@Catrope how would this project proposal fit into the plans with Vue?
With a very superficial estimate, assuming I have as much free time as in the last month, I would be ready with a prototype in a month and with a beta in 1-2 months more.

For an actual full rewrite to Vue.js, we should look into CdxDialog and split off its focus capture and background overlay into a component that we can use for the fullscreen component. Then with that, we can build a vue 'app' and replace all the other components. This is something I have been thinking about doing, as some of the rework to make it mobile compatible is going to be so invasive, that it might just be better to rewrite it in vue, with modern JS, modern CSS.

But that will essentially require reimplementing the whole thing, which is going to be one heck of an effort. It's probably better to keep looking at cutting up existing code a bit more first, simply to have a more manageable base.

Change #1048862 had a related patch set uploaded (by Simon04; author: Simon04):

[mediawiki/extensions/MultimediaViewer@master] wip Migrate MultimediaViewer to Vue

https://gerrit.wikimedia.org/r/1048862

Hi @Jdlrobson and @TheDJ, I've drafted a patch to get an idea how the Vue.js future of MultimediaViewer could look like. It's far from complete and lacks plenty of animations and interactions. Also, it should use the native Vue.js Codex components, eventually. What do you think? Is this (draft of) Vue.js migration worth pursuing?

@simon04 I'm personally in favour and think it will make the code more maintainable, easier to read and easier to adapt for more situations.

I will note that there is a team that is looking into making the images zoomable (T77151) and I'm not entirely sure what their timelines are, but it might be wise to talk to that team if we set out on this.

Thanks @simon04 for the patch! I think eventually this is what we want to get to, but I'd say consolidating mobile and desktop would be needed to be done first (e.g. reducing the bloat).

I would worry that rewriting and rolling out a Vue version would require creating a third version of MMV and rolling that out gradually which may slow down that goal and would mean maintaining 3 versions until that happens. On mobile we do not even use Vue.js right now in article views so there are unknown challenges ahead there. Switching to Vue will also likely increase the size of MMV. I think for that reason it could be important to have some team in Wikimedia Foundation sponsoring that work and making sure it gets seen through to completion.

In T249431#9950113, @Jdlrobson wrote:

I would worry that rewriting and rolling out a Vue version would require creating a third version of MMV and rolling that out gradually which may slow down that goal and would mean maintaining 3 versions until that happens. On mobile we do not even use Vue.js right now in article views so there are unknown challenges ahead there. Switching to Vue will also likely increase the size of MMV. I think for that reason it could be important to have some team in Wikimedia Foundation sponsoring that work and making sure it gets seen through to completion.

So generally I agree... BUT, I'm equally worried that in order to do a merge with mobile first, we are going to write a whole lot of code, and do stuff that Vue would could deliver for us and at the some time be more readable and maintainable. By the time it is then done, i wouldn't be surprised if Vue is going be present on mobile after all and it was wasted effort.

You can already see that the code to handle the buttons etc, which was quite cumbersome to implement. The other team apparently wants to add zooming and that too would then have to be reimplemented. I see it a bit as one of those "why spend twice the effort and time, when we can just wait a little and do it right in one go"-situations.