🚀 go-pugleaf

In <3nqo8n$ov6@csnews.cs.colorado.edu> Tom Christiansen <tchrist@mox.perl.com> writes:

>In comp.lang.perl, 
>    gert@greenie.muc.de (Gert Doering) writes:
>:Tom Christiansen <tchrist@mox.perl.com> writes:

>:Ummm... on my SCO Unix machine (486dx2-66, 40 Mb RAM), it takes about 5-10
>:seconds to start up *per invocation*... now *that* is what I'd call slow
>:startup :(

>Then something is really broken.  Taht's just brain-dead.  You're
>running wtih more RAM than I am.   I refuse to believe that my 
>Sparc Classic is 100-200 times faster than your 486dx2-66.  Bug
>in SCO, maybe?  Who'd'a thunk it?

I partially agree with Tom.  I don't think it's inherent in SCO, but
something is definitely hosed on Gert'ssystem.

It doesn't take 10 seconds to load perl from the Skunkware CD on 
my SCO system at home (486dx2-66).   Once it's in-core and off the 
CD, subsequent invocations are very fast.

Since Tom didn't publish his means of test, I'll offer this low rez
test.  Since we're talking about a couple of orders of magnitude 
difference, it should be adequate.  

On my 486/50 (no dx2)

$ time perl < /dev/null

real    0m0.15s
user    0m0.02s
sys     0m0.10s

[ and once it's loaded in the disk cache, subsequent loads are around.. ] 
(robertl) amsg:/usr/local/bin
$ time perl < /dev/null

real    0m0.08s
user    0m0.02s
sys     0m0.06s

I use perl a fair amount on SCO, and 10 second load times would have driven
me nuts a long time ago.   Gert, perhaps you should pull the Perl binaries
from somplace like ftp.celestial.com.


--
Robert Lipe    Sr. Software Engr.    Arnet Corp.     robertl@arnet.com