I have a difficult enough time just keeping my own systems up and running.
As regular visitors to this blog may recall, I’ve been dealing with desktop problems for just about the entire semester. Yesterday, the Corsair DIMMs they had graciously RMA’d came back, and eager to test out my latest theory for what could possibly be causing the freezing problems that had plagued me for months, I installed them, made the necessary tweaks, and fired up the sleeping beast, Ronon.
By raising the FSB to 1333MHz, and dropping the RAM speeds to 800, the system seemed to stabilize. “Seemed” is the key word here; without going into greater detail, let’s just say that I’ve already ordered a new motherboard, and it’s on its way here now.
Meanwhile, I undertook the task of installing Vista 64-bit to take full advantage of as much memory as possible; as a computational biology graduate student doing lots of research in sequence analysis, the extra processing power and memory available to me on this uber desktop has already come in handy, and the intensity of the computational work is likely to increase from here. Doubling my memory to 8GB would very much come in handy. Plus, with the 32-bit Vista installation, only 3 of my 4GB were addressable anyway (Windows likes to assign a block of addresses for other hardware…obviously necessary, but annoying nonetheless).
This “upgrade” had to be a clean install of the OS (going from 32-bit to 64-bit is, unfortunately, not an “upgrade” in the traditional Windows sense of the word). The install went smoothly enough; now came the step to install the numerous patches and updates.
Here’s where the trouble started.
Vista’s new integrated updater (honestly, I preferred the XP setup where you visited the Microsoft website…probably because I’ve had nothing but problems with this desktop-based bundle of angst) presented me with nearly 300MB worth of updates. Naturally, I selected all of them at once and let the system have at it.
Imagine my surprise (which quickly turned to ire) when the update ended, and out of about 50 updates, listed 3 as having “succeeded” and 47 as having “failed”, along with three separate error codes, all unknown to Vista. I rebooted as per the instructions, and when the OS booted, it entered an infinite loop of “Reverting changes” and rebooting. I couldn’t find a way to halt it, so I had to reinstall the OS from scratch and try again.
Reformat, reinstall, run update…lots of errors, reboot, infinite loop. Sonofabitch.
Googling the error codes didn’t help out a whole lot. There were certainly folks out there who’d encountered these error codes before, but precious few people who’d encountered all three at the same time. Those who had suggested something else: installing each update one, two, or three (at most) at a time.
Sweet Lord in Heaven. I gave the batch upgrade one more shot, and shocker of shockers, it loop bombed me again.
(FYI, about five hours have passed up to this point, because the Vista OS installer isn’t exactly what one would call speedy)
This time around, I installed updates just a few at a time, also ordering them by date of release, so I could be sure I was installing the earlier updates first (just in case later updates had dependencies). This worked pretty well, though it was mind-numbingly slow. I encountered a few errors along the way, though this time around no infinite loops greeted me upon reboot. If a batch of three updates threw errors, I would install those three one at a time, and the updates would install successfully.
And now, after installing MS Office 2007 and running update again, its been stuck for the past 30 minutes on the OneNote 2007 (KB950130) update. Fabulous.
This isn’t the sort of thing I could have done over the phone. Even during my co-op tenure at Georgia Tech, tech support was difficult enough for people in the same office as me; each day could feature several of the above scenarios. Typically, the culprit ends up falling into one of three categories:
Hardware Failure: Swap out the malfunctioning part and you’re good to go. Unfortunately, isolating the exact issue can be tricky as hell (Exhibit A: the problems with my desktop).
Simple Misconfiguration: The easiest to do and arguably hardest to find. It’s usually a one-line incorrect configuration setting, or a particular checkbox that needs unchecking, or a single dependency that wasn’t satisfied. And you won’t find it in the documentation.
User Error: Nobody likes this one because it is, unfortunately, the most common. The above misconfiguration is often caused by something the user unwittingly did, and something which the vendor never anticipated the user doing. Or it’s just the dumbest crap you’ve ever heard of, until something even more unheard-of happens the next day.
Really, I could handle the first and second. That’s one of the (many, many) reasons why I love working with computers – something either works, or it doesn’t. If it doesn’t, there’s a concrete reason why. It might take a hell of a lot of digging to find it, but it’s definitely there. It’s the third option that really pushes things over the edge for me. I have many stories from my days in IT support; I consider myself extremely lucky to have worked with the wonderful folks over at Georgia Tech (seriously, they were awesome), but that certainly doesn’t preclude frustrating IT issues from cropping up and people – my coworkers, random passers-by, or myself – from royally mucking things up.
Besides, programming is way more fun.