On 9 Jul 2009, at 10:32, Dr. David Kirkby wrote:
> > Ralf Hemmecke wrote: >> Thanks. That makes sense. >> >> But as you see, some people (me) ask why these .tar files are not >> zipped. I guess, a small note on the webpage would clarify that. >> >> """ >> The .tar files below contain collections of .spkg files (which are >> basically .tar.bz2 files). >> """ >> >> Or you put somewhere on that page a link to a .txt file that lists >> the >> contents of the .tar file. Some project add a manifest.txt file to >> the >> distribution. >> >> Just my 1/2 cents. >> >> Ralf > > I would agree with Ralf. The fact the file is already compressed is > not > obvious. In fact, since some of it is not compressed, and each .spkg > must have some overhead in compression, has anyone ever tried > compressing the tar file? It might save a few bytes, and reduce > confusion. (I'm aware it could also increase the size by compressing > compressed files.) No I have not tried it myself - perhaps I will > later. > I have tried compressing the sage.*.tar files using gzip and bzip2. The compression is insignificant, at least when using the default compression levels. % ls -l *tar* -rw-r--r-- 1 XXX staff 215244800 Jun 19 08:46 sage-4.0.2.tar -rw-r--r-- 1 XXX staff 214183786 Jun 19 08:46 sage-4.0.2.tar.gz -rw-r--r-- 1 XXX staff 215169684 Jun 19 08:46 sage-4.0.2.tar.bz2 But, there is the big advantage that if you have a compressed file you are alerted right away if the file is corrupt. I have had more than one occurrence where the tar file was bad, and only found there was a problem after several hours of compilation attempt ended in failure. I would prefer that the tar ball was compressed with gzip, so I can know right away whether it is good or bad. gzip is a better choice than bzip2, as it results in slightly smaller file sizes, and it is much, much quicker to compress the file, at least on my ancient PPC machine. -- Kevin Horton Ottawa, Canada --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
