Re: Policy wrt Important (was Re: dc and bc in Important?)

[Galen Hazelwood <galenh@micron.net>]

John Goerzen wrote:
> 
> OK, then I suspect the policy is at fault.  (BTW, I checked it out and
> I did find dc and bc on SunOS -- I had not known these programs were
> on other OSs.)
> 
> By the current definition of Important:
[snip]
>  * lilo should not be there because lilo is not part of UNIX

Now you're being silly.  Although programs which are "Standard and
Expected" are Important, the category can also include other things if
they are, well, important.

> And:
>  * gcc should be in Important because everybody expects a C compiler

gcc is Just Too Big.  With the singular exception of Perl, I think
development tools and really large packages (>2 Meg) should be excluded
from Important.

>  * libc5-dev should be there because everybody expects working
>    header files
>  * make should be there, I expect a working make in any Unix

Again, development tools don't count.

>  * lpr should be there, it is standard with just about any Unix

This is sort of a grey area.  Yes, lpr is standard and expected, but
it's useless if you don't have a printer...and many boxes don't.

>  * netbase and netstd should both be there, they are standard
>    on Unix

I actually think that these packages (or at least netbase) _should_ be
upgraded to Important.  Other opinions?

>  * csh/tcsh should be there (again, standard on various Unices)

I'm tempted to agree with you here...but no.  My judgement here is that
c-shells, though standard and perhaps expected, aren't _quite_ standard
enough.

>  * The list goes on...
> 
> Basically, it seems that this policy doesn't quite apply correctly.

The policy is a "rule of thumb", nothing more.  Attempting to apply it
mechanically produces ridiculous results.  It requires a judgement call
on the part of the packager.  You can disagree with a particular
decision, but that dosen't cast aspersion on the policy itself.

BTW, I was the one who suggested to James that bc and dc be made
Important.  I have been using dc for on-the-fly math for nearly as long
as I've been using unix, and missed it when it wasn't there.

--Galen


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org . 
Trouble?  e-mail to templin@bucknell.edu .