to group compress or not to group compress

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Stefan Monnier wrote:
>>> Speed of common operations is top of the list of metrics. We also rightly
>>> get assessed on total disk space usage, particularly for large projects,
>>> so we should include it as well.
>> I agree on both points. I don't really have a good idea of the relative
>> weight.
> 
> I think that size is not much of an issue in itself: it's only relevant
> to the extent that by reducing size you can reduce the time spent
> accessing the network/disk.
> 
> 
>         Stefan
> 
> 
> 

I don't think that is entirely true. It depends on the magnitude, but
people often use "this takes XXX MB" as a benchmark.

I realize in some senses it is only meant as a view into how performance
should end up, but it *is* used as a direct measurement on its own.

And personally, I wouldn't mind being a bit more compact. It isn't a
huge deal, but when I'm testing conversions, etc, I often have to have
multiple copies of a large project.

Anyway, it certainly isn't the highest priority, but I do think it is a
common measurement (if only because it is easy to do).

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAklk3c8ACgkQJdeBCYSNAAPeOwCgqHRrsqIfeDijLcrq100iXXh2
4DIAoKjJY/S6WAWRbwOaSsf4gGqu9VrX
=stjJ
-----END PGP SIGNATURE-----