Re: New tazpkg installs busybox-pam for tazpkg

[GoKhlaYeh <gokhlayeh@xxxxxxxxxx>]

There's two things to consider :
* As Julien said, is the decompression tie better than lzma ? In my benchmark I didn't test it but theorically it's near the same for all tools. The good compression ratio + good decompression time of lzma is what makes tazpkg install so speed.
* The memory needed for decompression. lzma tarball likes the ones we provide are theorically decompressable with 16Mb of RAM - I'm not sure if it's exactly that but I think to. SliTaz sometimes run with 32Mb of RAM; I don't know if some user use less but's it's possible.

I havn't take the time to benchmark 7z as I'm currently coding; I just tested it speedly on eet source and it's not better at compression ratio level, what tool/command-line have you use to compress?

On Sun, 12 Dec 2010 16:36:12 +0000
Christopher Rogers <slaxemulator@xxxxxxxxx> wrote:

> i used time '7zr a VirtualBox-3.2.10_OSE.7z VirtualBox-3.2.10_OSE' on dual
> core amd 4450e cpu with 4gb of ram. Here are my times:
> real   1m 51.17s
> user  2m 47.07s
> sys    0m 2.14s
> 
> 7zr a VirtualBox-3.2.10_OSE.7z
> real   0m 31.57s
> user  0m 6.56s
> sys   0m 4.72s
> 
> The real time is the one you want to look at. The user and sys are more
> about os timing.
> 
> I hope this helps.
> On Sun, Dec 12, 2010 at 3:36 PM, Alexander Medvedev <devl547@xxxxxxxxx>wrote:
> 
> > As for lrzip - it heavily depends on memory usage and compression
> > algorithm.
> > Also it's multithreaded and is developed for multicore systems.
> >
> > You can see it's benchmarks here -
> > http://ck.kolivas.org/apps/lrzip/README.benchmarks
> >
> >


-- 
GoKhlaYeh <gokhlayeh@xxxxxxxxxx>


---
SliTaz GNU/Linux Mailing list - http://www.slitaz.org/