Re: System integrity...

1999-06-17 Thread Julian Gilbey
> I'm currently unsubscribed to -policy - I should be resubscribing soon, > when my free time increases to that amount needed to follow policy. So > apologies if this has been covered before... > > Doesn't an md5sums file need to have file lengths to work? I > understoood that there were known

Re: /usr/local stuff [Was Editor and sensible-editor]

1999-06-17 Thread Julian Gilbey
> One of the things that really annoys me about prepackaged > distributions is the way they tend to ignore everything that isn't in > packages. E.g., the only way to get non-deb GTK themes in > /usr/local/share/themes recognized by the GTK config is to link them > into /usr/share/themes, which is

Re: System integrity...

1999-06-17 Thread Chris Leishman
On Thu, Jun 17, 1999 at 01:49:30AM +0100, Julian Gilbey wrote: > > Just a brief note about the thread there: if md5sums are included in > packages, they will *only* be included for system integrity checks. > They serve *no* useful security purpose. Given this, the MD5 sums > themselves should be

[forwarded] FHS pre-2.1 draft #2 on web site

1999-06-17 Thread Daniel Quinlan
New FHS draft. If you have any comments, please use the FHS mailing list (to avoid having two separate discussions) or you may send them to [EMAIL PROTECTED] Thanks. I believe this draft still addresses the concerns that some Debian developers expressed about FHS 2.0 even though it's not the com

[forwarded] FHS pre-2.1 draft #2 on web site

1999-06-17 Thread Daniel Quinlan
(Wow, that was brilliant. Let's try that again.) New FHS draft. If you have any comments, please use the FHS mailing list (to avoid having two separate discussions) or you may send them to [EMAIL PROTECTED] Thanks. I believe this draft still addresses the concerns that some Debian developers e

Re: System integrity...

1999-06-17 Thread Brian May
In article <[EMAIL PROTECTED]> you write: >Just a brief note about the thread there: if md5sums are included in >packages, they will *only* be included for system integrity checks. >They serve *no* useful security purpose. Given this, the MD5 sums >themselves should be adequate for the integrity t

Idea: /proc/sys/fs/binfmt_misc/register in the policy?

1999-06-17 Thread Peter Van Eynde
Hi people, In the 2.2.x series there is a feature that allows the user to define shells for certain types of files. Selecting on extension or a few magic bytes. This could be useful for languages where the "#!" trick wouldn't work, like lisp. For the moment I have in a documentation-file: #!/bi

Re: weekly policy summary

1999-06-17 Thread Davide G. M. Salvetti
* MS => Manoj Srivastava Hi Manoj, I value your input, but there are some of your points that I can't really understand. MS> Freedom of software should come on its merits, not because on MS> Debian it is hard to find good (but non-free) software. This is something I agree with. However: De

Re: weekly policy summary

1999-06-17 Thread Davide G. M. Salvetti
* GB => Goswin Brederlow Hi Goswin, a little note just to emphasize patents don't make free software any bit non-free. So GIFs and the Gimp isn't really a good example (and gimp-nonfree is misnamed), choose please another one. The only one thing that can make a piece of software non free is

Re: Editor and sensible-editor

1999-06-17 Thread Charles Briscoe-Smith
[EMAIL PROTECTED] (Joseph Carter) wrote: >Not even. pico CANNOT be packaged for Debian! The best that can be done >is offer the source and let you build it yourself. If you do that, pico >will provide the editor alternative. If you want it to be the default >system editor, anybody else using yo

Message subjects

1999-06-17 Thread Julian Gilbey
Hello all! Now that Joey is posting a "weekly policy summary", *please* ensure to change the title before replying to the group regarding points raised. Otherwise, there will probably end up being multiple threads with the subject "Re: weekly policy summary", which is really unhelpful. Julian

Re: weekly policy summary

1999-06-17 Thread Manoj Srivastava
Hi, Since Alex is otherwise busy, I have to take his mantle up and represent what I think is an important balancing aspect of Debian. >>"DGMS" == Davide G M Salvetti <[EMAIL PROTECTED]> writes: MS> Freedom of software should come on its merits, not because on MS> Debian it is hard to

Re: /usr/local stuff [Was Editor and sensible-editor]

1999-06-17 Thread Goswin Brederlow
Julian Gilbey <[EMAIL PROTECTED]> writes: ... > > --- 7,10 > > > > umask 002 > > test -x /usr/bin/check-sendfile && /usr/bin/check-sendfile || /bin/true > > + test -f /usr/local/etc/profile && . /usr/local/etc/profile > > Eeks, no! There's no such directory as /usr/local/etc. /

Re: weekly policy summary

1999-06-17 Thread Goswin Brederlow
"Davide G. M. Salvetti" <[EMAIL PROTECTED]> writes: > * MS => Manoj Srivastava > > Hi Manoj, >... > As you see, this whole issue stems from this one question: «What do > you want Debian to be?». > > MS> What if it is true? What if the non-free software does indeed > MS> provide functionalit

Re: non-maintainers seconding proposals (was: Re: weekly policy

1999-06-17 Thread Jim Lynch
Hi! :) Goswin Brederlow <[EMAIL PROTECTED]> writes: > Shaleh <[EMAIL PROTECTED]> writes: > > -BEGIN PGP SIGNED MESSAGE- > > > > Agreed. And again, this is not personal, just Debian must retain control. > > Any interested in being a part of this can simply apply for maintainer > > status

Re: /usr/local stuff [Was Editor and sensible-editor]

1999-06-17 Thread ferret
I'd agree with something along these lines, just from my own experiences with my network, and more recently having to reload both of my surviving machines pretty much from scratch. (Good reason for having multiple drives) My own thought would be to keep the regular config files in /etc or /etc/,

Re: weekly policy summary

1999-06-17 Thread Marcus Brinkmann
On Tue, Jun 15, 1999 at 01:45:09PM -0500, Manoj Srivastava wrote: > > What if it is true? What if the non-free software does indeed > provide functionality missing in Debian? We bury our heads in the > sand and pretend that it does not exist? We do our users a disservice > and make it h