Two things I’ve come to realize after leaving the Linux community :
- it’s actually very small
- if you left out of disillusion, you’re not alone
Two things I’ve come to realize after leaving the Linux community :
Despite the many predictions that CDs will soon disappear and be replaced by fully digital distribution, I’ve always thought that they would rather be displaced toward a “high-end product” niche. That is, mp3 is for the music you just like and listen to casually, but for bands and artists you really care about, you’ll gladly purchase a CD.
Back when the CDs first appeared, I was still mostly using audio tapes for stuff I didn’t really value but was interested in nonetheless. Then CDs became much more common, all stores started to have bargain bins, and lending CDs from friends or a library replaced the tape. Then came CD burners. And finally mp3s. Nowadays if someone tells me about this band he’s just discovered, the band’s name is usually enough for me to find out what he’s talking about.
So, it seems a study has somewhat confirmed my intuition. It’s surprising that a study commissioned by the British recording industry (British Music Rights) would reach conclusions which are (apparently) not totally biased.
Among other findings, 80% of the youngsters they polled claimed they would pay for “a legal subscription-based music service that would allow them to discover, swap and recommend music”. May be there’s room for this kind of service after all. While I still believe Jobs got it right when he said that people don’t want to rent their music, there could be a “don’t care so much about it” space where renting would be good enough. Time will tell…
By now you’ve probably already seen this leaked video from Microsoft, apparently an internal “motivating” video from the Vista sales department, about SP1.
Microsoft claims it’s actually a spoof, but I have a hard time believing it. I’ve seen my share of sales droids making complete fool of themselves in trying to be “cool”, embarrassing their audience to the point of physical discomfort. And for a spoof to be this long and so dull, it’s just not convincing at all.
At least when geeks behave in a ridiculous fashion, like when they go to a Sci-Fi convention wearing appalling costumes, they have the excuse of being out of touch with reality. That’s part of the definition. But sales people ? These ones are supposed to be very much in the real world, aren’t they ?
May be not so after all :-).
Well, I just tried that this morning, and… it worked, much to the amazement of the guy at the desk. Reading the comments on both posts, it seems there’s nothing new to it though, people have been doing this for ages, even on other devices.
So today I finally made official my move away from Linux, through a post on the rosegarden-devel mailing list : I’m an ex-free software Linux developer, aiming to be a free software OS X developer. My only use of Linux is now on my home server (except at work where I still write Java on Linux). It’s been 13 years since my first Slackware install, with kernel v1.2.8 back in 1995 on a 90MHz pentium with, at the time, 16Mb of RAM and 512Mb HD.
The first reply I got was a “me too” from an other user, who had made the same switch for the same reasons. He also mentionned that sound on OS X wasn’t that rosy either : USB peripherals recognition problems, stability issues with Logic Audio… However these problems are on a higher level than those on Linux. Actually, I wish Linux had this kind of problems, but instead we have to deal with the most basic “can’t hear anything” issues (or similar basic stuff like my mouse’s tilt wheel being broken by a kernel update). And I’m just so f*cking tired of this.
Plus there’s the fact that writing a music editor for an OS which already has the equivalent of alsa, jackd, qsynth and our own RG sequencer in it is somewhat more motivating than endlessly trying to debug random sound setup problems.
To tell the truth, even though I did gain a lot through these years, had I known that Linux would still be in such a state after all this time, I would have had jumped ship two or three years ago already, when OS X was starting to look welcoming enough. Core Audio might not have been as mature as it is now, though.
One thing which I’ve long wanted to do with my ipod was to easily keep it filled with a random selection of albums. Not songs. I don’t like listening to songs at random, likewise my squeezebox is most often used in “random mix album” mode (the only caveat being with albums fitting on two CDs – there’s an option to tell it to collapse multi-CDs albums in a single one, but while you want this for a concert spread on two CDs, it’s less useful on a boxed-set of 10).
So I wanted to have the same thing for my ipod, and the solution is simply a playlist which randomly selects albums. It’s actually very easy though not quite obvious.
As a digression, I was also looking for a way to automatically expire old mail in some folders under Mail.app. Again, no specific feature for this, but it’s very easily done with filtering rules. The common point in this is that a UI is kept simple by making some of its features powerful enough that they can serve other purposes, rather than adding more features which would be rarely used. It’s a hard balance to strike, though.
So my iphone was finally activated on friday. First impression : it is cool. Just as damn cool as the demo videos show it to be. It feels like you’re in a Sci-Fi movie because of how amazingly well done it is.
There are a couple of lacks :
But ultimately this is, without contest, the best designed piece of hardware and software I’ve ever seen. Today I smiled when I first placed a call while listening to a podcast : just select the phone number, the podcast’s sound fades out, talk to whoever you’re calling, bla bla bla, hang up, the sound fades back in. I had seen it in the demos, but this is one of the things you actually marvel over when you experience it directly : “damn these guys are good”.
But what really marks the difference between any existing phone and the iphone is how you set up your voice mail : Enter your code. Confirm it. Record your message on the iphone (and not through it), listen and re-record as much as you need, press “ok”, and it sends the whole thing over to your provider. And that’s it. And you don’t have to meddle with one of those dreadful voice menus (“press 1 if you’re happy with your message, press 2 to record it…” *hurl*). Once I had done it, it just struck me : there’s nothing special here. That’s just how it should happen, and every other phone which forces you to go through a voice menu is retarded.
One last thing (pardon the pun) : following up on my previous post on the iphone’s potential as a universal remote, I confirm that driving my squeezebox with it works like a charm (just use the Nokia 770 skin until v7.0 of SlimServer – now renamed to SqueezeCenter is released, then you can install ipeng, a skin dedicated to the iphone, which depends on SqueezeCenter).
It’s not activated yet, there’s a 5 days delay to port my current phone number from SFR (my current provider) to Orange. Cédric had told me that he was sure I would not get one… Sorry, you were wrong :-).
Now on to some scripting to move my Treo contacts to my Macbook’s Addressbook.
OK, just quickly ranting about this :
KMail doesn’t have a spam filter, it’s supposed to integrate existing filters through its filtering system (there’s even an “antispam wizard” which automagically creates the right filtering rules). All these filters add a header to the message, therefore modifying it. Apply that to an IMAP inbox where the non-spam messages are kept in the inbox, and you get kmail re-uploading every single non-spam message after it’s been piped through the filter. And I can’t find a way to do it otherwise. Even worse, you have to create a “non-spam” folder to put the messages in, otherwise kmail refilters all messages every time you enter the inbox.
Now on top of that, there seems to be a funny bug that causes kmail to get stuck on a message, endlessly refiltering it, and re-uploading it to the “filtered” folder so by the time you cancel the filtering job (and, incidently, get kmail stuck in a state where it won’t display the content of any of your imap folders, forcing you to restart it) you find yourself with multiple copies of the same message in your “filtered” folder. Great.
This is already annoying enough – I’ve just (painfully, as usual) upgraded to Kubuntu 7.10, using KDE 3.5.8, and there’s no improvement of the situation – but since the bug happens only on large messages, I figured I could just add a filter rule on the message’s size. Oh, great, there’s one. Except it tests the message’s size in, wait for it… bytes. And you can’t change the unit, like to something a tad more useful like kb. Seriously, when was the last time you ever felt like you needed to filter your messages according to their size with a byte-level precision ? And good luck for Joe User who wants to filter between “large” and “normal” messages (pretty much what I want to do). Typically I want to let messages over 1Mb pass through. So I should enter a six digit number in that filter rule ? Now that’s convenient, counting zeroes. Love it.
But wait, it gets better. There are no less that 5 operators for that “size in bytes” rule. 5. “less that”, and “greater than”, of course, but also… “less than or equal”, “greater than or equal”, and of course… “equal”.
This is akin to sorting freight containers with a scale precise down to the gram. Not only some moronic coder actually thought it would be useful to have such a filter criteria available, this went undetected and has remained so for years.
Next time, a rant on the kubuntu upgrade, with a config script crashing silently causing me to do this at the command line level, and ending up with a partially functionning mouse (which used to work fine – now half of the buttons aren’t recognized, so no tilt wheel).