Verily I say unto thee, that BearItAll spake thusly:
> Roy Schestowitz wrote:
>> Xandros Linux Server First to Receive LSB Certification
>>
>> ,----[ Quote
>> | Xandros today announced that Xandros Server 2.0 is the first product
>> | to be certified by the Linux Foundation through use of the LSB
>> | Distribution Testkit (LSB DTK).
>> `----
>>
>>
> http://www.linuxlookup.com/2007/apr/24/xandros_linux_server_first_to_receive_lsb_certification
> Would it be pedantic to say that if this is the first ever test and
> therefore the first ever pass then it can't really be called a standard.
>
> I tried to see what exactly this certificate covers, but haven't come up
> with anything that makes sense.
In the words of the Linux Foundation:
"The LSB defines a binary interface for application programs that are
compiled and packaged for LSB-conforming implementations on many
different hardware architectures. Since a binary specification shall
include information specific to the computer processor architecture for
which it is intended, it is not possible for a single document to
specify the interface for all possible LSB-conforming implementations.
Therefore, the LSB is a family of specifications, rather than a single one."
http://refspecs.freestandards.org/LSB_3.1.0/LSB-Core-generic/LSB-Core-generic/overview.html
Or IOW, the LSB is an attempt to define an official standard for what
precisely constitutes a Linux system.
As a trivial example, one of the packages I worked on at Fedora had a
legacy bug where its helper scripts were being installed into
%_sysconfdir (/etc) rather than %_bindir (/usr/bin). The LSB dictates
that %_sysconfdir is for configuration files only, not executables
(binary or otherwise).
It is important that distro vendors adopt and maintain LSB standards,
especially WRT security measures such as SELinux contexts, which
determine the ACL scope of objects based on (amongst many other things)
their location within the directory hierarchy (as well as ownership, and
more traditional things like permission bits). In the above case, for
example, if that package's scripts *had* been installed into
%_sysconfdir, and SELinux was active and set to enforce, then those
scripts would not be executable in their current location, even by root.
Read the LSB for yourself:
http://refspecs.freestandards.org/LSB_3.1.0/LSB-Core-generic/LSB-Core-generic/book1.html
For variations based on other arch's (e.g. x86_64 also contains an
additional /usr/lib64 hierarchy), go here:
http://www.linux-foundation.org/en/Specifications
As for Linux Foundation's right to call this a standard, it is something
which they wish to be a standard, rather than claim to be an established
standard. Although the *principles* are certainly well established.
> The main site tells you about the software that does the test and the
> web interface to it. But not what the tests are.
Try it yourself, and see how well your distro compares:
http://www.linux-foundation.org/en/Downloads
--
K.
http://slated.org
.----
| I found [Vista] to be a dangerously unstable operating system,
| which has caused me to lose data ... unfortunately this product
| is unfit for any user. - [H]ardOCP, <http://tinyurl.com/3bpfs2>
`----
Fedora Core release 5 (Bordeaux) on sky, running kernel 2.6.20-1.2312.fc5
23:04:40 up 8 days, 20:36, 2 users, load average: 0.45, 0.51, 0.44
|
|
