Commuterwelt

[hirez]

I am by no means a morning person. I am even less a winter morning person. It seems entirely wrong to scramble out of bed while it's still dark and for the sun to rise over the TA building while I'm waiting for a bus. But that is what I'll have to do until I get my cycling legs back

Comments

[reddragdiva]

The gibbering insanity was required in OS X to get it officially certified as Proper Gentleman's Unix™.

My cheap Chinese MP3 is 2GB and is my friend and comfort, with the buckets of crap particularly including MP3s of old John Peel shows and of trance compilations, a favourite form of disposable amusement.

[samoth]

Remind me tomorrow and I'll dig you out the 'rosetta stone' web page for translating standard *nix things you'd expect into OS X's rather 'quirky' ideas about things. On the NeXTStep front, it's all a rather interesting lesson in 'winning by losing' :)

[hirez]

That will probably be dead handy.

In point of fact, the more I dig, the more it seems that someone's actually thought about how it should all fit together. Although I'd guess that a lot of that thinking was done by the original NeXT team, since NetInfo is still in there.

Unlike, say, the CFI-weenix wireless implementation, which is quite obviously a nasty lash-up.

[gaius_octavian]

Yes, it makes a huge difference when someone sits down and makes a "Unix" rather than cobbling together lots of different things made by lots of different people who have lots of different ideas about the way things should work and running it up the flagpole and seeing if anyone salutes. What do you mean -x does something else here?! And so on.

[blue_condition]

> because it's a lot like Unix in much the same way that HP-UX is.

All the deliberate faffery reminds me more of AIX, which is of course exactly what Unix would look like if its user interface was invented by Nazis with 3270s.

SMIT was, er, never a favourite tool of mine ;)

[hirez]

The only good word I've heard about SMIT was 'shite!'.

swinstall on HP-UX is more or less baroque as designed. Apparently it was part of a standardisation effort on the part of the SVRn vendors at the time. Only no-one else could be arsed with it.

[blue_condition]

I never liked SystemV in the early days, particularly when it displaced BSD from Suns.

During the early versions of Solaris 2.x a friend proposed the existence of a cunningly hidden daemon called upgraded that randomly moved files and directories around between minor OS patches.

Personally I think Sun had something in there that was very good at subtly altering nsswitch.conf..... ;)

[hirez]

Oh Fuck Yes.

We managed to entirely shag our main 2.4 box by installing the 'recommended patches' from Sunsolve in the dim & distant past.

Unfortunately, this was when all concerned were still pretty green regarding Unix and it took A Long Time for us to fix it.

[jendama]

[hirez]

I'll see if I can post it to you when I've finished the thing.

[blue_condition]

> an enthnographic report on a software development team

I've never actually met a software development "team" and I think I'd be worried if I met one.

I've met (and led) plenty of bunches of individuals working cohesively on the same software development project.

But I don't believe the culture I grew up in (people who started off on home micros in the late 70s/early 80s and spent too long in academia) tend to identify as members of or even voluntarily take part in teams at work. Work interactions tend to be entirely functional -- development is generally a series of transactions between actors with independent goals, not a relationship -- and the social network at work is often completely orthogonal to the org chart.

For me the idea of "team" implies a group of people with some esprit de corps and that seems to be alien to the kind of people I work with. There is certainly rarely any concept of subsumption of individual identity into the team.