Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!sunybcs!pjg
From: pjg@acsu.Buffalo.EDU (Paul Graham)
Newsgroups: comp.arch
Subject: Re: Bloat costs
Message-ID: <26798@eerie.acsu.Buffalo.EDU>
Date: 31 May 90 05:41:10 GMT
References: <2793@crash.cts.com> <265D2FE5.2513@tct.uucp> <640@sibyl.eleceng.ua.OZ> <2662CE6C.3E68@tct.uucp>
Sender: news@acsu.Buffalo.EDU
Lines: 46

chip@tct.uucp (Chip Salzenberg) writes:

|According to ian@sibyl.OZ (Ian Dall):
|>Indeed, these days, an employer could be justified in berating a
|>programmer for wasting time reducing the size of a program instead of
|>increasing it's speed or functionality or getting it to market
|>earlier!

|Well, I'm not advocating a 16K limit on code size, either.  :-)

|However, I consider program size vs. programmer time to be a
|false dichotomy.  I would argue that, once gained, the skill of
|writing programs that begin and stay small makes a programmer
|more productive and valuable.  It's a part of the overall
|"think small, do one thing at a time" mentality that produced
|Software Tools, Research Unix and Plan Nine.  (But not System V
|or BSD, unfortunately.)

|The "think small" approach is, in my opinion, a big win over
|the "just write it the most obvious way" approach.  It seems,
|however, that the latter has a patina of respectibility among
|new programmers by virtue of all the bloated programs they use

woof.  i can't buy any of this.  as someone who spent his formative
years on an lsi-11 and thought you could do anything in 8k i find such
discourse on discipline silly.  i've run sunos, x, gnu and other
typical things all at the same time in 4M, and in 16M.  i'm not going
back.  nor do i write programs that are overly large.  i've used paper
tape based tools too, with 1k assemblers. i'm still not going back.

while not wishing to encourage waste in any particular thing i don't
really find it the least bit productive for various folks to keep
harping about how things are soooo huge, all the programmers are sooo
lazy and trillions of bytes are going to waste.  in the grand scheme
of things (and in this newsgroup) it sounds more like petulance than
discussion.  particularly when couched in an inflammatory or
agressive style generously leavened with loaded terms like bloat.

now if one wishes to discuss the relative merits of writing straight
forward code (the obvious way) versus bumming every last byte or
whether life-cycle costs are better served by long term maintenance
issues or short term resource issues it would seem another group
is in order.
--
"as a user i'll take speed over features."  hmmm, me i'll take both.
or was that bourbon over ice.