@Kostya: I have this feature (most likely a weakness) that I always hope that people who I consider smart, are making a bad decision simply because of an imperfect communication; and if one can help them understanding the situation as others see it, these smart people would do "The Right Thing (tm)". While sometimes I've had success, I've also been proven wrong in some cases. But I keep trying unless I get convinced that it's totally hopeless.
Your today's responses gave me a hint that you might be just not hearing a couple of arguments behind the noise about non-essential here technical details (PDFs, HTMLs, other decisions, other apps, other companies, etc.)
Hence, I will try this one more time from a different perspective.
I am 1000% in support of reflow -- if it did not cause scroll jumps.
But it did, and we got complaints.
But it did, and my attempts at fixing it did not succeed (and btw, WebView version 58 and newer acts different from earlier versions, just to give you an idea of how hard this is).
So we had to pick a compromise - and we did.
Kostya, in all this saga, you've been saying that killing the reflow-after-zoom was necessary: it was an "either-or" choice.
You have never commented on
why the 3rd option, the solution suggested by several long-term users would be a problem:
Make an option: "reflow after zoom": "disable"[DEFAULT, improves zooming experience] / "enable"(may cause problems with zooming on images).Setting its default value to the "new" behavior: no reflow/reduced jump [with a warning: "changing this may create focus jumping when you are zooming"] would reduce potential for the complaints from the users who might try enabling it.
That would be a win-win situation for the both groups of users (and will not create much of a problem for the support).
And as I also wrote previously
- in 2017, a lot of messages are optimized for mobile and do not need any "fitting" or "scrolling". We also did some work on that in 1.10 and with good results.
- other messages can be auto-fitted, again no need to scroll.
- but yet others cannot be auto-fitted because they use a rigid markup structure based on tables. Wish there was a magic way to fix those too, but there just isn't (without severely breaking their layout).
Kostya, sorry, but you keep repeating these arguments about this and that type of messages, while they are mostly irrelevant to the main point (that you seem to be missing) of people who are asking for
REflow-after-zoom: It is an
accessibility issue, when people need to zoom on the text due to the vision limitation. (And, unfortunately, no, font increase is not an acceptable universal solution, as I explained before). And once you zoom, re-flow is essential for these people. Even if it was imperfect, it was important; it was something people have been relying on heavily.
The fact that these two simple arguments have been mostly ignored, while the decisions were made on the basis of "who was louder" is the most disappointing. And that's why I called (rather strongly, I admit) the way this issue was handled --
seemingly hypocritical: the
essential usability needs of users have been ignored by hiding behind an argument of making it better for other users, while totally ignoring a possibility of making
everybody happy (except those who want the old dark theme).