-BEGIN PGP SIGNED MESSAGE-
It has been proposed that that is not the case, that conffiles
are an independent classification.
If that is the case, then under one interpretation (a) becomes
meaningless, since it only applies to the small subset of files that
are
Zed - Files listed as conffiles are still constrained by any other
Zed rules that may apply to them; e.g. configuration files that
Zed are conffiles must still go in /etc, variable data files
Zed still go in /var, and so forth.
Hmmm. Ok.
Zed After
Hi,
Marcelo == Marcelo E Magallon [EMAIL PROTECTED] writes:
Marcelo wmaker's author likes GNUstep. He made wmaker
Marcelo GNUstep-happy. He wants to install things in
Marcelo /usr/GNUstep/Defaults or /usr/share/GNUstep/Defaults, and
Marcelo stuff like that. I modified it to use
-BEGIN PGP SIGNED MESSAGE-
Zed I think that mindset is not in the best interests of the Debian
Zed project.
I think I disagree. Letting developers who can't write shell
scripts can in no way be in the interests of the project, one of
whose goals is *excellence*.
Let me
[You (Manoj Srivastava)]
Zed After having lurked in debian-devel for a while, I suspect that
Zed Manoj will object that developers for Debian need to be sufficiently
Zed proficient in writing shell scripts and whatever else that they can
Zed deal with this on their own.
Zed I think that mindset
: Zed After having lurked in debian-devel for a while, I suspect that
: Zed Manoj will object that developers for Debian need to be sufficiently
: Zed proficient in writing shell scripts and whatever else that they can
: Zed deal with this on their own.
: ;-)
: Zed I think that mindset is
Hi,
Adam == Adam P Harris [EMAIL PROTECTED] writes:
Adam Manoj, you miss the point that creating conditions where errors
Adam are possible makes errors inevitable, statistically speaking,
Adam for these errors to occur. Crippling, nasty errors.
And removing the possibility of errors
Hi,
Guy == Guy Maor [EMAIL PROTECTED] writes:
Guy Manoj Srivastava [EMAIL PROTECTED] writes:
In my opinion, I would ask for conffiles to be exactly the set of
configuration files, plus exceptions decided by consensus no the
mailing lists
Guy That's a bit much. I agree that conffiles are a
Hi,
Ok, that's it. You said the same thing about me on IRC, and I
let it go. But in this forum, I feel I should respond. I resent the
implication that I am a bare boned minimalist hick who is not upto
date on tools and modern software.
I would have preferred to not voice this
[EMAIL PROTECTED] (Gregor Hoffleit) wrote on 21.02.98 in [EMAIL PROTECTED]:
Vincent Renardias [EMAIL PROTECTED] wrote:
On Sat, 21 Feb 1998, Christian Schwarz wrote:
distributions. Frankly speaking, this looks a bit like the
`C:\WINDOWS'
approach: just add a new directory for your
[EMAIL PROTECTED] (Manoj Srivastava) wrote on 22.02.98 in [EMAIL PROTECTED]:
I personally do not find the info keys incomprehensible, but I
spent some time learning how to use info, and in a number of cases
info is my preferred documentation format.
There's no logic behind those key
[EMAIL PROTECTED] (Joey Hess) wrote on 22.02.98 in [EMAIL PROTECTED]:
Manoj Srivastava wrote:
As to the conffile not being a configuration file, I think you
are wrong on that issue (you doubtless feel I am wrong). I think this
is a time for some deadlock breakage to occur.
Yes,
[EMAIL PROTECTED] (Rob Browning) wrote on 22.02.98 in [EMAIL PROTECTED]:
It's an old observation, but how to you recursively search an html
tree for the info you want (like Ctrl-s in info, Meta-s in emacs, or
/ in manpages with less?)
And the old answer is, of course, that just as with other
Hi,
I should apologize to the people who write helper scripts for
my outburst; I regretted it as soon as I had sent it.
I have the deepest respect towards the authors, who are trying
to do a very hard job.
I should keep a closer rein on my temper. I still do not use
Manoj Srivastava wrote:
I should apologize to the people who write helper scripts for
my outburst; I regretted it as soon as I had sent it.
I have the deepest respect towards the authors, who are trying
to do a very hard job.
Thanks.
Personally, I don't care if people use
Hi,
I was just reading the proposal, an I think I have gotten
myself confused. Let me try to see what an emacs add on package
maintainer needs to do.
1) /usr/lib/emacs/packages/emacsen-install/vm
/usr/lib/emacs/packages/emacsen-remove/vm
So, since macs =
Hi,
Kai thinks that emacs key bindings have no logic behind them;
I find there to be an elegant consistency there. He hates the
bindings, I have them coded into my DNA. He finds things easier to
search for in man pages, I find the multiple document scanning
features of info wondeful.
Hi,
Hamish == Hamish Moffatt [EMAIL PROTECTED] writes:
Hamish The problem I see is that with the source in main, although
Hamish you can build www-pgsql just fine, calling debian/rules
Hamish binary would have to make www-mysql too (if the same sources
Hamish are to be used), and making
On Tue, 24 Feb 1998, Hamish Moffatt wrote:
Some sources produce both main non-main packages. www-sql, for example,
produces www-mysql (used to be www-sql) for contrib (since mysql is
non-free); currently the www-sql source is in contrib too. Now www-sql
can be built for postgresql too, so I
Shall they be gzipped or not? (Up to the copyright file, where a
policy exists)
--
unsolved [/usr/doc] $ find -name *README* | grep gz | wc -l
152
unsolved [/usr/doc] $ find -name *README* | wc -l
315
Jens Ritter [EMAIL PROTECTED] writes:
Shall they be gzipped or not? (Up to the copyright file, where a
policy exists)
As Luis Francisco Gonzalez pointed out offline, this is point 5.3 in
the current policy manual. Sorry I missed that.
---
[EMAIL PROTECTED] [EMAIL PROTECTED]
Key ID:
On 23 Feb 1998, Manoj Srivastava wrote:
Marcelo stuff like that. I modified it to use /etc/GNUstep/Defaults
Marcelo (sue me, I didn't ask here first -- now I realize I should
Marcelo have)
Huh? There is an misunderstanding here. /etc/GNUstep/Defaults
is great -- it is under
On Mon, Feb 23, 1998 at 09:09:07AM -0500, Branden Robinson wrote:
So finally, here's my proposed solution:
xtrs should install two zero-length files:
/usr/lib/xtrs/model1rom.bin
/usr/lib/xtrs/model3rom.bin
and flag them as conffiles, so they don't get clobbered during an upgrade.
Manoj Srivastava [EMAIL PROTECTED] writes:
I was just reading the proposal, an I think I have gotten
myself confused.
That's OK. It *is* confusing. I'll be happy to revise the docs if
you or I can think of what to change, but...
1) /usr/lib/emacs/packages/emacsen-install/vm
Manoj Srivastava [EMAIL PROTECTED] writes:
I think I disagree. Letting developers who can't write shell
scripts can in no way be in the interests of the project, one of
whose goals is *excellence*.
Sure it can. It depends on what they're doing. You don't have to be
a good shell
Manoj Srivastava [EMAIL PROTECTED] writes:
I should apologize to the people who write helper scripts for
my outburst; I regretted it as soon as I had sent it.
Christoph Lameter [EMAIL PROTECTED] writes:
Maybe we should also try to have developers who are up to date on the
tools
I've noticed what seems to be a common problem lately: daemons that do not
chdir / on startup. The problem is, if you mount a debina cd on /mnt, cd to
/mnt, install some daemons, then /mnt is always busy after that and cannot
be unmounted. The solution is to add a chdir / to the daemon's startup
[discussion about conffiles/configuration files]
I just read the whole tread and now I'm really confused 8-) Are there
real questions you are talking about or is this just for fun?? (Sorry, I
couldn't resist ;-)
Let me just comment on one clear question I've seen in the discussion:
Should
On Fri, 20 Feb 1998, Joey Hess wrote:
[snip]
Policy says:
3.3.5. Symbolic links
-
Most symbolic links should be relative, not absolute. Absolute links,
in general, cause problems when a file system is not mounted where it
normally resides (for example,
Let me just throw in a few notes:
(1. I'm in the process of taking over the maintenance of the Packaging
Manual. I'll release this as a new package as soon as the dpkg package is
released.)
2. I'm wondering why it's so hard for people to get the difference between
configuration files and
On 24 Feb 1998, Manoj Srivastava wrote:
Hi,
Hmm, my reading of policy is that packages refers to is .deb
Debian packages, not sources or the so called ``source packages'',
which are never mentioned in policy.
Huh? See section `2.4 Source packages'.
We bundle upstream
Christian Schwarz wrote:
I agree that the policy should be clarified. In short: Symbolic links
_within_ a top level directory should be relative, symbolic links
_between_ top level directories should be absolute.
Great. That's simple and seems perfect to me.
--
see shy jo
On Tue, 24 Feb 1998, Christian Schwarz wrote:
Recall, that we discussed how to handle `variable data files which are not
edited by humans but by scripts/programs' a few days/weeks ago. The result
of the discussion was that `conffiles' may only be edited by humans, not
by scripts. (This result
Christian Schwarz wrote:
2. I'm wondering why it's so hard for people to get the difference between
configuration files and conffiles. Perhaps, the name conffiles is not a
good name (but either way, it's unlikely that we'll change the name
because this would be a _lot_ of work). Here is my
Christian Schwarz wrote:
I just read the whole tread and now I'm really confused 8-) Are there
real questions you are talking about or is this just for fun?? (Sorry, I
couldn't resist ;-)
No, it's not just for fun. Some examples of other conffiles that Manoj
thinks should be disallowed because
35 matches
Mail list logo