MPlayer

As explained last time, the PHP script I wrote as a contribution to MPlayer’s website could not be accepted, and I was encouraged to rewrite it in pure BASH. That is indeed what I did, and I include the code for this new version below. In addition, it would be natural to give a link to the output of the script, hosted live on their server, assuming that such output did in fact exist. Unfortunately it does not, and will not, because even my new script was rejected, without so much as a “Thank you for your efforts” or “Sorry you went to so much trouble.” As such there is a bit of a story to tell, which I will do here.

My views on and dealings with MPlayer are becoming quite a saga, or at least a blog category. I now have something to report after emailing the webmaster@ address I found on their website, but (as perhaps I should have guessed given my previous criticism of the project) communicating with them turned out to be rather difficult. To start with, I never got an email in reply, and had to go into the their IRC channel to get a response, which may innocently be explained by a lack of clearly defined roles in the project. Also, I am prepared to admit that I am not without my own failings, for instance accepting, as this post will explain, that PHP was the wrong language to choose to write the Atomise script in. If I had checked a little harder into the project, like their SVN repository, I may have found the Makefile-based system they use to generate the website (as opposed to a system using files written in a web scripting language). I don’t think that this oversight on my part, however, is as unacceptable as the fact that the first reply I got in the #mplayer channel was a death threat.

As I mentioned, my desire to be above reproach when criticising MPlayer lead me to produce a PHP script that converted their website into an Atom feed. I have now emailed that script to their webmaster and await a reply. In the mean time, I hope they will not see it as a lessening of their security that I publish the source code here, especially since the project itself has benefited from the openness of their code. Also, in case we forget, I should point out that contributing to an open source project is not just a form of political activism (and even if it were “just” that it would be worthwhile in itself), but it has short term benefits (the only type which some people appreciate) in that it has increased my understanding of the Atom format so I am now confident I could use the format in my own projects, whether at home or at work. In terms of the benefit to people reading this, well, they may be looking for a script that converts a site to an Atom feed (as I was before I endeavoured to make my own), they may be interested in screen scraping in general (which is the method I am using), or they may just want to check someone else’s PHP code for style tips (or possibly examples of mistakes to avoid). Being GPL code you may do all those things, as long as you respect the requirements of the licence.

I try not to be critical in my blog posts, or at least not without giving both sides of the story, as I believe that giving constructive support achieves more than standing at a distance, making complaints without context. Unfortunately, there have been several facts I have been keeping note of recently which together are worthy of a post and which all point out limitations or failings in the MPlayer project or software. Hopefully my previous post makes clear that my noting of these problems is not without a great respect for the good work that has gone into MPlayer, but that does not put them above criticism. On that point, it is worth clarifying something, namely that it would be unreasonable to make a demand of MPlayer, as the developers are generously giving their time and effort to the whole world for Free, but that doesn’t mean those developers should never have their decisions or actions questioned. Only this week someone thanked me for a contribution I had made to a Free software project, but also pointed out a lack of functionality in that contribution. Generously, they offered a patch to introduce the functionality they desired, and I in turn thanked them for doing so. I therefore do not raise these concerns with a lack of understanding for the nature of non-commercial projects.

I like to be able to watch videos on the computer without worrying about missing codecs, annoying upgrade messages, spyware tactics with embedded adverts, and business models that involve locking up your data then selling you the key (or a virus). I also come into contact with a lot of boxes running Debian GNU/Linux. The result has been a need to install MPlayer — one of the most powerful pieces of video-playing software, supporting over 100 codecs out of the box — on one of the most popular Linux operating systems. For various reasons, this has not, in the past, been as simple as installing other software under Debian, which leads one to question whether Debian or MPlayer is at fault. Personally, I think that (despite a lot of obviously great work by the developers) history has now shown that the MPlayer project has come off the worse. I will explain this below, and it is further justified by some other observations I will make in a later post.