Slashdot Mirror
Interview With Linux Flash Player's Lead Engineer
An anonymous reader writes, "Ryan Stewart of ZDNet has an interview with Mike Melanson, the lead engineer behind Adobe's upcoming Flash Player 9 for Linux. It covers what the plans are for the player, what kinds of things won't be in the Linux player that are in the other players, and ways to give Adobe input on the Linux player."
Good, no need to start my browser with alsa-oss anymore if I want the mixer to work (So that multiple sound sources can play at the same time), or having the flash hogging the sound device.
How big is the team working on the Linux version of the Flash Player?
There is a core team working on porting and testing. There are various volunteers within the organization who have jumped into the effort out of general platform enthusiasm; and if we need any advice with particular areas, we bring in people from the rest of the Flash Player team as needed.
Of course, we're not making cheese sandwiches here. Throwing more programmers, any programmers, at the problem will not necessarily speed the process along.
Shit.
Is there a flash 9 for Linux, where?
So tired of shit not being developed for x86_64. Get with the times. Didn't RTFA but I assume they'll ignore it like they always have...
Tom
Someday, I'll have a real sig.
I got into Linux when I wanted to use a free relational database called MySQL for a web project
Why does this quote remind me so much of Data (from Star Trek, an obscure TV show):
Data: "It is from an obscure language known as French"
Picard: "Data, the French language for centuries represented civilization"
from one of these companies is going to contain a rootkit tailor-made for Linux.
I, for one, will not forget why I'm using a free (Open Source) platform. It sure
as hell aint for viewing snazzier adverts. Let's also not forget alternatives like
http://www.gnu.org/software/gnash/
This is not an indictment of Mike, I'm sure he's a nice guy.
I'm not a zealot --I use closed video drivers, but these kinds of needless (IMO) upgrades
just smell like trouble to me. And for the record, Flash 7 has performed flawlessly for me
on Debian Testing...
"Didn't RTFA but I assume..."
that aside, it's a valid point.
That's not a comparable situation. The PDF format specifications are freely available for anyone to use. Not so with Flash. The specifications are available, sure, but the license to get them includes a provision about not creating a player. Therefore the only way Open Source players can exist is through reverse engineering. Some do exist though, like GNUFlash, but it's not an easy task.
It's like giving nine women the task of gestating a baby in one month.
Click here or here.
Yes, but does it run on linux?
echo $SIGNATURE
... but one of the reasons Flash is so successfull is that it's considered the securest plugin (tm) in existance. Flash allways has been extremely picky about security - that's one of the reasons why it's the easiest cross-plattform VM to deploy in corporate enviroments.
We suffer more in our imagination than in reality. - Seneca
The only problem I have with the current Linux Flash is that the sound is out of sync with the picture. This makes using Google Video or You Tube a bit a hassle. So my question is
Has the sound synchronization problem been fixed?
On a side note, if you don't like flash ads with screaming sound, just install Flash Block.
But install Firefox first
Summation 2
Stopping? No, you're right. However it's not the same as PDF, PDF was (is) an open standard, they told people how to make viewers on various platforms with various tools, it's the main reason it caught on so quickly and so strongly.
As a friend of mine explained, the computer world is much differrent now, there isn't umpteen different OSs that companies have to deal with, in fact, they could (and do) get away with only supporting one. The percentage of Windows users is so high as to make everything else not even appear on many charts. The second and third places are covered by OS X and Linux, but those are so small compared to Windows that many companies don't even take a second glance.
I think this is very bad as it only makes people gravitate towards Windows more, thus making a vicious cycle. I think it would be wonderful if more companies started seeing the advantages of open standards and open source. Apple doesn't make the money on iTunes, it makes it from the store and iPods, so why not make the pprogram open source?
I got really pissed at Adobe recently for their idiotic canned support emails ("You seem to be having trouble with , you can find help with that at our FAQ, if you need more help please reply to this email" Dude, if your FAQ had the answers I was looking for I wouldn't have emailed you...). It sucks that a) there hasn't been a new flash player for linux for quite a while and b) there has never been one for AMD64. Having had several back-and-forth emails with Adobe support I got the sinking impression that the Linux versions had been dropped and were never going to be updated. I'm glad this is not true, and I applaud Adobe for doing the (mostly) right thing in releasing a Linux version of their player. Of course, if they opened the standard we'd get better flash players quicker and they wouldn't have to pay as many people to do it, win-win, if you ask me.
I don't understand why more companies aren't seeing the advantages of open source, but at least some of them aren't ignoring us completely.
There is a open source flash player, though it doesn't work too well.
...I've got to say that this long-winded dragout of the next Linux version of Flash Player (hey, isn't both 8 final and 9 beta out for Windows already - neither of which we've seen in *any form* for Linux?) is getting rather tiresome. Sadly, the current Linux Flash development "team" (who is involved exactly in writing the Linux-specific code? The article doesn't really spell it out - you do suspect only one person has been assigned to do that and Adobe don't want to publicly admit that) haven't helped by spinning things out with their blog.
I find it amazing that "obvious" steps haven't been taken by the Linux Flash team, namely:
* Some sort of release schedule announcement - don't care if it slips by a few weeks here and there.
* A set of pre-releases (heck, have them time-bomb out if you don't want them being used in the long-term) coming out to showcase its current alpha/beta/RC status. Note here - Windows gets beta releases, why can't Linux?! It's utterly shameful there is no pre-release version for Linux, especially since the latest Linux Flash blog entries brag how stable the player now is at all the major sites it's been tested on!
* A definitive statement on whether they'll support 64-bit (i.e. "it'll be released at the same time as the 32-bit version" or "it'll be released X months after the 32-bit version" or "it'll never be released"). Sadly, Adobe are somewhat pig-ignorant w.r.t. the 64-bit platform and don't even have a 64-bit version for XP!
* Explain the exact differences between, say, Windows Flash 9 and Linux Flash 9 - there's some woolly stuff on this in the article really. After all this time in incubation, you'd have thought that the two platforms would have identical version 9 players, but I wouldn't it past Adobe to release a half-baked Linux Flash 9 player, since they have not yet demonstrated to anyone at all that they take Linux seriously (does the word "vapourware" mean anything to Adobe? That's exactly what Flash 8/9 on Linux currently are).
* Start a merge of the Linux development environment and the Windows one, so that ultimately they work from the same codebase to avoid the ridiculous delays in platform releases we've seen in the past. It's not clear to me if the Linux effort is fragmented - have we been told how much code is common on all platforms and how much is specific to Linux (and how they keep the specific code to a minimum)?
* Open Source the player! If Adobe have coded the entire player in-house (which I believe they have), then why not Open Source it...it's a free download after all! Even if they've patented some methods used in the source code, they own the patents and the copyright on the source code, so that shouldn't stop them open-sourcing it surely? Just exactly what is Adobe's objection to open sourcing the player? Sheer bloody-mindedness?
The major problem I got with Flash for Linux is that keyboard simply don't work! There are a lot of mini-games out there that use keyboard for controlling the character, but I get nothing trying all the keys. Maybe it's because I'm using a pt-BR (ABNT2) keyboard, but it dosen't seem to be a problem for Flash Player in Windows :-(
why does flash 9 for linux suck horribly compared to all the earlier releases?
There *isn't* a flash 9 for linux, sucky or not. It doesn't exist. That's what the dude is working on.
When are you going to release a fixed version that actually works right or at least comperable to the mac or windows versions?
TFA said "early 2007." That's what he's working on right now!
Sheesh.
Software Wars
He's a troll because the article specifically mentions at one point exactly what to do if you want a 64-bit player. You keep pounding the adobe wish list with requests.
The Tao of math: The numbers you can count are not the real numbers.
There are 2 big reasons why it is unlikely that Macromedia will change and allow the spec to be used to build players.
Firstly, just like with Sun and Java but much more so, flash truely is "develop once, run anywhere". Any web browser on any platform running the relavent version of the flash player plugin for that platform can play any shockwave flash file out there. (which is probobly why web designers love it so much...)
And, just like Sun and Java, if Macromedia goes open source or open specs, how can they be sure that "GnuFlash" can play ALL the flash files the same as how the Macromedia player can.
The other reason is the mobile devices space (PDAs, cellphones, smartphones etc). Right now, Macromedia is pushing heavily into the mobile space and trying to convince mobile device manufacturers to ship "flash for mobile devices". I dont know details but I imagine mobile device makers have to pay Macromedia to ship "flash for mobile devices" in their device (especially when a source code licence is required and its not just a binary provided by Macromedia). If the specs or code were open, the mobile device manufacturers wouldnt need to pay macromedia.
What the hell is PPC? Some old, arcane hardware? Is that like an Amiga? Damn, you guys _never_ shut up, do you?
They've made their position on 64-bit support pretty clear.
Ignoring the 64-bit world seems shortsighted to me. Sure, most users are 32 bit at the moment, but are new 32 bit machines even sold any more? Old stock, maybe, before current models push it out of the supply chain. Even Semprons are 64 bit now.
The whole question about wanting the player to "just work" on other distributions really brings up a question that the linux world has been avoiding for at least 10 years, and that is the question of why so many ideosyncratic distributions exist. What makes open source so powerful and effective has also made a mess of interoperability where Linux is concerned.
./configure --put-this-there --this-is-there --look--for-this-here --my-init-scripts-are-here --use-this-and-not-that;make install everything yourself, you'll not be bothered by a lot of this. But suppose you are the mythical, mainstream Linux dekstop user who doesn't know wnaymore about Linux than it installed from the CD no problem. If you are looking for a piece of off the shelf software are you reall going to see something on the label akin to the following :
Why is that each distribution of Linux has to be so ideosyncratic that a body cannot produce a binary installation that "just works"? Why should that even be a question? Isn't this a stumbling block in terms of mainstream, desktop adoption of Linux? Sure, if you can
Compatible with RedHat Linux, SuSE, Slackware, Debian, Gentoo, Mandrake, Ubuntu,SlackHat Redbian, Mandrux, Unbonux, Seus, ZuSE, Debware, Mandhat, Slackdrake, Jesux, Paulux, Vitamin-C, and Bean Crock Enterprise
Even though you can really categorize most into a few base types, what is to gurantee that my Rhinestone Pantux will run something as easily as my Blue Sude Linux even though they are both based on RedHat?
Is this GNUFlash the same as Gnash?
factor 966971: 966971
The current (but outdated) Flash player 7 for Linux has big problems with audio/video synchronization. They are hoping to solve this by getting rid of OSS support and using ALSA exclusively. This is a good move. But I also see that they do not plan to support the current version of Video For Linux (V4L2), although the older V4L is being phased out of the kernel. And in that engineer's blog, I saw a brief statement about the fact that the Flash player will not use Gstreamer. This is bad.
Why don't they use Gstreamer? This would solve the synchronization issues (the current gstreamer-0.10 is very good at keeping everything in sync, unlike other multimedia frameworks) and it would also provide good support for both V4L and V4L2. In addition, it would provide a good cross-desktop integration, because Gstreamer will be supported in KDE4 (through Phonon) and in GNOME.
Currently, Gstreamer allows me to configure multiple sound cards correctly and decide in one place which one is the default one. If the new Linux Flash player ignores Gstreamer and codes for ALSA and V4L directly, then I bet that it will have problems picking the right sound card automatically. And it will probably ignore my gstreamer filters as well, which is a pity. Not to mention that it would force me to keep the obsolete V4L code in my kernel instead of using V4L2 (gstreamer would do the switch transparently), just like Flash Player 7 forces me to keep the OSS API (alsa-oss) while all other programs have moved to a more mature interface (ALSA).
By the way, I have read some comments in the blog saying that Gstreamer should not be used because its API or ABI is not stable. I say: bullshit. There were some incompatibilities while moving from gstreamer-0.8 to gstreamer-0.10, but this was a long time ago and the interfaces have been stable since then. If I remember correctly, the Gstreamer developers stated that they intend to keep the interfaces stable now. So those who reject Gstreamer for that reason are just spreading FUD.
factor 966971: 966971
Just keep hounding Adobe with requests for Linux versions of all of their products. They will eventually realize it's wise to cater to a growing market. If they continue to ignore Linux, they risk being rendered irrelevant, especially with Xara, Inkscape, krita, and the gimp all quickly maturing. I'm sure that alternatives to Adobe's flash player aren't far off if Adobe continues to drop the ball. I've read that the so-called "lead engineer behind Adobe's upcoming Flash Player 9 for Linux" is the ONLY engineer assigned to the upcoming Flash Player 9 for Linux - I have NO idea if it's true (but the thing is so late I suspect it is) but if it is, that's pretty sad. We'll end up getting the Flash Player 9 for Linux right around the time that Flash 10 for everything else has alreaqdy shipped.
The Christian Right is Neither (Christian nor right). See: Matthew 23, Matthew 25, Ezekiel 16:48-50
From the article:
I started doing some homework and began contributing to, and occasionally leading, various multimedia-related open source projects and efforts, such as xine, FFmpeg, and MPlayer.
So I'll say yes (at least for FFmpeg)
Guvf vf abg n EBG zrffntr
Flash Player 9 is not just a C program any more. It now has a JIT, and so they have to write a separate JIT backend for every architecture they want to support (although an interpreter might be a good stopgap). They haven't written the x86-64 JIT yet, and it's going to take time for them to do it. Sure, it might have been better if they delayed the release of Flash Player 9 until after they developed x86-64, PPC, ARM, IA-64, MIPS, and Alpha JITs, but it's too late now.
You're clueless, and you just let everybody know it.
64-bit code does indeed run faster. The data may be bigger, but the code (.text segment) is actually smaller. There are more registers. The calling convention for 64-bit Linux can pass up to 6 integers in registers. Stack accesses are greatly reduced. This is enough to be noticable to casual observation.
A typical modern Linux has only a handful of 32-bit binaries, certainly not including the web browser. OpenOffice.org is even 64-bit now. Some systems don't even include the 32-bit libraries anymore, thus being unable to run 32-bit apps.
Running 32-bit stuff is gross. It's like running 680x0 code on a Mac, or 16-bit DOS apps on Windows. Normal people hate that.
Bandwidth-wise, if nine women had nine months to each generate a baby, then you can say each woman will have created one baby in one month.
The Latency simply was nine months.
The Throughput is the twelve hours of labor it took to push the brat out.
Consider the ramifications from a Network Admin's point of view.
If you could completely eliminate the latency, then you could push a baby out twice a day. You could make millions off of welfare alone!
However, the lowest latency I've ever seen, was that big-bellied chick I took home to get it on with, and she gave birth the very next day. It was only 24 hours between doing her and making my first child support payment.
I'm trying to apply these findings down at our data center, but all I have to show for it is some slippery floors and some unusual-tasting coffee. My boss must have taken an interest in the concept, because I often see him and his secretary attempting a similar effort after hours.
Solomon
"Twice half-assed makes an ass whole." --Solomon K. Chang