Re: Is Systems Software Research Irrelevant

[Daniel Weinreb <address@hidden>]

Kragen Sitaker wrote:

>Daniel Weinreb writes:
>
>>Well, as a matter of fact, Windows XP is vastly superior to the
>>Windows available in 1990!
>>
>
>Well, I might be selectively blind, but it looks to me like Windows XP
>is mostly superior to Windows 3.0 by adopting OS features from the
>1970s, UI features from the 1980s, and Web features.  The Web features
>are the only ones that tend to disprove his thesis, and even they
>didn't come from "systems software research".
>
I was just reacting to his claim that nothing has changed. Things in the 
outside, commercial
world do change.  If your point is that things take a long time to move 
from the academic
world into the commercial world, well, yes, they do. If you go back and 
look at the set
of slides by Richard Gabriel that were part of the "advanced reading" 
for the LL1 workshop,
there's a lot of good information about this topic.

>[quoting Rob Pike]
>There has been much talk about component architectures but
>only one true success: Unix pipes.
>
>
>I'd be interested in your views on this.  From my POV, the Web is a
>successful component architecture, as is COM.  What else is?
>
Well, it depends a lot on how you define your terms: what does it take 
for something
to be "successful"? You could easily say that anything based on Lisp was 
not successful
because Lisp was not successful, in the sense of commercial pervasive 
success that's
still ongoing.

I don't think "the Web", per se, is a component architecture at all. 
 HTML screen-scrapers
don't count.

Unix streams, with their total lack of both syntax and semantics, do not 
strike me as a
component architecture to be proud of.  Yes, you can do cute hacks with 
pipes and so
on, but that's not what I mean when I say "component architecture".

-- Dan

>
>