Subject: Re: Is gcc slow? Or is our gcc slow?
To: None <juengst@saph1.physik.uni-bonn.de>
From: Curt Sampson <curt@portal.ca>
List: current-users
Date: 04/09/1996 17:45:36
On Tue, 9 Apr 1996 juengst@saph1.physik.uni-bonn.de wrote:

> Why is -O in the system makefiles?

Because it won't seriously inconvenience any of the machines that
NetBSD runs on? There are a fair number of Sun 3/50s out there (I
should know--I own five of them :-)) and, while those in search of
better performance will probably think of putting -pipe in first
thing, those who are not and are trying to get a system up and
running on one of the smaller machines may be excused for not
checking to see if -pipe is there and removing it if it is.

> "Convenient way" means you have to know this missing option and hack the
> system.

And having to know of the option and hack the system if you don't
want -pipe is more convenient?

> MFS is really stupid. You are wasting resources for the most of time.

No, the point is that you are *not* wasting resources for most of
the time. Most of the time /tmp has almost nothing in it. Therefore
you can use that memory for other things. When /tmp does fill up,
it uses free memory first, and then goes to swap.

> "/tmp" and "/var/tmp" are stupid at all. Common temporary resources for
> all users are the best way to produce problems.

So you think that on a 24-user system giving each user 1 MB of his
own scratch space is better than just assigning a 24 MB shared pool?

> If there is just one user who made a mistake and allocates the entire
> space, then other users are not able to compile their program (and most
> don't know why).

Yes, much better that none of the users can compile their program in the
first place because there isn't enough space to begin with.

cjs

Curt Sampson    curt@portal.ca		Info at http://www.portal.ca/
Internet Portal Services, Inc.	
Vancouver, BC   (604) 257-9400		De gustibus, aut bene aut nihil.