PE> Rod Speed, Brenton Vettoretti and Frank Malcolm, you are
PE> sending me a 00000000.PKT, when you could instead be sending
PE> me a 00000000.MO0 which is compressed.  In the name of not
PE> being inconsiderate arsewipes, why don't you add a "zip -9k
PE> 00000000.MO0 00000000.PKT" to your batch files?

RS> I'd be surprised if it would even save a second.

PE> I wouldn't.

Depends on which case you are talking about.

RS> The modem is doing the compression instead of the archiver
RS> and the compression achieved is about the same. The current
RS> way is more thoroughly tested.

PE> zip vs V42bis?  I think I'll back zip on that one.

Yes, there isnt any doubt that an archiver will certainly compress
better than an on the fly compression like V42bis. However, what
actually matters is what the nett saving is, no one disputes that
there will be some. PKTs dont actually have the sorts of data where
you see the biggest differences between on the fly like V42bis and
archivers. And the other problem is that most people arent actually
uploading much in total PKT either.

PE> Anyway, today you sent me 58323 bytes.

And those others you listed would generally send much less.
And they dont pay any extra for extra time anyway.

PE> I zipped it here and it went to 23187 bytes. The 58323 transferred
PE> at 202%.  A large compressed file was transferred to you at 112%.

Those figures are pretty meaningless on such a small file.

PE> The 58323 took 20 seconds to transfer. So the time to transmit the
PE> 23187 would be 23187/58323*202/112 = 14 seconds, a saving of 6 seconds.

Yes, calculating it properly still gives about that answer, 6 secs.
My total session time is normally over 3:30, a saving of 1c. Whoopy do.

PE> Why don't you zip up the *.pkt into a file called "fred.zip" and
PE> transfer that at the same time one day, and I can tell you from the
PE> Binkley logs what happened.

No real need to do that, I know I get about 1650cps on an archive file,
3000cps is typical on the uncompressed files, I can actually archive up
all the outgoing PKTs which have been sent and get a rigorous analysis
of the compression ratios that say PKZIP gets with them. I could even
without too much trouble use the FD log to calculate relative compression
ratios that V42bis and PKZip actually achieve on each individual PKT too.
I might save 1c a day. Whoopy do. I would actually save more money by
dropping some echoes I only read occasionally.

--- PQWK202