[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bastille linux : secure linux distribution



On Thu, Jun 03, 1999 at 11:40:12PM +0200, Raphael Hertzog wrote:
> The distribution will be based on Red Hat for several reasons: 
[...]
>    2.The packaging  format is good (does one critical thing which Debian
>    packages don't and which we need: multiple patches per package) though
>    not perfect.  (Nothing is, though.) 

Someone please ask these people to retract this vile slander.

Debian *CAN* support multiple patches per package *IF* the maintainer
chooses to do things that way.  My proof?

apt-get source xfree86-1

(This implementation of source handling is thanks to Adam Heath.)

>    3.The install process is GPL'ed, so we can muck around with it to our
>    heart's content. (The SuSE installer is NOT, AFAIK, and doesn't even
>    seem to qualify as open source, according to what I've heard.) 

Too bad we doesn't use free software for OUR installer.  Wait a second...

>    4.I've mucked with the installer before in some ways, so I'm at least a
>    little comfortable with it. :-) 
>    
>    5.It's reliably multi-platform. While x86 is our primary target, it's
>    certainly not our only target.  

I guess I've been hallucinating those reports about X running on
Debian/m68k, Debian/Alpha, Debian/PowerPC, and Debian/SPARC.  (I have the
crazy notion that if you've got X working, your system is probably mature,
stable, and fairly reliable.)  Debian/Hurd-i386 and Debian/ARM also appear
to be in the pipeline.

>    6.The installer appears to be the best for installations of multiple
>    machines, a must for organizations with large numbers of similar boxes.
>    The install also appears to have most of the pieces needed for CD
>    customization. 
> ------
> 
> You know what is important for Debian now ... :-)

Point one is irrelevant by definition ("Red Hat is the most popular").  It's
just bandwagon jumping.  Points 2, 3, and 5 are just as true for Debian as
they are for Red Hat.  Point 4 is also not an objective criterion -- it
just points to the developer's pre-existing pro-Red-Hat bias.

That leaves us with point 6, which is a vigorous topic of discussion among
the Debian developers.  Wonder of wonders, people are even coming up with
CODE to implement it.

> But that's not the intent of my message, many of you may have already 
> read the announce on bugtraq. Although they are basing their work on
> RedHat they want to share their experience and we should from time to
> time take a look on what they did in order to incorporate the best 
> changes ...

It doesn't look like they've taken a very close look at Debian at all,
since half their list of reasons are blatant falsehoods, and an additional
third are not even criteria based on technical merit.

Wow, my random .sig selector is manifesting AI again...

-- 
G. Branden Robinson              |
Debian GNU/Linux                 |   Never attribute to malice that which can
branden@ecn.purdue.edu           |   be adequately explained by stupidity.
cartoon.ecn.purdue.edu/~branden/ |

Attachment: pgpXkd0glK2K7.pgp
Description: PGP signature


Reply to: