🚀 go-pugleaf

(Love the message-id's and reference lines ... keep 'em coming.  David
needs your headers for the record books ...)
 
bob@allosaur.cis.ohio-state.edu (Bob Sutterfield) writes:
>  While I'm impressed with gnews' flexibility, I'm depressed by its
>  speed.  I've switched to GNUS 2.10.

(Umm, 3.1, bob?  I did get it from you ...)

I'm afraid I have to agree with Bob.  Why afraid?  I dunno ... people
always say things like that.  In any event, Weemba sent me a message a
short while ago that said, paraphrased, 'Gnews 2.0 has none of the
problems you mentioned about GNUS.'   (Quick aside, I wish both of
these weren't pronounced the same way ...)

Well, he was mostly right about that.  And unfortunately no one else
addressed the problems I mentioned in an earlier message.  So, I got
Gnews and tried it out.  It was, sadly, MUCH larger as a package than
GNUS (for which I couldn't afford the space to byte-compile, too) and
somewhat slower too.  That's where it lost the race.  I'll give it
points for flexibility, but I am pretty much settled on GNUS right
now.  I like the way it does things and it does do them faster and in
less space.
 
I also, by the way, did a benchmark (very unscientific one, to boot)
with GNUS regarding spool files versus NNTP ... our local USENET host
is rpi.edu, which is also mounted as /usenet on pawl.rpi.edu.  From
pawl.rpi.edu I ran GNUS both via NNTP and via the /usenet spool files.
The results were that NNTP was noticeably faster, but, as I said, they
were unscientific tests and the speed difference could be attributed
to many different things and not necessarily reflect a consistent
situation. 
 
If anyone comes up with some answers/fixes regarding automatic
subscription of groups, translations of kill-files and anything else I
mentioned in that earlier posting, please let me know.  Thanks.
 
Dave
--
--
      tale@rpitsmts.bitnet, tale%mts@rpitsgw.rpi.edu, tale@pawl.rpi.edu