Month: December 2004

  • Plesk Coming to Debian

    While I was loitering in the requirements section of SWsoft’s Plesk website I noticed a curious thing:

    Standard configuration of Debian 3.1 (coming out in the end of December)

    Hey, check that out! I’m assuming that once Debian 3.1 is final Plesk is planning to launch support for the platform. Traditionally Plesk has only supported RPM-based distros (first Red Hat and now RHEL, Fedora, SUSE, and more). I also notice that they’re planning to roll out support for Mandrake and FreeBSD.

    While it’s obviously running a bit behind, this move further underscores the move away from legacy Red Hat, expensive Red Hat Enterprise Linux, and bloated/slow Fedora Core and towards other systems. I think it also adds to the viability of Debian as a replacement for traditional Red Hat systems. I’ve already seen anecdotal evidence of Debian being rolled out in traditionally Red Hat shops, but I’m always glad to see more evidence.

    Now that I’ve babbled a bit, allow me to explain what Plesk is for those of you who don’t know. Plesk is a web-based control panel similar to Ensim, CPanel, and even (sorta) Webmin. It’s commercial glue on top of a LAMP architecture that allows colo customers and web hosting providers to sell and resell their services. I didn’t originally have a particular affinity towards Plesk but ended up deploying it on one of my servers. It requires a little bit of maintenence and updates of course but in general it’s a very good fire-and-forget platform manager.

    I might just have to upgrade to Plesk 7.5 reloaded when I can redeploy it on a Debian box.

  • Apache2 and PHP5 in Less than 10 Minutes

    This weeks Debian Weekly News featured a link to some unofficial debs of PHP5. I’ve been meaning to tinker with PHP5 for some time now and jumped on the opportunity to do so.

    I took a fresh and minimal Debian Testing install (updated from a tiny stable install) and added the following line to my /etc/apt/sources.list:

    deb http://people.debian.org/~dexter php5 sid

    After an apt-get update I used apt-get install to install the following packages:

    • apache2-mpm-prefork
    • libapache2-mod-php5
    • Many of the standard Apache2 modules
    • Many of the standard PHP5 modules listed here
    • A few suggested and recommended packages.

    I ran the install command with -s a few times to see what would be installed, but zero to working Apache2 + PHP5 took me less than 10 minutes. Of course this is running unofficial stuff on a Testing system (though if you ask nice enough you should be able to pull this off on an otherwise Stable distro, but why would you do that?), so use it at your own risk. I’d suggest that you not try this on a production system. Until we see PHP5 incorporated into the various distros, this might be the quickest way to give it a go.

    I’m excited to tinker with the new OOPness features in PHP5. I’ve also got my eye on a particular MVC framework called Mojavi.

    Update (1/12/2005): As some have noticed it looks like there has been some moving and shaking in this unofficial unstable php5 build. After an apt-get update; apt-get dist-upgrade a few days ago, I lost the ability to use DomDocument. I checked the dists directory and changed my sources.list to reflect sid-old.

    I changed my sources.list back to sid after noticing that there were fresh new (5.0.3-0.3) pakages dated this morning at 4am. The new packages are excellent!

    Let me say again, don’t use this on a production system. But do look around, have some fun, and get a feel for php5. Special thanks to Piotr Roszatycki (dexter) for generating these packages.

  • Mobile Data Conundrum

    I’m between a rock and a hard place.

    T-Mobile is my mobile service provider. I love them to death. Their customer support can’t be beat, the people in the retail stores are actually plesant, they actually honor the warranty on phones, and they just generally rule.

    But man, their data plans are killing me.

    Here’s a bit of history for you. I’ve had T-Mobile service for almost two years now (eff Sprint) and love them. When I activated the service on my 3650 through Amazon I added their $9.99/mo T-Zones Pro plan. That got me unlimited on-phone data (plain old http port 80, WAP/WAP 2.0, IRC, etc) and even allowed me to surf via a bluetooth connection on my laptop even though I wasn’t supposed to be able to. For awhile I upgraded to the $19.99/mo unlimited GPRS plan because I was traveling and wanted full access to everything via GPRS over Bluetooth. After a few months I went back to the $9.99 plan because I just wasn’t using GPRS often enough (think T-Mobile hotspots).

    All was fine and well for a long time. I used email (outgoing) all the time, hopped on IRC via WirelessIRC when I was out and about, and was generally happy. Somewhere along the line I picked up a used taco for $60 at Gamestop and swapped my main T-Mobile SIM back and forth, using Opera on the taco and WirelessIRC, Services, and mail on the 3650.

    Then things came to a grinding halt. All of a sudden I couldn’t use WirelssIRC any more. That’s a big bummer, as WirelssIRC is probably my personal Series 60 killer app. It’s frugal on data usage, communication is instant, and I can talk to my online mobitopian family anytime anywhere.

    It turns out that I should have been paying $19.99 for access to anything other than web and email ports the whole time, on phone and off. T-Mobile realized this some time back in November/December and silently started blocking ports.

    This brings me to my conundrum. What’s a geek supposed to do in a situation like this? Did I mention how painfully slow a GPRS connection is in a Wi-Fi, EDGE, and 3G world? It’s pretty pathetic actually. It’s like when I was still using my 1200bps modem when all the cool kids had 9600’s. GPRS is slow, but it’s been very useful over the years. Until I couldn’t do something basic like hop on IRC from the phone for any less than $19.99 a month I was quite happy with it.

    I just can’t justify $20 a month for a slow connection that I’m just going to use to connect to IRC a few times a month. That’s just absurd. So I started looking at my options.

    My only real option (being a GSM guy that likes his phones to work places other than the US) was to look at the giant behemoth that is ATT+Cingular. Before being snatched, AT&T Wireless seemed like the wireless carrier with the best strategy for mobile data. EDGE is rolled out nationwide and true 3G service is available in several metropolitan areas. Before the merger, going with AT&T would be a no-brainer.

    The thing that scares the crap out of me is that Cingular appears to be so stupid when it comes to mobile data. Russ has been urging me to go out right now and pick up an EDGE or UMTS phone at an AT&T Wireless store (you know, the ones with the Cingular banners over the AT&T logo) before they go away. Why’s that you ask? Because Cingular is stupid about mobile data. Russ has an unlimited EDGE data plan tacked on to his AT&T Wireless account for $24.95 a month. Now we’re talking. EDGE is significantly faster and would be quite usable via a Bluetooth teather. If I were on an EDGE plan I could skip the hunt for a starbucks and just connect my laptop to the internet anywhere anytime. I’ve done so via vanilla GPRS and it’s just not a happy thing.

    Why am I not running out right now and picking up an AT&T phone right now? Well, because Cingular is stupid about data. The other reason that I am cautious about switching to etiher carrier right now is that they’re in the process of merging their billing systems, accounting systems, support, and infrastructure. I’d like to avoid that potential mess if I can.

    Why is Cingular so stupid about data? Take a look at their web site and look at their data plans. $30 for 10 megs, unlimited for $80? No thank you. I’d much prefer unlimited EDGE for $25, thank you. I looked at addons for their voice plans and didn’t see anything intelligent there either. Actually, I saw more stupidity. A 2MB GPRS addon called “MEdia Net” for $9.99 a month. Wow, that’s lame.

    That brings me back to my conundrum. I’ve got several options, but they all seem to suck in one way or another.

    Option 1 is to stay with T-Mobile even though they’re not going to get serious about 3G until at least 2007 and they’ve made absolutely no mention of EDGE. They’re also going to be way behind the pack of ATT/Cingular, Verizon, and Sprint/Nextel. I don’t know how long I can stick with option 1 even though I love T-Mobile.

    Option 2 would be to wait for the ATT/Cingular merger to settle down a bit and switch to Cingular once the billing kinks get worked out and all that. The problem with option 2 is that Cingular is dumb about data, and I honestly don’t think that any of their data plans come even close to fitting me or my budget.

    Option 3 is to run to my local AT&T Wireless store, pick up a Nokia 6620, and get on an unlimited EDGE plan while it’s still available. The thing that worries me most about option 3 is that I’m setting myself up to get screwed. At least there’s a lot of potential for me to get screwed. What happens if my only option in 6 months is to go on one of Cingular’s stupid data plans? What if Cingular screws up the momentum on EDGE and 3G that AT&T has been building up for years? There are a lot of questions that have not yet been answered.

    I’ve been thinking about my options for some time now and just haven’t been able to come up with a good answer. What should I do? What would you do?

  • Python for Series 60 Wiki

    It has been mentioned other places but not here. Several people have been keeping track of links, news, and information about Python for Series 60 on the Python for Series 60 Wiki. It’s running on the same semi-old MoinMoin install that houses the official LOAF wiki.

  • Delorme Bluelogger: Bluetooth GPS for $149?

    Earlier today Christopher Schmidt was looking for a cheap Bluetooth GPS receiver on #mobitopia. I think he stumbled upon a hidden gem: the Delorme Bluerunner. It looks like a no-frills but useful and usable receiver that would be perfect for GPS applications on mobile phones and PDAs. The truly best part about it is the price point: $149.

    There are tons of GPS units out there. There are tons of cheap but good GPS units out there. However, I think this is the first case of a good cheap GPS unit that also works over Bluetooth. Here’s a roundup of information about the Delorme Bluelogger:

    I’m pretty sure that this is a fairly new product, but I’m suprised that a compact inexpensive bluetooth GPS reciever such as this seems to be flying under the rader. A quick check of Technorati and Feedster comes up pretty short. I would have thought that people would be all over this. Am I missing something? Is this too good to be true? $149 seems to be a new price point for a decent Bluetooth GPS receiver.

  • Hang In There

    I can’t tell you how good it is to hear from Brian Buck. I can’t tell you how long I’ve been reading his blog, but to me it feels like we go way back. Hang in there man!

  • PyCon DC 2005 Registration Open

    Steve Holden:

    Following my last message, I am pleased to be able to
    announce that you can register for PyCon DC 2005 on the
    web at

    http://www.python.org/pycon/2005/register.html

    Rock on! PyCon DC is probably one of the best bang/buck conferences out there, and it’s right in my backyard! I hope to see everyone there this year.

  • Extending Python for Series 60

    This morning Jim pointed me to some extension modules for Series 60 Python by the Personal Distributed Information Store project. While I’m very excited to see Series 60 Python specific releases, you don’t have to go to that much trouble to get a working xml parser. There’s a perfectly good xml parser built right in to Python 2.2.2.

    You’ll probably not find this written anywhere obvious, but here’s what works quite well for getting a much more complete Python environment on your Series 60 device:

    1. Grab the Python 2.2.2 source tarball.
    2. tar xzvf or do whatever it takes to decompress the file.
    3. Insert your MMC to a card reader or plug in your taco via USB.
    4. Drag the boatload of .py files and subfolders in lib to E:SYSTEMLIBS (you did install to the MMC, right?) making sure not to overwrite any existing files.
    5. Import xml.dom.minidom or whatever library you need.

    I’m sure there are modules that just plain won’t work. Nokia have already included a lot of the basics (thank you thank you thank you for getting urllib working!). If the public release is anything like the pre-release software, the xml parsers and a few other things should work just fine. I spot-cheked a few examples from Dive Into Python on the prerelease and the xml-related stuff worked fine.

    Update: In the comments attached to this post, effbot clarifies:

    Adding the xml python package won’t help you if you don’t have an XML parser; pyexpat adds exactly that.

  • Python for Series 60 Released!

    Jim, Ewan, and Erik have already spoken about it, but Python for Series 60 is in the wild!

    I’ve been tinkering with the public release and should have some fun stuff to share shortly.

  • /me is back.

    It’s been a long couple of months and I apologize for the hiatus. It’s a long story for another day, but lets put it this way, I’m back! I’ve moved from Radio Userland to WordPress. I promise that I’ll share my (semi-painful and procrastination-ridden) migration process in due time.

    The .css that is currently driving the site is Dots by Alex King, which I’m currently tweaking. I’ve still got some random bits that I need to find and url rewrite to fit the new engine, but I’ve done my best to keep the old permalinks. If you find something that’s whacky, please drop me a line at matt at the domain ooiio.com. Thanks!