On Wed, 2004-12-22 at 15:42, Eric Smith wrote:
> Guy wrote:
> > Because Bliss-11 was a cross compiler under TOPS-10 (and was a
> > multiple pass compiler) it would have to swap out in order to
> > bring in the next pass. This was done for every routine
> > (function for you 'C' types). To keep the swap-in/swap-out as
> > fast as possible, you tended to write small (1 page) routines.

>
> I don't understand that reasoning. It seems like you'd want to write
> routines as large as possible in order to minimize the number of swaps.


The issue was the fact that large(r) swap-ins were lower priority than
small(er) swap-ins.

This was on TOPS-10 running on a KA10 with between 100-200 users.

>
> Of course, writing large routines tends to be inconsistent with
> maintainability, though as with all generalizations there are exceptions.


As I recall (dim memory) my typical Bliss-11 routines were on the order
of 30-40 lines max.
--

TTFN - Guy
----------
To unsubscribe (or subscribe) from (to) this list, send a message to
info-pdp11-request@village.org, with the first line of the message
body being "unsubscribe" or "subscribe", respectively (without the quotes).