?SPAM? - Re: [tex-live] Re: [tug-board] Any news of DVDs ?

[Philip TAYLOR]

Reinhard Kotucha wrote:

 > You certainly will not understand what I mean unless you had been in
 > Vietnam (12 PCs connected to a single voice modem in an internet cafe
 > in Sa Pa).

Just returned from there, Reinhard (Viet Nam, I mean, not
Sa Pa on this occasion); ADSL ubiquitous :-)  But no-one
in Vietnam is going to try to download the TeX Live DVD from
an Internet cafe in Sa Pa, nor would anyone in their right
mind try to download it with less than broadband, in which
case 587Mb v. 616Mb is a noise-level difference.

> I absolutely agree with Karl.
> 
> And I absolutely agree with the phrase "I don't think it's our
> job...".  

And I absolutely disagree with both.  If you use standard utilities,
I accept there is no reason to document them; if you use something
as non-standard as BZ2, then there is every reason to document it.
> 
> I think that even Windows users are smart enough to pass an error
> message to google when they don´t know how to proceed.

The instructions tell them not to bother if they don't know how
to proceed.  That's gratuitously offensive and will lead to
others doing what I did : downloading the ISO file at great
cost to everyone because no clue is given as to how to undo
BZ2 compression.

   And it should
> not be too difficult to install software which can extract tar.bz2
> files even under Windows.

Chr1st, you're now telling me that under the BZ2 there's another level
of indirection ?  This is insane.  ZIP compression does both
seamlessly.
> 
> The compression algorithm ZIP uses is amazingly inefficient if a lot
> of small files are involved (for instace the tex4ht font files).  In
> certain cases zip files are ten times larger than tar.bz2 files.

But overall, on TeX Live Live, the difference is less than 5%.
noise-level, IMHO.
> 
> Sticking with ZIP is as annoying as sticking with 8.3 filenames.
> Sorry, but in which world do we live today?

A world in which ZIP compression is uniquitous and BZ2 niche.
That's the /real/ world, Reinhard, not some ideallistic view
of how it might be.
> 
> Efficiency of compression algorithms does not matter very much if you
> have DSL (and maybe even a flatrate).  But many people have to live
> with their slow and expensive internet connection.

And they should /not/ be downloading the TeX Live ISO if that is all
they have; the probability of success is (I estimate) less than 1%
over a 56K modem.

** Phil.