Please fix this.  Showing a close button on tabs makes tab management VERY prone to accidental closures.  This needs to be fixed.

Despite trying to take extra care, I still kept on getting caught out by the close buttons on the tabs. As Chrome is too frustrating to use, I've switched away to Vivaldi. Vivaldi supports many of the extensions I was using in Chrome.

Does Undo Close Tab (ctrl-shift-t) not restore the typing?  The goal here is to make undoing easy, rather than make closing hard.

There are reasons we can't always restore things correctly, but in general we should try to fix that problem first.  We've intentionally decided in the past to leave close buttons on inactive tabs when the tabs are larger than a certain minimum width, and we've spent a lot of time over the years looking at click data and re-debating this.

No, it doesn't. It's been some time since I've switched away, so I've had to go back to a few sites to check. I've found that various contact forms on ebay.co.uk as well as forums such as hexus.net and forums.autosport.com will not restore typed text for a tab restored via Ctrl-Shift-T.

While it would be nice if was possible to restore data in this way, it does seem more like trying to boil the ocean rather than fix the underlying usability problem.

While click data must be taken into account in decision-making, I'd expect it has to be interpreted in the context of product positioning and the range of user capabilities. It's not ideal to have mainstream-oriented products be unforgiving of people with less than perfect hand-eye co-ordination. I understand that time has been spent on looking at this, but with the current compromise I have no option but to not use this product, and this does not show up in click data.

Regardless of whether we make changes in this area, please do file bugs for the specific cases you mention of form data not being restored, ideally with specific URLs and repro steps.  We should aim to fix those.

You are right that it's difficult to solve that problem in general, and that allowing users to mis-close a tab is a usability problem.  However, not allowing users to close tabs they want to is also a usability problem.  There's no easy answer here; it's not a case of simply fixing a bug, it's a case of trading one issue for another.

"There's no easy answer here; it's not a case of simply fixing a bug, it's a case of trading one issue for another."

It's actually incredibly simple to fix this usability issue.  Other browsers have had this fixed for years.

The the incredible fix is to have an option 'show close button on tabs'.  If this is unchecked, then there's no close button!  It's not rocket science.

If someone wants to disable the close button, it's going to be pretty obvious that they'll need another way to close the tab (ctrl-w, menu, right click, etc).  This is not a usability issue because the user *chose* to do it this way.

Options solve problems for the 1% of users that actually ever find and change them, assuming the new behavior is monotonically better than the old.  They don't solve problems for the other 99% of users, and they cause scale-up problems since we get thousands of requests to add options, and options UI with thousands of options is unusable.  And again, that's assuming that for a given user one behavior is monotonically better.

I think I speak for everyone who has ever commented on this issue...

It usually goes more like, XX% of the people actually want feature Y. 1% of those people actually get bothered enough to write about it.  And a small fraction of that actually write about it on the bug tracker.  So, think about the multiplier about how many people actually ARE affected, since only a small amount of that original 1% that writes, writes about it HERE.  And if you spend a few minutes searching, you'll find thousands of posts all over the place complaining about this exact issue.

So... having been in customer-facing software development myself for many many years, I completely understand you can't add the million options that people actually request.  But I also understand you need to appease the loudest and most consistent reporters of issues after fixing all your showstoppers first. We all understand that here.

Also think about how much time you and other chromium personnel make contact on the forums about issues like this, when the fix can't be more than two or three lines to implement, 5-10 minutes of actual code, 30 minutes of reviewing and testing if that, 5-10 minutes of documentation and make an announcement, another 5 minutes to pull this into a release branch, throw in an hour of 'buffer' time for anything that takes longer than it should, round up to 2 hours for the heck of it, and the problem is solved. You've appeased thousands of people you've never heard of... some of which who have already switched to other browsers who may switch back.

There are of course the addons, but as far as I'm aware, the API for tab handling is pretty locked down and this is not publicly exposed.  And even if there was an addon that could put this obvious, core, required feature into chrome.  It's a *core* feature and we shouldn't need to dig for addons for core functionality.

Thanks for the feedback.  I appreciate it.