PE>> Sorry, what you need is infozipp.zip which is the OS/2 version.

 PM> Yep, thanks, infozip works fine. The only problem is that it records full
 PM> path names, so when it extracts the packets they went into directory
 PM> \SQUISH\BINKLEY\OUTBOUND.SQ instead of \BINKLEY\FILE so Squish couldn't
 PM> find them. I have put the -J (ignore directories) option on the unzip for
 PM> tomorrow to see if it fixes it. You may also want to use -J on the when
 PM> you zip.

Hmmm, I'll have to do some testing on that (I did some testing but I got an
internal logic error using -j).  I reported the internal logic error on a
previous version as well!  Dear oh dear.  When I get the source I'll see if
I can fix the bloody thing myself.

 PE>> compressing your mail with zip -9 which gives tight compression and
 PE>> can be decompressed with pkzip 2.04 or infozipp.  I am compressing
 PE>> mail on Maximus with zip -1 which gives better compression than pkzip
 PE>> 1.1 for OS/2, and much faster.

 PM> Does -9 help much? I guess every bit helps, especially for STD callers.

Something like 10-20% from memory.

 PE>> I have asked David to send me the source code, so that I can compile
 PE>> zip for OS/2 under Borland with 486-specific instructions etc which
 PE>> may help my cause!

 PM> We'll see how portable their code really is! Remember, don't send me the
 PM> 486 version as it probably won't run on a 386.

Good thinking!

 PE>> P.S. I have file-attached it to you.  I forgot to put HOLD on it, so
 PE>> it tried sending it to you some time ago.  Dunno if you answered the
 PE>> phone or not though!  I got No Carrier after a long time.

 PM> No, I was out all day Saturday. Maybe you should make the default
 PM> attributes hold.

Yeah, I'll look that one up next time I'm fixing up Squish.


Paul

--- GoldED 2.40