Grumpy Camper

I have had to deal with some very annoying things these last few days. As usual, when problems are out of my control, they are extra stressful to deal with. While it felt like a waste of time, I think it was all a useful learning experience. </p> <h3>HDD Pre-Crash</h3> <p> I noticed yesterday that piggy was spitting out strange messages in the system log. Looking the messages up on Google, I noticed that the common suggestion was to backup and replace the hard disk immediately. Thankfully I had a spare drive pulled from an older system to work with. </p> <p> I was home sick yesterday, so this was a highly unwelcome activity for the day. To make matters worse, the 17 inch monitor attached to the machine decided to totally burn out on me (black screen, smell of char-broiled PCB) during the process. Luckily, my LCD project still functioned, and I was able to continue by using that instead. </p> <p> Luckily, this transfer happened without any problems whatsoever. The machine is operational, and I can sleep a little better at night knowing that my HDD isn't on its way out. </p> <p> Lesson: Moving Linux to a new HDD is not as tough as it sounds. </p> <h3>USB Audio Hell</h3> <p> Having written device drivers in the past, I am extra annoyed when I am faced with poorly-written device drivers. I bought an M-Audio Transit USB to use as a test device for my measurement software. This device is pretty good at doing what it's supposed to do, when it behaves itself. </p> <p> Ever since I purchased this horrible little devil in a plastic enclosure, I have had nothing but trouble on my Mac at home (beaker.supermegaultragroovy.com). As long as I've owned beaker (it'll be a year in November), it has only experienced a kernel panic once. Since I purchased the Transit USB, I have had numerous kernel panics, and system lockups. This is absolutely ridiculous for such a high-priced piece of audio hardware! </p> <p> Had I noticed the issue earlier, the device would have been returned instantly. Unfortunately, I didn't get to use it extensively until just recently, and cannot return it directly to the retailer. Yesterday I had tried a few tests using the Transit USB, and it wouldn't last longer than two test cycles in my application before it would just stop working! I would reset the machine to get it back to life, and it would keep on failing me! (Simply unplugging the device would result in a lockup, or a kernel panic, or my system slowing to a crawl until I removed it.) </p> <p> This device driver is terrible. I know it is the driver since I use the device with no issues on my Win2K machine at work (better quality audio is great when listening at low levels). Just as I was about to fire an evil letter at them this morning I noticed that there was an updated driver posted on the 19th of September. I will have to check it out and see if it solves this flakiness. I will be sure to post an update once I have had a chance to stress it a bit in FuzzMeasure. </p> <p> Oh, and for those of you wondering why the heck I would want to bother with a USB audio interface in the first place, I was trying to find a low-cost (which rules out FireWire), high-quality device that would work in a laptop configuration. This killed two birds with one stone for me, since I also got to muck around with sample rates above 44,100. I plan to purchase a high-quality PCI (likely) or FireWire (preferred, due to most having a phantom mic power option) audio interface so that I can ensure my application works with more serious gear in more serious situations. This will happen, of course, once I sell enough copies to do so. :P </p> <p> Lesson: Don't assume that a well-known audio hardware company will have great products across the board that work equally well on all platforms. Also, be sure to stress what you buy in the first few days of owning it. </p> <h3>Subversion Failure</h3> <p> As a result of the above fun, my Subversion repository managed to get affected as a result of the faulty M-Audio driver. I was using XCode at the time of the last crash, and I think that it was updating the subversion status at the same time. I was alerted to this fact by a build failure e-mail this morning. (Those daily builds are really paying off!) </p> <p> I don't know how I fixed the problem, but at least it's fixed. I futzed around with all sorts of commands, and it managed to start working somewhere in the middle. I couldn't begin to try and retrace what I did if I tried. I decided to implement a weekly automatic backup of my Subversion repository as a result. I was pretty freaked out while it was broken, since I hadn't had a chance to back it up once yet!. Now that piggy's in a somewhat stable state, I may add my svn repository and bug tracking system to its list of duties. </p> <p> Lesson: Be sure to backup important data regularly, and automatically (if possible). If you don't have time to do it yourself, your computer would be glad to make sure it's done for you. </p>