That said, Flash was/is an accessibility nightmare. It broke screen readers, and the animations and custom interfaces regularly created unnecessary cognitive load and confusion.
That said, Flash was/is an accessibility nightmare. It broke screen readers, and the animations and custom interfaces regularly created unnecessary cognitive load and confusion.
Later, the designers showed up and started abusing <table> and <img> and body background, thinking the web was a word processor or Photoshop, and everything started going to hell. Now we have JavaScript and <canvas> and I as a user have pretty much lost all control over my browser. It’s just been handed over to the web site publisher, who tries to get rid of my scroll bars and worries whether some element is rendered 3 or 6 pixels wide on different browsers.
In my view, we sacrificed a lot at the altar of “predictable layout”.
<map> was the true monstrosity, <table> and <img> were improvements upon it, but were burdened by padding and margins added by early browsers that were difficult to modify and retain consistency.
The situation hasn't really improved much; creating a website nowadays is a matter of dedicating yourself to a custom content platform, like Blogger/Squarespace/Wordpress, or learning the arcane language of HTML+CSS+JS.
That's a far cry from the drag, drop, click, publish or share of Flash. Just how does an individual create a .EXE of their animation/game/document and send it via email, nowadays?
They also weren't nearly as good as Flash at creating layout and appearance that would be consistent on all browsers.
> It’s just been handed over to the web site publisher
This is an illusion. It is a fact that the evolution of the web as a medium empowered publishers with more control. However, the browser is running on our computers. The truth is we let them have control. We trusted the publishers and allowed execution of foreign code on our machines.
We gave them that power. We can take it away. It began with popup blockers. Then people created ad and script blockers. Now we have general content blockers. I bet one day people are going to make GPL'd custom clients for popular websites.
People do, it's just hard to do so against any website that has an incentive to prevent you from creating a third-party custom clients, which is every site that serves advertising, which is... uhh... basically the internet.
But, see Hacker News: https://play.google.com/store/apps/collection/cluster?clp=gg... (At least, I hope that link sends you to a search on "Hacker News" on the Google Play Store. If not, well... search for "Hacker News".)
I posted about this idea on reddit and someone showed me this:
Something like this could change the web.
[0] - Except maybe Firefox, but it has to follow Chrome on most things to stay relevant.
Somehow nowadays the group think has kind of changed into what Slashdot was 10 years ago. Which is funny, since the objective of YC has not changed.
I mean I knew this mentality exists, (after all Bill Gates was rather influential), but it's not the OP who is not willing to support developers...
neither am I
This is not to claim that what we have isn’t less than ideal, but the fundamental problem with Flash is not the authoring tool itself, but that it effectively monopolized the space.
For as rough as PDF tooling is, flash is even worse.
1. Find a bug in the Flash IDE where something basic (e.g. the debugger) crashes or doesn't work in an obvious, easily tested manner.
2. Report it under the paid support plan (hey, that's what we're paying for, right?) with a straight-forward test case
3. A year later, get an email telling you that a new paid upgrade is out and could you pay another $800 to have the privilege of testing whether the problem is fixed?
The key question is if these are mutually exclusive goals or not. Can we take the simplicity of Flash's concepts and improve the accessibility, or will fixing accessibility inherently ruin the simplicity?
It seems to me that accessibility meta data can be supplied to the client independent of layout info. If Paragraph A comes after Paragraph B, the meta data can indicate that regardless of where A and B show up on the screen. Ideally they don't contradict, but my point is that they can be independent.
Conceptually think of two different markup tag sets. Set A is for visual positioning only, and Set B is for semantics (accessibility). The visual rendering engine would ignore tag Set B and an accessibility "reading machine" would ignore Set A. For practical/convenience reasons we may want to tie them together at times, but it's not a requirement for making both the visual screen "happy" and making ADA devices "happy" at the same time.
They can be independent info.