APNG Support

[istech]

I think this feature would be great for LC.

If we could have Animated PNG support replace Gif or supplement as a animated option. This would be excellent in my opinion. As APNG is superior in nearly every way.

Just thinking out loud.

[Klaus]

Yep, sound like a good idea!

[FourthWorld]

It would be even cooler if we could have all animated file formats displayed asynchronously. Right now, animated GIFs stop looping during script execution and some other activities, sorely lowering the value of using them.

[Klaus]

It would be even cooler if we could have all animated file formats displayed asynchronously. Right now, animated GIFs stop looping during script execution and some other activities, sorely lowering the value of using them.

Yes, and APNG support

[richmond62]

I remember a similar discussion about MNG images a while back. Came to nothing . . . . .

[endernafi]

There is and was no widespread interest in mng format.
It's dead in the water.
And same thing is about to happen to apng.
PNG group officially rejected apng years ago, seemingly for technical reasons but in fact with other considerations.
There's little support for apng, unfortunately.
Opera has moved to Webkit from their own Presto-engine; which means goodbye-apng-support.
This leaves only Mozilla which supports apng natively.

I was in favor of apng really, it's lightweight, not over-designed unlike mng.

For LiveCode, it's a risky decision to try giving support for apng.

It seems that the future is in svg and animated-svg.
And since it's approved by W3C and got support from all major browsers out there;
it was the right path to choose for Livecode which RunRev did with the vector shape objects stretch goal, of course.

We'll be missing apng.


~ Ender

[istech]

Some valid points

However I think we need to consider who will use this feature. I agree not all major browsers support apng.

But (in my opinion) I intend to use this feature in applications and not for web development purposes.

We have a free GPL apng decoder available and tests that show apng is one best formats for size and features.

This could be seen as a unique selling point for LC and I think one that should be considered.

[endernafi]

I agree with you about it's being one of the best formats both for size and features.
And I'm well aware that Livecode is not a web browser or web authoring tool.

But internet gives direction to the computing era.

If consortiums like W3C doesn't accept a technology as standard,
browsers don't likely support it { except IE, it doesn't support anyway };
if browsers don't support, end-users won't aware;
if end-users won't aware of or interest in a technology, the developers won't bother to code {Windows Phone?};
those developers {us} who develop apps targeting end-users
and those developers {adobe, oracle, etc.} who creates authoring/creating tools.

This leaves the company alone who tries to support that underrated technology.
It's hard to support a technology all alone.
Coding and maintaining libraries/frameworks of this scale is hard and requires vast resources.
Even Mozilla community has secondary thoughts about apng nowadays, have a look at bugzilla.

RunRev's developers are next to nothing in numbers when compared to Mozilla.

That's why I wouldn't put much hopes in this.


Svg is another story; it's being already actively developed by hundreds if not thousands of developers.
So, all RunRev has to do is finding a way to integrate it into the engine
and keeping it neat, maintaining, updating, fixing minor consistency issues.


As I said before, I'm a major fan of apng for quite a time;
but there are facts.

Unfortunately, being better than others doesn't always result with success.
Anyone remember BeOS?
What an excellent piece of code it was, way ahead of its time.
NextOS was just bloatware compared to it, in my opinion.
However, it was NextOS who succeeded and came to these days.

How about Hypercard?
Adobe's blackmail of "pulling Director support from Macs" to Steve Jobs killed it, well almost
Thanks to its predecessors, we are still benefiting its innovations.
S*ck it adobe


Sorry for belaboring but that's because I really loved apng and wanted to see it's been supported widely.


All the best,

~ Ender

[FourthWorld]

This leaves the company alone who tries to support that underrated technology.
It's hard to support a technology all alone.
Coding and maintaining libraries/frameworks of this scale is hard and requires vast resources.
Even Mozilla community has secondary thoughts about apng nowadays, have a look at bugzilla.

One key difference between LC and Firefox is that Firefox has to support things that are also supported by other browsers, whereas LiveCode only has to support things that run in LiveCode.

So with Mozilla, it would indeed require caution to adopt things that could result in browser-specific web pages. We've seen too much of that from companies who use only Microsoft tools internally and have somehow forgotten that there's an entire other world outside of Microsoft, so they just blithely write IE-specific code, and their pages are useless to the rest of the world.

But as you say, LiveCode is not a Web browser. It doesn't need to consider what Web browsers or any other packages do or don't do. It only needs to consider what it supports.

While it may be a low priority for the RunRev team (the Road Map still has a lot on its to-do list), anyone in the community could add that, and I'm sure the pull request would be welcome.

Svg is another story; it's being already actively developed by hundreds if not thousands of developers.
So, all RunRev has to do is finding a way to integrate it into the engine
and keeping it neat, maintaining, updating, fixing minor consistency issues.

I hope so. SVG is increasingly important. Now that LC is using the Skia graphics engine under the hood, I suspect SVG support would be easier now than ever before. Fingers crossed.

Unfortunately, being better than others doesn't always result with success.
Anyone remember BeOS?
What an excellent piece of code it was, way ahead of its time.
NextOS was just bloatware compared to it, in my opinion.
However, it was NextOS who succeeded and came to these days.

An opportunity lost, IMO. NeXT was quite cutting-edge when it premiered in 1988, but BeOS' extended metadata and fully integrated Unicode were quite remarkable, and arguably may have accelerated many types of app development that would be unique to the platform - that is, not easily reproduced on other platforms thereby providing a marketing advantage to the company bundling the OS with their hardware.

But BeOS was missing one critical feature that came bundled with NeXT: Steve Jobs. There's a lot to be said for the human side of technology adoption, and even with BeOS' advantages I would hesitate to suggest it was a mistake for Apple to bring Steve back, even if it meant using older technology (heck, they still use HFS, whose infamously-brittle b-tree and extents structures have been largely unmodified since it premiered three decades ago, so cutting-edge is apparently not as high a priority <g>).

Have you played with Haiku OS (the FOSS build of BeOS still being maintained)? I keep a copy on a VM here just because it's interesting, but sadly without an ecosystem I doubt it could ever be anything more than that.

[endernafi]

I have a VM with Haiku, too, and update it constantly.
It's fun to play with it and dream what it could be with a modern interface and integrated with modern technologies.
Heck, once I installed it onto an old PC in 2006.
It worked surprisingly well despite its apparent lack of driver support.

Haiku's problem is same with apng.
There isn't enough developers who work on it.

That's why I brought up the *browser* issue.

RunRev's sources wouldn't suffice to support an *island* technology.
Skia was a right choice, in my opinion; choosing open-source route was, too.
Community-driven technologies will help to save scarce resources.

Eco-system, you said.

Apng is losing its eco-system, sadly.


~ Ender

[richmond62]

The BROWSER ISSUE probably goes back to the fact that RunRev released a web plug-in and a way to run 'revlets' in a browser,
and just about "5 minutes later" drop it like a hot potato.

Pity, really.

[endernafi]

Pity, really.

EcmaScript was just another language, there isn't any jaw-breaking, paradigm-changing aspect of it.
Yet, Mozilla started to use it;
and Firefox was the dearest son of David against big bad wolf Goliath.
{ David = Netscape, Goliath = IE }

Firefox became popular, so did JavaScript.
With the contributions of community, look what it metamorphosed into: http://www.unrealengine.com/html5/
Now, that is indeed jaw-breaking

RevTalk was more suitable to web-scripting than any other languages out there.
HyperCard was who created the concept of hyper-text, hyper-link, etc.

I wonder what revtalk/livecode would become today if community has chosen it as the script language of browsers.
Popularity is important.

~ Ender