I have started listening to podcasts on my drive to and from work each day. It sure beats surfing the net on my phone while I’m stuck in traffic, or listening to corny political talk shows on the radio (I’m don’t deal well with the local stations music playlists, so tuning to all music shows is out of the question, and as I only have standard audio CD player carrying enough of my own music with me to keep me entertained is not feasable).
Anyway, currently I’m listening to TLLTS and its mostly very interesting.
On #131, the guys have an interview with Miguel Di Icaza, of Mono fame (and mc, evolution and what not), which is mostly an interesting guy, and he talks about Mono which I’m personally very interested in, but listening to him talk about other products – this guy has a huge programming langauge bias! He’s not the only one – a lot of serious hackers which are passionate about what they do have this problem, but its really annoying whenever I see it.
Miguel Di Icaza talks in the interview about creating F-Spot (a photo management application) and Banshee (A music manager), and the TLLTS guys ask him why he’d need to write these when there are great open source applications out there that already do this, and they mention Digikam and Juk (and maybe Amarok). And the main thing that Miguel has to say about it is that he doesn’t like C, and hacking on Mono is more fun then doing it on C. Which is perfectly ok – I’m all for having fun with what you write, and I fully agree that if you want to work on something and the current incumbent app isn’t fun to work on (because you don’t like the technology, you don’t agree on the design, or for whatever reason) – then go ahead and write your own.
But its one thing to start your own project and say “I started this because its fun to work on and it scratches my itch”, and its quite another thing alltogether to say “None of the current applications are worth anything, and mine is superior” – which is not exactly what Miguel says, but he’s hinting at it constantly. Specifically, he mentions some of the best features of F-Spot, and hes basically enumerates everythign that Digikam does. Now, I haven’t used F-Spot (more then a cursory glance at the interface) and I do plan to test drive it, soon, but comparing F-Spot’s most important features, except for being able to handle more then 50,000 files Digikam does it all and very easily with a very usable interface (I don’t know about scaling to 10s of thousands of pictures because I don’t have that many, but I doubt there will be any problems with that).
I cam compare Banshee to the current Linux market leader – Amarok – as I’ve tested any and all Linux music managers that I could lay my hand on in the last year, and lets face it – Banshee just sucks. Its very easy to fall into the trap of having your music manager have an “artists” list, an “albums” list and a “track” list, let the user view all them together and apply them as filters. Basically any music player come media manager since Winamp 3’s media library (which came out, like 5 or 6 years ago ?) has gone that way, and it never worked properly (unless you know exactly what you want to play but can’t get at it any other way then to filter your library, like in Rhythmbox). It works pretty well for mobile players with minimal interface where you are expected to fiddle with your player constantly, but its completely useless for PC use. Contrast to the humdrum interfaces that you get from the current crop of iTunes wannabes like Banshee, Rhythmbox, BMP and even Juk, the KDE music manager, Amarok, stands in sharp contrast – it has a usable “getting out of your way” interface which is much much easier – for the power user as well as the newbie – then anything else that I’ve seen, and it has tons of features which are currently unavailable anywhere else. You can check the LinuxQuestions.org 2005 members choice awards in the Audio Multimedia Application category and see Amarok at first place, before XMMS and much father then anything else. Banshee isn’t even a nominee.
Anyway, my main problem with this “we are superior” attitude (which is very common with the GNOME developer community in general), is that it has no basis in reality. My second problem with it is that it prevents the the culprits from learning from other people’s experience – rejecting things that have failed in the past and integrating things that have been shown to work. A good example of this is the GTK file chooser which is broken in so many aspects that its not even funny any more, and specifically the ongoing discussion about properly integrating the very well hidden text entry box which has been going on for the last two years with no end in sight – possibly a bit due to the fact that doing what the users want will cause GNOME to look too much like KDE :-(.