2009/12/2 ????? ???????????? <olga.kryzhanovska at gmail.com>: > 1. I like to get the names of the people who attended to lunch and > discussed this. > 2. Converting scripts to executables reduced observability. But: > Converting scripts to compiled scripts does not reduce observability > because ksh -x binscript, truss, dtrace u.s.w. all continue to work. > 3. The size reduction is very important for the Live CD and USB > sticks. The change I was to propose is the experiment to prove that > the basic idea is solid. > 4. If the experiment works more scripts get converted and the real > size reduction would benefit the Live CD.
I don't have a strong feeling one way or another on this, but I do have the following observations and questions. - Shrinking anything can be of great help in live media situations where getting a megabyte or two back can allow for the inclusion of another useful utility. - I'm not sure about formats such as iso9660, but with UFS or ZFS there seems to be no benefit to shrinking a file from 500 bytes to 100 bytes. This still occupies a 512 byte disk sector. - It looks like disk sectors will be getting bigger (http://arc.opensolaris.org/caselog/PSARC/2008/769/final_spec.txt) - How compressible is bytecode compared to uncompiled code? That is, once it is put into a compressed disk image format, how much space is really being saved? -- Mike Gerdts http://mgerdts.blogspot.com/
