it's never lupus

Dec 31, 2008 22:54

One of the great things about my job is how much, on a good day, it resembles an episode of House. But, you know, with less blood and screaming because we have computers for patients.

The amusing thing is how often the principles in TV show medicine apply. Dr. House can tell us a few things it's good to remember: ( Everything you need to know about computer systems analysis you can learn from listening to the esteemed Dr. House )

Leave a comment

Comments 7

psychohist January 1 2009, 06:12:25 UTC
Fixing a hypothetical problem is the most reliable way to verify it was the real problem...

Shows the difference from my job! I've seen way too many messes from people fixing hypothetical problems, then failing to unfix them when it turns out they weren't the real problem.

Reply

harrock January 1 2009, 17:16:29 UTC
Oh, I don't think we're saying it doesn't cause problems down the road. Just that it's quicker and more reliable than trying to formally prove anything about a complex system.

But I suspect that hypothetical fixes in your genre are more likely to corrupt data, etc, than hypothetical fixes in our genre...

Reply

psychohist January 1 2009, 18:35:50 UTC
I'm kind of thinking treptoplax might be the kind of person who would back out 'fixes' to problems that turn out not to be real, anyway.

Reply

treptoplax January 1 2009, 21:05:22 UTC
Ok, I probably exaggerated a bit for comic effect there; certainly I don't mean to condone "shotgun debugging", where we've given up on even having a working model and are going to just randomly frob a bunch of stuff (usually starting with a reboot) in hopes the problem goes away.

But to elaborate on what harrock said; the only 100% (almost) reliable test is to try it, so if that's the easiest test why not do it first? (And, yeah, that's a lot better idea if the change is a volatile one that'll revert at the next reboot regardless unless the configuration is changed, or at least can be flipped back on the fly...)

Also, we often to get apply another Houseism to that: "If I'm right, he gets better; if I'm wrong, he's dead either way."

And yeah, it's certainly also true that a lot of the issues we run into are the result of complex solutions to simple non-problems ("Just to be safe, every time we write a row to the database, let's read it right back to make sure the update took!" WtF?)

Reply


twe January 1 2009, 07:01:21 UTC
Snort!

Reply


harrock January 1 2009, 17:21:59 UTC
I don't watch "House", but I'd add a principle to that list:

"Some patients are suicidal, whether they know it or not". Everyone wants to live, but some of those who land in your hospital are there because they made bad decisions which they will repeat.

Reply


greyautumnrain January 1 2009, 18:47:14 UTC
Yep, especially the second one.

Reply


Leave a comment

Up