Well, I won’t make that mistake again. Last year I used Skype’s $30-per-year unlimited outbound calling feature (to call regular phones), and didn’t have any problems with it. This year I added a “SkypeIn” number so people could call me. Since I actually tried to start using it, I began to notice some weird problems, including local callers being told they had to dial +1-[areacode]-[my number] instead of just [my number], and other callers automatically getting voicemail even though the client and account are configured to permit all calls.

I sent a support request to Skype on August 10, 2008. Today, September 1, 2008, I finally got a response (and it didn’t help, either — it amounted to “we don’t control any of that” for the long distance thing and “change your settings to something else then change them back” for the voicemail thing — it didn’t help ).

I’ll stick with my new VoIP system, methinks, and let that Skype service lapse instead of renewing. 22 days to respond to a five-minute inquiry is a pretty bad sign. I guess they’re just eager to make sure nobody thinks they’re trying to be a phone company. Even pre-breakup AT&T’s support infrastructure never sucked so bad

I’m undoubtedly tempting the fates by complaining about the very service that’s hosting this site (heh — especially if any tech or other employee reads this, I’m boned), but it needs to be said anyway.

I’m sure Microsoft is shitting bricks at this news: you can now buy a Dell laptop or desktop with Ubuntu 8.04 pre-installed and pre-configured. Complete with full 3D acceleration, video and audio codec support (MP3s, DVD playback, etc.), and full driver support for built-in hardware, like the wireless cards, fingerprint readers, bluetooth, HDMI, and so on.

It seems the excuses people use for avoiding this superior platform sure are evaporating at a steady clip, eh?

Color me impressed. The laptop’s already back in my grubby paws, with a new motherboard, and all is in working order again. The timeline:

• Sunday (July 13, 2008): First noticed funky behavior in the wireless card.
• Monday: It actually died; found HP’s support article about the problem. Called HP, scheduled a repair pickup.
• Tuesday: HP ships out an empty box to me.
• Wednesday: Empty box arrives, via FedEx Standard Overnight. I disassemble the notebook (removing hard disk, memory, and battery as instructed), pack it in the box, and drop it off for shipment.
• Thursday: It arrives at HP. HP evaluates, repairs, and ships it back to me, all in the same day.
• Friday: The repaired unit arrives back at the house. Purrs like a kitten now.

Damn. That’s good service — two day turnaround, and they actually fixed it. Spiffy. Great job, HP!

HP’s not screwing around this time (unlike my encounter with them when it came time to order a new power supply for my laptop early this year) — right on schedule, a box arrived for me today containing a prepaid shipping label, two strips of shipping tape (heh!), and packaging materials for the laptop. The prepaid label was for FedEx Standard Overnight service, and I managed to get the beast wrapped up and turned in at a local FedEx drop off spot, so this thing will actually get there tomorrow. Of course it may sit there in their lab for a week or two before it comes back, but at least it’s in the system and out of my hands now.

Well dammit, I guess my laptop felt left out of the Dead Hardware Parade. I’m posting this via that very laptop, tethered to my cellular phone for internet access, as its built-in wireless card has given up the ghost. According to Hewlett-Packard and Compaq, this laptop apparently has a bit of a “design issue” whereby it damn near melts itself down. The bulletin I linked there offers up a BIOS update that makes the system fan far more aggressive, but apparently it was too little, too late for this crazy little machine.

Wireless works occasionally, but mostly it’s dead. The Wi-Fi card doesn’t even show up in the device list in Vista, and in Linux, lspci -v doesn’t list it either. I actually had to go through the ritual of hard-resetting my cell phone to make its damned “wired tether” mode work again just to get this thing online at all. To their credit, HP are doubling the warranty period (just for this issue) and are sending a box along with a prepaid shipping label so I can send the monster back to them for repair. Given that they know about this specific issue and had trained the phone monkey in India (he had an obvious accent; sorry guys) how to deal with this directly, I suspect they’re just going to ship me a refurbished unit instead of mine back.

But we’ll see; it goes out the door to them Wednesday afternoon. This move coming up (to Tallahassee) is going to make the return shipment rather interesting

PyKaraoke doesn’t currently support pitch-shifting natively during song playback, and this is about the only missing feature that prevents PyKaraoke from being a complete digital replacement for the traditional analog CD+G player.

This is the case because PyKaraoke relies on Pygame’s MP3/Ogg/WAV playback engine, which does not (currently) provide any means of applying effects filters to the sound before it hits the sound device. It would be possible to replace this with our own threaded MP3/Ogg/WAV player, and use libsoundtouch to perform pitch-shifting (it is a very high quality, faster-than-realtime pitch shifter and time compressor/expander), but that’s a significant undertaking. There are no Python bindings for libsoundtouch, so using it would mean mucking with SWIG, which has always given me fits.

On Linux at least, there is an alternative solution that ultimately provides more flexibility anyway in building a sound layer that permits all sorts of cool sound effect possibilities.

This HOWTO attempts to document how the whole thing fits together, how to implement it on your own system, and how to use it in production at a karaoke (or really any DJ’ed) show.

Earlier today I had need of a disc image that was only available in the irritating MagicISO “UIF” format, and my Linux box was the only machine available to deal with it.

MagicISO is an irritating, closed-source, proprietary Windows-only application that is available in free and paid versions; only the paid version enables all “features.” The tool’s officially-declared purpose is to permit easy conversion between all the various (and largely useless; the ISO and BIN/CUE formats are standardized and satisfy all CD and DVD imaging purposes) disc image formats. Such tools are only useful in the first place because of all the competing (and equally proprietary) disc formats being pushed by assorted companies, and the insistence by idiot newbies and posers of using the damned things.

What makes MagicISO so irritating isn’t the fact that it only runs on Windows, is closed-source, and requires payment of a fee to enable all its features — what makes it irritating is that it introduces yet another proprietary format for storing CD-ROM and DVD-ROM images. Naturally, because it only runs on Windows, other people who run Mac OS X, Linux, or any other platform at all can’t deal with the format. It introduces compression and encryption (again in proprietary ways, which is unnecessary because compression has been done before (and far better, by much smarter people), and encryption is something that should be done separately (do you really trust the encryption algorithm developed by some guy bolting on a feature to a proprietary disc image tool?).

Fortunately, Luigi Auriemma has written a utility to convert UIF images back to the standard-issue ISO file. Unlike MagicISO itself, this conversion tool is open source. Here’s how to build and use it.

I actually began the process of migrating my domains (willfe.com, willfe.net, and willfe.org) over to a different registrar that supports dynamic IP addresses in their DNS service last week, but because of a screwup at the current registrar (sigh … so nice of their “protected domain services” to “forget” to forward mail to me…) I had to cancel that transfer and start it again, but this time the confirmation mails actually arrived and so the move is officially in progress.

I’m told that the registrar switch doesn’t actually kill anything in DNS but I don’t completely trust this; it is entirely possible that the site, or mail to it, or the domain itself might temporarily vanish if something goes wonky. Because we live in a world occupied by humans, something is likely to go wonky

Consider that fair warning that there might be an outage over the next couple of days.

Recently a Debian developer discovered that a change made (by Debian) in OpenSSL (the underlying toolkit that powers OpenSSH and OpenVPN, as well as providing a stable SSL system for open-source web servers) a couple years ago significantly reduced how much genuine “entropy” (randomness) was available at key creation time for users of the OpenSSL library. Oops The end result is that it is theoretically possible for an attacker to guess your private key with just a little bit of information from your system.

Ubuntu users already have updates waiting when they next check for updates, but there’s the additional step of re-creating your private keys too (and the requisite “installing them on your systems” step that follows) to be free of the impacts of this update. It’s actually pretty straightforward to do (and if you don’t know what the hell I’m talking about, you’re not affected anyway ).

The relevant advisories are:

• OpenSSL vulnerability
• OpenSSH vulnerability
• OpenVPN vulnerability

A handy little tool called ssh-copy-id is a wonderful little bit of shorthand to help install your key on new systems quickly. On a regular box where you’ve just got the one key, you can just copy it over with a single command:

$ ssh-copy-id username@host

You’ll be prompted for your normal login password for username@host, but after that the tool automatically installs your public key in the right place on the target and from then on you shouldn’t need to take any extra steps for it to all “just work.” It’s spiffy.

BTW, anybody who’s gotten a bit rusty on their care and feeding of ssh and needs a refresher on how to regenerate a key, just run

$ ssh-keygen

by itself for an RSA key (add “-t dsa” to get a DSA key).

To the nay-sayers out there whining that this is an “obvious demonstration” that open source isn’t as secure as it’s claimed, bullshit. This is exactly how open source is more secure — somebody noticed a problem and it was fixed and rolled out within hours, not days, weeks, or months like with a closed-source vendor. Also, there is no known exploit for this vulnerability. Finally, this was a Debian-only (Ubuntu is based on Debian, so it and any other Debian-variant Linux distribution was affected as well) goof, and wasn’t a problem with the original OpenSSL on which this package was based.