Re: [w3c/editing] Move execCommand to WICG (#184)

All of the 4 browsers we are currently dealing with represent really large organizations. Such large organizations at times mean that not everything is super well coordinated. I am not so sure all developers, project leads, spec writers, etc. in those browsers agree on everything about editing, for example.If we don't have the Editing taskforce be the one place where everything related to text editing is dealt with, there may be no such forum.

A few years ago we had such a situation. When I had joined the editing taskforce I met a lot of browser people very enthusiastic about going beyond execCommand. On the JavaScript side the production ready editors with larger programming teams behind them there was little concern about moving away away from execCommand because in practice that's what the editors had been doing anyway and unless browser makers would invest a lot of resources into fixing it, it wouldn't be used going forward anyway. So we were all in agreement when it came to editing - or so we thought.

Yet we had been getting requests from various to fix this and that about contenteditable and execCommand. We then had to explain that that spec was actually somewhere else and the editor had turned into a fulltime student and had stopped responding to emails. That caused a lot of confusion as people turned to us for things they needed to have standardized concerning editing - afterall we were the editing taskforce -- yet what they saw as the key technology was somehow not maintained. When Ben Peters left and I took the Input Events spec over, I therefore made sure to find out if we could also take on those draft specs concerned with contenteditable and execCommand - not because we were going to finish them, but so that we could have one place of discussion about text editing.

This is when we suddenly found out that there was a whole different community of browser developers, also working on related things although slightly different (I believe it was clipboard stuff). Yet their view on execCommand and inputevents was just about the opposite of what we had. And while he that that the execCommand document was a dead document, I suddenly had requests from these browser devs to please add or change this or that because it was going to ship in the next version of browser X and one had followed the practice of using the spec draft to document the browser behavior and compromises one had agreed to in that field even though the spec had no official standing in the W3C.

This is when we had a lot of very heated debates on whether Input events was reinventing the wheel or whether execCommand was dead. The good thing though was that now we had everything in just one forum. And we did negotiate a truth from that and both an immediate way forward (Input Events spec + contenteditable) yet keep execCommand around as something similar to an Irish border backstop just in case things didn't go as we had anticipated as some were - and probably still are - convinced that we will fail with creating something that can work entirely without execCommand.

Right now everyone in this discussion (but not all browser devs filing issues with us) seem to agree that execCommand is not going anywhere. I am also part of the group that that thinks this. But I think we should look at our recent past and accept that something similar could happen again if text editing is being split up into different fora again with different groups of browser developers creating different, uncoordinated APIs, all of which are half-implemented and JS editors continue to have to build their editors on a mishmash of interfaces.

I also think this situation is limited in time: We are dealing with a situation that really had gotten out of hand, but really it should not always be like that. Once we really have something that can fully replace execCommand in all browsers that works for JS editors, we may finally see that the question of possibly reviving execCommand will be much less interesting.

Disclaimer: I am reporting on historic events purely from memory, so some of this may actually have occurred somewhat differently. Most of it is probably documented in the form of email messages and github posts if someone really wants to research it.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/w3c/editing/issues/184#issuecomment-438443214