Welcome to the future
Feb. 17th, 2009 10:21 amNamespaces. They're common in IT - things like "the people who use this service" have names according to a consistent formula, like BillG at Microsoft, because to do it otherwise is madness. When you have several namespaces, it's important that they don't clash. One set of username servers here, for example, had Andrew_G; the other GatesA. Each service had a user drive that was named after the username.
Two years ago, the Powers that Be decided that the old Andrew_G naming rule was fuddy-duddy and embarassing, and changed midflight to the new rule.
Can you spot the schoolboy error? It's absolutely fine as long as none of the services ever mash up. Because if they do, say, when politics or scalability make it necessary, then some poor sod has to grovel around in thousands of 'em unpicking the ones that don't mash right. One namespace's Fred Bloggs, Fred_B, would go fine alongside the other's BloggsF, Ferdinand Bloggs. But when you change rules midflight, we've got BloggsF's on both services - and a hairball of rules as to which one becomes BloggsFe or BloggsFr, and then one of the buggers will probably leave, and -- gah!
I warned them about this hairball two years ago!
Namespaces: One instance where "stay the course" isn't boneheaded.
Two years ago, the Powers that Be decided that the old Andrew_G naming rule was fuddy-duddy and embarassing, and changed midflight to the new rule.
Can you spot the schoolboy error? It's absolutely fine as long as none of the services ever mash up. Because if they do, say, when politics or scalability make it necessary, then some poor sod has to grovel around in thousands of 'em unpicking the ones that don't mash right. One namespace's Fred Bloggs, Fred_B, would go fine alongside the other's BloggsF, Ferdinand Bloggs. But when you change rules midflight, we've got BloggsF's on both services - and a hairball of rules as to which one becomes BloggsFe or BloggsFr, and then one of the buggers will probably leave, and -- gah!
I warned them about this hairball two years ago!
Namespaces: One instance where "stay the course" isn't boneheaded.
no subject
Date: 2009-02-17 01:28 pm (UTC)no subject
Date: 2009-02-17 11:06 am (UTC)So, when I found out that malcolm@ed was free, I grabbed it. Pointed it at my tardis.ed.ac.uk account.
Then, one day, they decided that everyone should have xxx@ed.ac.uk addresses - and they thought they'd just promote all the existing ones.. From multiple namespaces, into the ed namespace.
There were four people with malcolm@xxx.ed.ac.uk domains. So, there was a huge argument among three of them as to who should get malcolm@ed. The highest ranking professor won, and I was told to hand it over.
I did warn them this was dumb. He would spend YEARS getting email that was intended for me. No problem, they said, he can live with that.
Of course, it would have been irresponsible to sign up for lots of mailing lists etc before I handed it over, wouldnt it.. Or to post on some mailing lists I knew got trawled by spam bots saying "not using malcolm@ed.ac.uk" any more without obfuscating the address.. :)
A friend of mine has a similar story from Imperial College. He had andrew.smith@imperial... A new professor joined, and wanted that address - so he got bumped to andrew.smith.1@imperial... And the professor wonders why he gets email for the wrong person ALL the time.
Mashing namespac3es is dumb. Changing what the point to.. even worse.
no subject
Date: 2009-02-17 01:36 pm (UTC)no subject
Date: 2009-02-17 07:07 pm (UTC)I get angry asshats.
Count yourself lucky.