jducoeur: (Default)
[personal profile] jducoeur
Fascinating article in Ars Technica today, about the network administration soap opera going on in San Francisco. It's worth a read: it's not terribly long, and a good cautionary tale about how to screw up the running of a network. Lots of he-said / they-said, with the city essentially claiming that their network architect is a devious criminal and him countering that their bureaucracy is really the cause of the problems.

But what really strikes me is that the defense seems to partly rest on the notion that the city's information architecture is so complex and elegant that only the guy who built it can maintain it -- that's why he has needed to centralize so much power in his own hands. And I'm sorry, but that just replaces malicious intent with incompetence in my book. If he's managed to build a system that *only* he can maintain, then he is *not* "very, very good at what he does". Maintainability is one of the most important aspects of any computing environment; if you miss that, you haven't done your job.

Or to put it another way: if your company / client / family / whatever is going to have insuperable problems when you get hit by a bus, you've done it wrong...

(no subject)

Date: 2008-09-12 05:01 pm (UTC)
dsrtao: dsr as a LEGO minifig (Default)
From: [personal profile] dsrtao
I'm going to sort-of-defend the idea, though not this instance: it is possible that in a small company with limited resources, the designer and builder of a complex system may be the only one at the company who is qualified to run it.

What that means, more than anything else, is that it is an excellent idea that, ASAP, someone else qualified should be hired, and immediately brought up to speed. And then, probably, a third person. The money to do this may not be available, but it should always be on the horizon as a sensible thing to do.

Now -- the city of San Francisco? Too big to have this excuse. And this project, by all reports, is too big to have such a stupid oversight. Anything this size should have been thoroughly documented and vetted from day one, including an operations crew of 3-5 people.

(no subject)

Date: 2008-09-13 02:58 pm (UTC)
From: [identity profile] be-well-lowell.livejournal.com
Yep. And a lot of the time for the small-company scenario, the best advice is KISS, because the ability to support the thing isn't going to arrive any time soon.

(no subject)

Date: 2008-09-12 05:58 pm (UTC)
From: [identity profile] dagonell.livejournal.com
You've just described my previous job! :) The company's entire IT department consisted of the IT Manager who wrote all the code and a sys-op who couldn't write code to save her own life. There were no user manuals, internal documentation, nothing! I was hired to learn the system and document it. It was all written in Cobol-78 (that's 1978) and was spaghetti code to boot. He's never heard of structured programming. It was the Aegean Stables. When he went on vacation for a week and I handled three crisis without having to call him for help, he realized I could replace him tomorrow and I was out the door within the week as 'unable to handle position'. I was handling it just fine! If he gets hit by a bus, they're screwed!

(no subject)

Date: 2008-09-13 02:51 pm (UTC)
From: [identity profile] be-well-lowell.livejournal.com
I've been known to take insult at managers assuming nobody could take over my work if I were hit by a bus...

Profile

jducoeur: (Default)
jducoeur

July 2025

S M T W T F S
  12345
6789101112
13141516171819
20212223242526
27 28293031  

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags