removing v1 API support.
Alright, great to hear! Thank you for confirming 😊
Drop it like it’s cold
Thank everyone for your ideas! I dropped support for v1 in the latest version and you now get this alert if you try to add a v1 instance:
