...more recent posts
CSS Gradient Text Effect. Simple and very clever. Pure CSS (plus png alpha hack for IE 6.) Who knew?
Macworld San Francisco keynote by Jobs in a few hours. Will he announce my new laptop? (MacBook Air? WTF?) Everyone at World Headquarters is on the edge of their seats. Stay tuned...
I've been working fairly steadily on client jobs, but in terms of finishing my big software project (what I've called Geneva here before,) I have been totally stuck. Since about early December. Haven't moved more than a few inches. And it's been getting me down a bit (which then adds to the stuckness, which then gets me down a bit more, which then....)
But yesterday I made some progress and then today I had a pretty big breakthrough. Or I found a shortcut, which I guess amounts to the same thing. It's not quite as elegant as the piece I couldn't implement, but I need to prioritize. I really have to fight the desire to just keep working on this thing forever (it's really fun for me to be in the middle of it.) I have to finish up, yet I've never been good at finishing up.
So I'm going to start telling myself that I will launch this thing on April 1. It's software, but it runs on my server and the product is something you use in your browser over the web. Basically, it builds websites (what else?) My target market is people who want to build websites, and know a little bit about HTML, but not all the other stuff you need to actually build a modern website.
Geneva is nothing revolutionary or world changing, but at the same time I don't know of anything exactly like it. Possibly because there is no market for such a service. But it at least has a shot of turning into something. Not sure I can really finish by April 1st, but I guess I can get something out the door by then. Polish can always come if there is any interest.
We'll see.
Graph of New Egg hard drive cost/GB. Updated hourly.
ZFS for OS X is now available. Doesn't boot yet but everything else is a go. Wow. If this actually works as it's described it is so amazingly cool I can't believe it.
I'm really curious how this will turn out. Usually when someone says "I'm going to to really rethink this problem and just start from scratch" it is not a good idea. Especially if the complexity of the project is high. And building a filesystem is just stupendously complex. So you'd think these guys would have a very high chance at failing. And yet maybe they did it?
Inside Baseball Apple bit: I wonder if there is an internal tension between the ZFS people and the Time Machine team. I'm sure Time Machine is some impressive work (seeing as it works on HFS,) but you pretty much get it all for free with ZFS!
Still, it's a lot of risk to move to a new filesystem. Even if everything goes perfect I'd be surprised if Apple moved to it before 2009. And maybe it won't ever make it (or maybe only on OS X Server.) You really have to test and be 200% sure.
From the interesting because it isn't interesting department: Apple released new Mac Pros (their tower computers) on Tuesday. They are nice and extremely powerful (8 cores across the board.) This used to be Apple's flagship product. Almost it's only product. But these days it just doesn't matter. They've become so powerful that they are just niche products. Unless you are doing insanely demanding graphics work (especially 3D graphics or video,) you just don't need one. The iMac and the portable line are way more than enough power for doing anything else, and they are priced much lower than the towers.
I remember when I bought my PowerBook (5 years ago and it's still fine for my work - think about that when pricing Apple's vs. PC's) I wondered why I would ever buy a tower again. And indeed, that's the way it seems to be working out. I could maybe see buying an iMac - especially if I had an iPhone or if Apple comes out with something in the UMPC space - but the days of the consumer tower are over. Just sort of interesting.
Wired article on the creation of the iPhone:
But as important as the iPhone has been to the fortunes of Apple and AT&T, its real impact is on the structure of the $11 billion-a-year US mobile phone industry. For decades, wireless carriers have treated manufacturers like serfs, using access to their networks as leverage to dictate what phones will get made, how much they will cost, and what features will be available on them. Handsets were viewed largely as cheap, disposable lures, massively subsidized to snare subscribers and lock them into using the carriers' proprietary services. But the iPhone upsets that balance of power. Carriers are learning that the right phone - even a pricey one - can win customers and bring in revenue. Now, in the pursuit of an Apple-like contract, every manufacturer is racing to create a phone that consumers will love, instead of one that the carriers approve of. "The iPhone is already changing the way carriers and manufacturers behave," says Michael Olson, a securities analyst at Piper Jaffray.
As usual, just more notes to myself.
I run qmail with virtual domains. To pipe incoming emails for a certain account (foo@example.com) through a PHP script you have to:
1) create a .qmail-foo file in /home/vpopmail/domains/example.com which contains the following line:
|/usr/bin/php -f /path/to/phpscript.php
2) the php script can read the raw email message from stdin like this:
$fd=fopen("php://stdin","r");
while(!feof($fd)){
$email.=fread($fd,1024);
}
fclose($fd);This only works (I think this is right) if foo@example.com is *not* an actual mailbox in qmail. Also I did a qmailctl restart which I'm guessing is needed as well (after creating .qmail-foo file.)