Adobe Confusion

As I pointed out earlier, Adobe has just donated their action script VM to Mozilla to be used in Firefox. This is really cool. For one, the VM is supposed to provide huge performance boosts. Secondly, it will ensure that your JS code works in both Firefox and Apollo to ease the migration path. However, I have to scratch my head and ask, "Why the hell didn't they donate it to WebKit/Safari?" After all, they are writing Apollo on top of WebKit, not the Mozilla's code. It is really quite amazing to me that they are investing so much in Mozilla when they are building their own platform on top of a completely different code base. Why not just build Apollo on top of Mozilla and reap the benefits of your Mozilla improvements?

Besides that, part of the philosophy of open source is that you are supposed to give back and contribute when you use someone's code. Way to rip off the WebKit engine for Apollo and then slap them in the face by giving Mozilla the VM code. If I was a WebKit dev, I think this would tick me off a bit.

[1] http://weblogs.asp.net/jezell/archive/2006/11/07/Mozilla-_2B00_-Adobe-_3D00_-Love.aspx

PS: Good luck Mozilla team, I've seen some of the code that has come out of Macromedia's doors in the past, and it wasn't pretty :)
Published Wednesday, November 08, 2006 3:33 AM by Jesse Ezell
Filed under: , , , , , ,

Comments

# re: Adobe Confusion

Tuesday, November 07, 2006 11:04 PM by John Dowdell

Howdy Jesse -- I don't have internal knowledge of the various business partnerships. I do know that WebKit best satisfied Apollo's need for a compact and focused HTML renderer, and I also know that SpiderMonkey, used internally in several Adobe products, was looking for JavaScript improvements anyway. But that's all the context I've got on all the factors that went into those two agreements, sorry... hope it provides another view, though.

(btw, you're welcome to examine the Tamarin sourcecode yourself, and work for any improvements you wish.)

jd/adobe

# re: Adobe Confusion

Wednesday, November 08, 2006 2:03 AM by Jesse Ezell

Thanks for the feedback as always JD. I've heard some of the reasons behind the WebKit usage, and regardless that Mozilla still would have been a better choice (IMO), I think it's strange that WebKit didn't get the donation instead of Mozilla. Seems like Adobe should give back to the WebKit guys, not the Mozilla guys, since they are getting that nice clean HTML rendering engine for free from them.

# re: Adobe Confusion

Wednesday, November 08, 2006 10:45 AM by Stuart Ballard

The nature of Open Source is that there's absolutely no reason for Webkit not to use the same code anyway. Sure, it's not quite the same as having Adobe's active help in supporting it and achieving a first class integration. But donating to Mozilla is NOT exclusive with donating to WebKit, and furthermore I imagine it's quite likely that in practice Adobe developers *will* help with the latter as well, since they'll need to integrate the two internally for their new platform.

My bigger worry about the Adobe/Firefox partnership is that one thing that ought to be a no-brainer for the Mozilla developers would be to integrate one of the many open source PDF viewers with Firefox. Everyone I know finds clicking on a PDF link to be an "arrrrgh!" experience because of the pain that is Acrobat Reader (actually my pet theory is that Adobe makes Reader painful on purpose as a motivator to buy their expensive alternatives). Firefox is a very good document viewer; there's no reason it shouldn't be just as good for the pdf format as html. Similarly - although further in the future - there are several attempts at open source implementations of Flash, and if one of those reaches the point of compatibility that open source pdf viewers are at today, Mozilla ought to bundle one of those too.

But with Adobe as a prominent high-profile partner, they might have enough influence on the Mozilla project to stop either of those things happening.

As far as the choice of WebKit for Apollo, from everything I've heard about WebKit it's clearly a better choice than Gecko today, for the same reasons it was the right choice for Apple when they made Safari (according to, among others, a few prominent Mozilla developers - I remember one saying they were disappointed that Gecko wasn't used but completely understood why). Gecko is getting its act together as far as architecture and bloat, and my understanding is some very large improvements will be made in the Firefox 3 timeframe, but WebKit still has a large lead.

# re: Adobe Confusion

Wednesday, November 08, 2006 1:12 PM by Jesse Ezell

"The nature of Open Source is that there's absolutely no reason for Webkit not to use the same code anyway. Sure, it's not quite the same as having Adobe's active help in supporting it and achieving a first class integration."

Mozilla could easily have done the same if Adobe donated the code to WebKit. My point isn't that WebKit can't use it, but that it seems like a slap in the face to donate all the code and resources to their main competition.

"Gecko is getting its act together as far as architecture and bloat, and my understanding is some very large improvements will be made in the Firefox 3 timeframe, but WebKit still has a large lead."

IMO, that's exactly what Adobe should have helped with if they are willing to invest in improving the engine. They apparently are willing to do that since they are giving them the VM and helping integrate it. Gecko's goals were to make a portable and lightweight engine, so I'm sure they would have welcomed the help to slim it down. Regardless, mobile devices were named as a main reason for Apollo using WebKit... yet Apollo 1.0 won't even support mobile devices. There is a good chance that Gecko will be trimmed down enough and mobile devices will be fast enough by the time Apollo 2 (or whatever the mobile device version is) has arrived. It certainly could have happened with Adobe's help.

# re: Adobe Confusion

Wednesday, November 08, 2006 1:22 PM by Jesse Ezell

"I imagine it's quite likely that in practice Adobe developers *will* help with the latter as well, since they'll need to integrate the two internally for their new platform."

Thats a really good point... they will most likely integrate the VM into their own WebKit build... but since WebKit is BSD license, they don't have to donate that code back to WebKit if they don't want to. I wonder if the lack of announcement of this means that they are planning on keeping their changes to themselves?

# re: Adobe Confusion

Wednesday, November 08, 2006 3:24 PM by Stuart Ballard

That'd be really mindbogglingly stupid of them. They'd just be repeating Apple's painful now-corrected mistake. Of course I've learned never to underestimate the stupidity of large corporations or their blindness to history...

# re: Adobe Confusion

Wednesday, November 08, 2006 3:27 PM by John Dowdell
For fast viewing of static PDF, try the beta Adobe Digital Editions... 2.5 megabytes, optimized for reading experience, but no collaborative features: http://labs.adobe.com/technologies/digitaleditions/ (On one of my computers I use Apple Preview as default for PDF rendering, and while it works well for many documents, I cannot use it for all static documents, much less collaborative documents... I'm not sure what Firefox "should" do here, myself.) I don't have any privileged knowledge of WebKit and their goals, but I do see that many of the JavaScript improvments on their wishlist are already addressed by the openly-available Tamarin code: http://webkit.org/projects/javascript/index.html jd/adobe

# re: Adobe Confusion

Wednesday, November 08, 2006 3:47 PM by Stuart Ballard

"it seems like a slap in the face to donate all the code and resources to their main competition."

I'm not sure I agree with this bit either actually. I don't think any WebKit developers wish Mozilla to fail any more than the Mozilla developers wish for WebKit to fail. Both sets of developers recognize that the measure of a standard is in the number of interoperable competing implementations. (I don't actually think that most of them even want IE to *fail* either - just to stop *sucking* so damn much). I think as long as Adobe has the sense to help with the integration into WebKit too, nobody will be upset. And you can't blame them for wanting to make a bigger splash about the Mozilla deal - Firefox is a much better known brand at this point.

I'm not closely in touch with the mozilla devs these days and never was involved in the WebKit community. But I am involved with the GNU Classpath project which competes directly with Apache Harmony, and despite the fact that there's been some acrimony between the projects over unrealistic expectations of the level of cooperation when Harmony started, to the best of my knowledge nobody on the Classpath team wants anything less than success for Harmony.

# Tamarin commentary, day 2

Wednesday, November 08, 2006 7:06 PM by JD on EP

Tamarin commentary, day 2: Fewer links today... it's off the memetrackers, the blogsearch engines by now are filled with AdSense splogs, and pro news sites are all in syndication. I'm hunting for significant original commentary; if you come across more,

Leave a Comment

(required) 
(required) 
(optional)
(required)