In November 2012 the Mozilla Foundation announced “Project Shumway”, an effort to create a “web-native runtime implementation of the SWF file format.”
Two-and-a-bit years, and a colossal number of Flash bugs later, Shumway has achieved an important milestone by appearing in a Firefox nightly, a step that suggests it’s getting closer to inclusion in the browser.
I was unaware Flash needed a ‘killer’ at this point.
I’ve heard some bad product names, but that is right up there.
I like it!
Flash Gordon –> Gordon Shumway
Thom, I do not know what sort of Internet you use, but where I roam there are still plenty and plenty more websites that insist on using Flash for even the most basic functionality. Some entities simply refuse to consider dropping Flash, no matter how hard you try, and as you should know by now Adobe’s Flash often comes with a whole boatload of vulnerabilities in it. If you can’t wean websites off of it, then what other choice is there than creating a client-side alternative to it?
One can only hope Shumway will prove more secure than Adobe’s Flash.
Doesn’t HTML5 offers some kind of similar functionalities ? Canvas, video, sound, …
Yes. Shumway actually translates flash into HTML5, thus they say “web-native”.
In short no. HTML5 offers similar functionality sure, but they are all harder to use, aren’t as well optimized (if you thought Flash drained your battery…), and the workflow is simply no where near as good as the old Flash CC -> SWF workflow (without ever using a single “polyfill”). HTML5 tools and runtimes have closed the gap from years ago, when everyone started declaring Flash is dead, but there is still plenty of gap left. You still can’t get audio working cross platform easily, and most polyfills just use Flash.
Then there’s mobile, ostensibly the whole reason we dumped Flash. Have you ever tried to make anything work on an Android browser (and all it’s versions, with constant regressions)? Compare that with Adobe AIR, which usually works just as you think it should (AIR is Flash for mobile apps.)
I’d say we are still 2 or 3 years off from having some truly great creative workflows that rival what Flash used to provide. Unity in particular is doing it’s job. There are some others that look promising. I miss Macromedia though.
I thought AIR was HTML5 in a plugin.
I understood it to be Flash mushed into a specialized copy of WebKit which granted extra permissions.
On mobile it’s literally the Flash Player (the Actionscript 3 part – no legacy support), with a couple of additional APIs to access device and app capabilities (including one that uses the device’s built in HTML WebView), and some hooks for Native Extensions (ANE). The desktop version packs an older version of WebKit, but the mobile version is really just Flash. On iOS it does AOT on your ActionScript bytecode to turn it in to native code. Pretty neat tech.
Highly unlikely it wouldn’t be, given that it’s replacing Adobe’s ActionScript runtime with unprivileged code running inside the browser’s existing JavaScript runtime.
(Same type of approach PDF.js uses to remove the need for Adobe Reader in most cases.)
And, of course, Mozilla and Google have both proven themselves much better at providing a secure runtime environment for an ECMAScript dialect than Adobe has.
Yeah, lots of TVs only have Flash streaming for their media libraries.
That’s right. Here in Brazil almost all newspapers offers online subscriptions, but their readers is nearly always Flash only. Same for TV stations here that offers free streaming and online radios.
Netflix uses HTML for it’s user interface, and it’s nice and snappy and easy to use. They are moving from Silverlight to HTML5 for the video interface as well. It works quite nice.
Shomi is the Canadian alternative to Netflix … and it’s a giant Flash applet! In 2015! It’s slow to load, slow to navigate around, and the video interface is nowhere near as fluid as Netflix. Why, oh why, oh why did they choose 2015 for a new project!?!
Unfortunately, this means that I can’t get rid of Flash on my HTPC.
The only upside is that it’s built into Chrome, so it auto-updates every 6 weeks along with Chrome, and I don’t actually have to think about it.
Will be interesting to test it with the Shomi site to see if it’s an actual Flash replacement, or just a Flash video player replacement (two very different things).
Netflix using HTML5 has always kind of made me happy and sad at the same time. Because Netflix is one of the primary reasons HTML5 now indirectly has support for DRM:
http://www.osnews.com/story/27731/Mozilla_DRM_and_the_challenge_of_…
It’s not a Flash-killer, HTML5/Native has already done that. It’s a Flash-legacy. A safe way of supporting all the Flash content out there for the future. That is a good thing, unquestionably.
This is a massive, massive win for security and a massive win for the backwards/forwards compatibility of the ‘Web. That you will be able to watch old Homestar Runner episodes in 10, 20 years time without having to set up a VM with an ancient operating system won’t occur to you right now.
They are talking about plugin, not format. Actually, Shumway is more or less a life preserver for Flash, as it brings flash support to the platforms where plugin was unavailable.
Flash just needs to die. I hate it. I hate it even more that too many people/organizations cling to it as if it were manna.
Try running Firefox with Flash set to Ask to Activate and browse. Notice how many websites have absolutely zero flash content, yet they have the plugin embedded.
See, closed flash is a tamper-resistant sandbox for storing persistent cookies.
If I could have the possibility to set up Shumway to run Flash content but send cookies to /dev/null, I’d be happier
Well, presumably that’s exactly what you’ll get, since Shumway is basically translating Flash into native JS/HTML5 elements. You’ll be getting the standard cookie handling provided by Firefox…
It’s how Sean Connery pronounces “Some way”.
Connery: “I need shum way to shee these shtupid shwf webshites”.
Mozilla: “Here you go”.
Anyway, the real origin of the name:
http://www.hanselman.com/blog/JavaScriptHasWonRunFlashWithMozillaSh…
…I liked your explanation better.
Edited 2015-02-18 17:49 UTC
I’m waiting for it for a while. It’s still quite unusable for video (probably the major bulk of use cases where Flash appears now).
Edited 2015-02-18 18:09 UTC
Have been waiting to have systems w/o flash installed for a long time, looks like that is finally coming true! Yay!
People need to realize that there is more Flash based content than just websites and streaming services.
There are alot of web applications that are written in Adobe/Apache Flex, and they require the Flash runtime to work.
Without the Flash runtime, those applications will be completely dead, and it can be a massive undertaking to rewrite the applications in (for example) HTML5.
I’ve been working on a pretty massive web application for several years at my job and at the time, Flex was, unfortunately, the best alternative available.
Today, pure HTML5 would undoubtedly have been a better choice, but after years of development and lots and lots of cash thrown into the project, rewriting is simply not realistic.
So, if we can make a translation layer for the browsers, that can translate ActionScript+MXML to JavaScript+HTML, then we’d be in heaven.
This is a HTML5 runtime for compiled swf files, so it’s very very good news.
Also, I know Apache is working on a cross compiler that can compile ActionScript+MXML to JavaScript+HTML, it’s called FalconJS.
For our needs, that’s even better than a HTML5 swf runtime.
FalconJS seems like a better fit for a project like that than something like Shumway. I used to follow that pretty closely a few years ago, and they were already pretty far along. I wonder how that’s going.
Seems to have stalled somewhat.. :/
Apparently, there were multiple different approaches to the problem over at Apache, and FalconJS was one of them and the one Adobe started before them.
There’s also FalconJX and FlexJS, which seem to be the most up to date initiative, but it’ll be many years before they are anywhere near being able to compile a project like ours and make it work.
In the meantime, if Shumway gets traction, it might be a good steppingstone while we wait.