[1798] in SIPB bug reports
freeze, patch01
daemon@ATHENA.MIT.EDU (Mark W. Eichin)
Sat Mar 30 10:05:07 1991
Date: Sat, 30 Mar 91 10:04:31 -0500
From: "Mark W. Eichin" <eichin@ATHENA.MIT.EDU>
To: Jonathan I. Kamens <jik@ATHENA.MIT.EDU>
Cc: bug-sipb@ATHENA.MIT.EDU
I've installed patch 01 for freeze, changed the Imakefile to support
statist as well, cobbled together a man page for statist out of the
README file (I should really write a real one and send it back; this
will do for now, though.) All of the changes went under RCS. Important
features of this patchlevel are: my bug fixes for big-endian architectures
(scanf("%d") into a short only works on little-endian. tsk tsk, this wasn't
tested when it was first installed...) and a new +n,n,n,n,n,n,n,n command to
allow the numbers generated by statist to be specified directly (for optimal
huffman codes in the second stage compression. Usually saves less than 1% for
text files, but binaries and tar files often benefit more. (You may not
realize it, but freeze can often get 30-40% more compression than "compress"!
it's worth using for that alone... and although it's based on Unix compress,
I've suggested to the author that he look into using Y-compress for the first
stage instead.)
What's the right way to do an install? And why doesn't the Imake
generated makefile have a "make links" target? (This patchlevel added
a patchlevel.h file, as well as my statist.man. Consider this a bug
report on the Imake config files...) Most of this should be in
/mit/sipbsrc/src/MAKEFILE_RULES, along with the instructions already
present on how to create a new project.
Little things which should be "written down":
Imakefile use - ComplexProgramTarget vs. ComplexProgramTarget_1
remember to add the directory to src/Makefile (freeze wasn't...)
attach what? where? and do what make? for each architecture...
what AFS commands are needed (or who needs to be tweaked? or at
least, how long do we wait for it to happen automatically?) for this to
appear in the SIPB locker?
Many of these things have changed since I last did serious
maintenance of the SIPB locker; while jik is doing a great job of it, he
occasionally leaves for Ohio... and this should really all be documented (so
the jik script can display it? :-)
_Mark_ <eichin@athena.mit.edu>
MIT Student Information Processing Board
