Off the Beat: Bruce Byfield's Blog

With the rise of the HTML 5 video and formats like Google's WebM, Flash may be on the way out. But for now, it remains the dominant format for viewing videos on the web. Consequently, from a desktop user's perspective, few free software projects are as important as a free-license Flash player. And that, in turn, is why the announcement that GNASH .8.8 was released last week, and is supposed to be compatible with all Youtube videos is important news -- and disappointing when it proves not to be completely true.

Not, you understand, that I would be unduly disappointed if I could never read the latest stupid video posted on Facebook. In fact, in some moods, I believe that the productivity of the entire wired world would increase several hundred percent if videos had never come to the web.

But videos are commonplace on the Internet, and many people want them. And, personally, I am tired of having to choose between the guilts of pretending to friends that I have seen their latest link or of using the non-free Adobe Flashplayer so that I have at least a chance of understanding what my friends who use other operating systems are talking about.

For better or worse, I can think of no other single tool whose development would do more to give GNU/Linux parity on the desktop. You can tell the priority that many groups in the community place on Gnash from the fact that Debian, Ubuntu, Fedora, and openSUSE

Still, the Gnash project has announced this milestone before. In 2007, Gnash announced compatibility with Youtube videos -- only to have the goal posts moved by the site's decision to standardize on a more recent version of Flash. In 2008, Gnash could, in fact, handle Youtube videos, but not small bits of Flash on sites like WordPress, where it is used to give a graphical representation of page hits. So, I was prepared to be disappointed.

Good thing that I proved myself capable of learning from past experience. Gnash .8.8's Debian plugin for Google Chrome does, in fact, run Youtube videos and the other bits of Flash I encounter regularly on the web. I would have to test it more extensively to be sure, but, so far, the Chrome plugin appears to be the closest to a free Flash implementation that Gnash has reached.

Unfortunately, the same cannot be said for the plugins for Mozilla Firefox/Iceweasel or Konqueror. Trying to use the Firefox plugin on Youtube consistently results in the message, "An error occurred. Please try again later" -- although, so far as I can see, later never comes (at least if you define it as a moment when the plugin actually works). The Konqueror plugin is even less successful, leaving Konqueror trying to load a video for several minutes, and eventually giving up. Even when you delete all Youtube cookies and restart the browsers, as the Gnash news release suggests, the results stay the same.

Definitely, as the now-standard ad lib in Gilbert and Sullivan would have it, a case of modified rapture, at least for those of us who prefer Firefox or Konqueror.

Of course, I am sure that if the Chrome plugin is working, fixing the Firefox and Konqueror plugins is only a matter of weeks for the Gnash team -- maybe even days.

Still, with all respect to the Gnash developers (who are working on an important project for the desktop with limited resources), this is at least three times now that the project's news releases have -- shall we say -- anticipated the reality. No doubt the Gnash team is under some pressure to perform, but could it not check that the news releases are accurate?

As a matter of fact, I suspect that the team did feel some caution, since the news release announced that "100% of all YouTube videos should work" -- apparently not noticing the contradiction between "100%" and "should."

But the continual gaps between promises and implementation are disappointing to those of us who want to see the project succeed. Worst, they undermine the credibility of the project. Possibly, when Gnash finishes reverse-engineering Flash 10 -- the current standard -- free software users will simply sigh with relief and start to use it. But it seems just as likely that, having been let down so often in the past, at least some users will react skeptically and continue to use the Adobe Flashplayer, which while non-free, is at least relatively consistent in its behavior.

In other words, by falling short of its promises, the Gnash project risks working against its own goals. Too many people now use proprietary software on their free desktop in the name of convenience, and somehow I doubt that Gnash wants to give them another reason not to keep people away from free software. Yet that seems a distinct possibility, if the gap between promises and implementation remains unresolved.

Possibly, all that is needed is a little more testing before releases. But, better yet, why can't those who support the development of the free desktop -- especially organizations like the Linux Foundation and the Free Software Foundation make a point of seeing that this key project is well-enough funded that it can reach its goal as soon as possible? That way, the community of users would not be disappointed again, and the common cause would benefit immensely.

Comments