[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ProgSoc] ipod battery



On Tue, 2006-06-13 at 13:27 +0000, John Elliot wrote:

> Roland Turner wrote:
> > On the latter, rather more reasonable basis, UTS has long since ceased
> > being a technical school. Indeed its largest constituent part at
> > establisment (NSWIT) had arguably already started to transcend being a
> > technical school.
>
> I'm not so sure. My last experience of UTS was sitting in a room being 
> told to click about on some bogus Cicso (TM) University Degree In A Box 
> (R) material while listening to a tutor attempt to explain a bitmask to 
> a dumbstruck second year.

It's not clear which of my assertions you feel that you're contradicting
here:

- That an institution's scope (of fields taught/researched) is a more
reasonable basis for deciding whether it warrants the title "university"
than some ill-defined measure of goodness.

- That, on that basis, it is no longer reasonable to refer to UTS as a
technical school.

- That NSWIT was the largest constituent part of UTS at UTS's
establishment.

- That NSWIT (with its physics, pure mathmatics and business students)
had already started to transcend being a technical school at that time.

You also appear to be repeating the error of assuming that your
experience in a single class/course is in some sense representative of
the university as a whole. (The same error in your previous post was to
assume that your encountering one boorish academic was a reasonable
basis for deciding that UTS does not warrant use of the term
"university".)

> I've actually been to TAFE (I was training as a chef), and frankly 
> TAFE's standards felt higher by comparison.

Further, you appear to be repeating the error of assuming that this
conversation is about standards, rather than scope.

> Who knows, maybe I have a chip on my shoulder? :)

So it would appear :-)

> p.s. Since our recent conversation I'm up to page 220 of Advanced 
> Programming in the UNIX Environment, Second Edition. I know what 
> O_CREAT, and dup2, and *everything* is now. :) Still haven't got to the 
> bit where they explain how the kernel goes about ensuring atomicity of 
> writes without locks,

It doesn't. That's an application-level problem. flock(2) is one
convenient way of tackling this, but so are all of the SysV IPC
mechanisms (semaphores, message queues, shared memory) and so is
assigning a single process control of a file and having it listen on a
socket; the kernel doesn't seek to mandate a specific co-ordination
mechanism. Nothing stops an uncooperative application from buggering up
the sequence of writes, but then, once a process has write permission,
nothing stops it from truncating the file either.

> but I understand that the file isn't 
> pessimistically write-locked by a process (as it is in Windows). There 
> has been some mention of record locks, but I'm not up to the details of 
> that yet. Overall, it's been very enlightening reading.

I'm glad to hear that it's making some sense!

- Raz


-
You are subscribed to the progsoc mailing list. To unsubscribe, send a
message containing "unsubscribe" to progsoc-request@xxxxxxxxxxxxxxxxxxx
If you are having trouble, ask owner-progsoc@xxxxxxxxxxxxxxxxxx for help.