Manifest v3 migration timeline and implications

[Jon Saft]

We are about to break ground on upgrading our Chrome Extension Manifest from v2 to v3. I've looked all over this forum and others, but I still can't seem to find any specific timing for the v3 migration timeline outside of vague references to "Google plans to support v2 for 1 year after v3 reaches stability."

As far as I can tell, that timeline puts the migration deadline somewhere around January 2022. Can anyone confirm that timeline? If so, is there a specific date in January? Additionally, after that date passes, what happens to Chrome Extensions that have not yet migrated?

Any additional information here would be very helpful.

Thanks!

[Cuyler Stuwe]

That’s the most specific communication from within Google that I’m aware of.

Feels a bit rough and rushed with so many major problems (e.g. persistent state) unresolved.

Maybe we’ll get a more specific update sometime.

--
You received this message because you are subscribed to the Google Groups "Chromium Extensions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-extens...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-extensions/60fb47b2-3a1c-4a6c-aa72-7815b1d8e91bn%40chromium.org.

[Shu]

I hope it's not January! Is mv3 stable already? I check regularly and see a lot of bugs and problems so I still wait a stable version to start the migration, but seems they have a lot of problems with some features :/

[Cuyler Stuwe]

According to many people who regularly work with Google Cloud, it’s a sort of Google tradition to deprecate old APIs while their replacements are still incomplete, buggy, and/or undocumented. 🤷‍♂️

--

You received this message because you are subscribed to the Google Groups "Chromium Extensions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-extens...@chromium.org.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-extensions/38dd5020-6aec-4bae-98c4-035ddea6db70n%40chromium.org.

[Robbi]

The more we continue to test their buggy APIs and to report the related issues, the more they will soon deprecate the manifest V2

in favor of V3.

Furthermore because it is likely that after two\three years the V3 will be replaced by a V4 (and so on)

and we dumb programmers will have to adapt our code again.

So let's stop doing their job for free.

Let's the Chrome team calmly does the job that it is up to him.

We are not in hurry

[wOxxOm]

ManifestV3 is nowhere near being stable, it still has glaring omissions and lots of bugs in its core. Advanced extensions like Tampermonkey, Violentmonkey, Stylus can't even start the migration because the necessary API aren't implemented. If MV2 will be deprecated/removed in January, 2022, I think a lot of people will just stop using Chrome.

[Evan Carothers]

Yeah I agree many things are far from finished and stable (including, importantly, the permissions - see HERE). I've personally seen all kinds of wonkiness with new APIs, even with a dead-simple mv3 proof-of-concept I've built to test some new features. This feels like a HUGE ball drop by the google team - we are in this limbo state where we don't know when MV3 will viably work to build against, and when MV2 will go away and are supposed to make product/planning decisions based on this, which is obviously impossible. 

[PhistucK]

From previous experience as well as guesstimates by Simeon, manifest version 2 will stay for a year after manifest version 3 is completed. I think a year is quite long and you should be able to make product/planning decisions based on that...

☆PhistucK

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-extensions/01e4cc68-c2b0-4c6f-8a65-b44bcdc1feb4n%40chromium.org.

[Evan Carothers]

The issue is that the chrome team said one year "after v3 is released in stable branch" which happened in Jan 2021 - however the APIs aren't complete/stable. So if we go based on what google initially said, mv2 will stop being accepted in the CWS in Jan 2022. MV2 -> MV3 has massive changes for many developers so organizations are weighing weather they should commit ongoing feature development to mv2 extensions when they might (read: will) have to rearchitect their entire extensions at some point in the future. If we had clear line of sight into the mv2 switchover, and when all mv3 features would be complete, we would be able to actually make these planning decisions.

[PhistucK]

I am sure there will be an announcement of the timeline no matter what and I doubt the announcement will happen less than a year before manifest version 2 is gone.

☆PhistucK

[apell...@copper.com]

That strikes me as wishful thinking... If that's what Google intends, then they could just say that and put a lot of minds at ease. "We will announce the start of the deprecation timeline and then you have one year." Actually they did say that, and then Chrome 88 released and started that timeline in January 2021. Just because they've scrubbed the announcement from the docs doesn't mean that the clock has stopped ticking from their perspective.

[hrg...@gmail.com]

I haven't even heard a word about how MV3 is going to support user scripts.

The Extension's team have mentioned a number of times through official channels that they plan to support user scripts in MV3. But still no news about it.

The last thing I heard was that they are still designing the mechanism to support this. Or maybe I misinterpreted and they have yet to start the design of such a mechanism.

Whatever the case, I find it difficult that they will announce any concrete timeline before MV3 has the minimum set of features that allows everybody to start the migration. I would be quite unreasonable that they did that.

[Ioannis Charalampidis]

> I haven't even heard a word about how MV3 is going to support user scripts.

Just want to say +1 on this, and to put a personal opinion on this matter, I think that disallowing custom code execution overall sounds like a half-thought idea.

Apart from pissing-off thousands of developers it isn't really contributing towards any kind of "security improvements".

Remember that JS is still a dynamic language. Even without `eval`, I can still write my own domain-specific language that I can download from any resource, or even create a full WASM emulator that can run WASM bytecode without the need of `eval` (as a matter of fact once MV3 is enforce, I can totally see this being a community-driven project in a matter of months).

Therefore apart from making the actual code execution slower and a dozen times more annoying, the "security" implications remains exactly the same...