reasoning exactly describes a "Recommends" relationship. It's not
the package's fault that apt-get doesn't support Recommends.
Richard Braakman
reasoning exactly describes a "Recommends" relationship. It's not
the package's fault that apt-get doesn't support Recommends.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
.
ftplib had a similar problem. I wrote a little python script for it
(debian/html2man.py) which converts those html pages to manpages.
You might be able to use it for ideas.
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
.
ftplib had a similar problem. I wrote a little python script for it
(debian/html2man.py) which converts those html pages to manpages.
You might be able to use it for ideas.
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
--
To UNSUBSCRIBE,
le is a lot better, actually. The installation
instructions are _not_ the reason for including the file. So it should
not be named "INSTALL".
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
le is a lot better, actually. The installation
instructions are _not_ the reason for including the file. So it should
not be named "INSTALL".
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
--
To UNSUBSCRIBE, email to [EMAIL
ial interpreter for the script itself. The Essential ones
are bash, perl, and probably awk. (awk is not itself Essential but
base-files depends on it.)
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
Essential interpreter for the script itself. The Essential ones
are bash, perl, and probably awk. (awk is not itself Essential but
base-files depends on it.)
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
--
To UNSUBSCRIBE, email to [EMAI
it to maintaining binary compatibility.
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
it to maintaining binary compatibility.
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
On Thu, Jun 28, 2001 at 07:08:33AM +0900, Yooseong Yang wrote:
> > Yes. poEdit is a horrible name. Command names should be all lowercase.
>
> What item in Debian Policy? you mean policy 2.3.1 or something?
It's not in Debian Policy, it's a Unix tradition going back 30 years.
Richard Braakman
On Thu, Jun 28, 2001 at 07:08:33AM +0900, Yooseong Yang wrote:
> > Yes. poEdit is a horrible name. Command names should be all lowercase.
>
> What item in Debian Policy? you mean policy 2.3.1 or something?
It's not in Debian Policy, it's a Unix tradition going back 30 ye
to avoid the
name collision if you point it out to them.
Did you contact the maintainer of the other package? It might be
easier to change the name there, because its poedit is not the
primary tool of the package.
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/resume.html
name collision if you point it out to them.
Did you contact the maintainer of the other package? It might be
easier to change the name there, because its poedit is not the
primary tool of the package.
--
Richard Braakman
Will write free software for money.
See http://www.xs4all.nl/~dark/re
re may be some other things which should be blocked
> out, but this does not seem intrinsically difficult.
It's slightly more difficult than that, because you do want to keep
the warnings about the control file and other metadata.
--
Richard Braakman
Looking for a job writing free softw
re may be some other things which should be blocked
> out, but this does not seem intrinsically difficult.
It's slightly more difficult than that, because you do want to keep
the warnings about the control file and other metadata.
--
Richard Braakman
Looking for a job writing free softw
pkg-shlibdeps
with a LD_LIBRARY_PATH setting that makes it find the libaries in
the build tree. This way you get a warning that it couldn't find
what package provides them, but you don't get a bogus dependency.
I haven't found a cleaner solution.
--
Richard Braakman
Looking for a
pkg-shlibdeps
with a LD_LIBRARY_PATH setting that makes it find the libaries in
the build tree. This way you get a warning that it couldn't find
what package provides them, but you don't get a bogus dependency.
I haven't found a cleaner solution.
--
Richard Braakman
Looking for a
x27;s no need to put it in the standard search path. So a
directory under /usr/lib might be a good place for it.
Richard Braakman
x27;s no need to put it in the standard search path. So a
directory under /usr/lib might be a good place for it.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ion about this section in November and
December, you can find it in the archives for debian-vote.
I can't really comment on the specific problem you found, maybe you
can use it to restart the discussion :)
Richard Braakman
ion about this section in November and
December, you can find it in the archives for debian-vote.
I can't really comment on the specific problem you found, maybe you
can use it to restart the discussion :)
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject o
my patch, it'll still be in the Debian diffs,
and I can nag them again.
Richard Braakman
my patch, it'll still be in the Debian diffs,
and I can nag them again.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
4-bit related, I'm wondering
> what happens next.
The best thing is probably to ask the submitter.
Richard Braakman
4-bit related, I'm wondering
> what happens next.
The best thing is probably to ask the submitter.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
all the changelog entries after 1.1 in your
changes file.
Richard Braakman
all the changelog entries after 1.1 in your
changes file.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
d
deal with that when it happens, not before.
Richard Braakman
d
deal with that when it happens, not before.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
on-free package, because of license (perhaps supplying the
source requires an NDA), or a build dependency on non-free components
that the admins will not install.
Then there are non-free non-US packages, which cannot be build on
machines in the US.
Richard Braakman
on-free package, because of license (perhaps supplying the
source requires an NDA), or a build dependency on non-free components
that the admins will not install.
Then there are non-free non-US packages, which cannot be build on
machines in the US.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMA
chitecture:" field, which
lists on what platforms the package can be built at all.
(I don't know what testing does with later builds of a package on
new architectures -- are they moved into testing immediately,
or held for ten days? Is it possible to get such builds into
the archive even if there's already a newer version in unstable?)
Richard Braakman
uot;Architecture:" field, which
lists on what platforms the package can be built at all.
(I don't know what testing does with later builds of a package on
new architectures -- are they moved into testing immediately,
or held for ten days? Is it possible to get such builds into
the archi
ency. If your package depends
on all of those directly (i.e. it invokes their commands or includes
their headers), then you should list them all. But if for example
it only needs B because it needs A, and A needs B, then you should
let A take care of its own dependencies.
Richard Braakman
ency. If your package depends
on all of those directly (i.e. it invokes their commands or includes
their headers), then you should list them all. But if for example
it only needs B because it needs A, and A needs B, then you should
let A take care of its own dependencies.
Richard Braakman
The best solution is probably to have an autoconf macro specifically
for testing for C++ functions.
Note that using C++ kind of defeats the whole purpose of using
autoconf :-) Autoconf is not even very happy about using ANSI C,
because it's not available everywhere.
Richard Braakman
ge.
The best solution is probably to have an autoconf macro specifically
for testing for C++ functions.
Note that using C++ kind of defeats the whole purpose of using
autoconf :-) Autoconf is not even very happy about using ANSI C,
because it's not available everywhere.
Richard Braak
ved it this way:
(p & q) | (r & s)
is
((p & q) | r) & ((p & q) | s)
is
(p | r) & (q | r) & (p | s) & (q | s)
It's three applications of the distribution rule for & and |.
The end result looks complicated but has the same meaning as the
original, and it can be expressed in dpkg's Depends syntax.
Richard Braakman
ved it this way:
(p & q) | (r & s)
is
((p & q) | r) & ((p & q) | s)
is
(p | r) & (q | r) & (p | s) & (q | s)
It's three applications of the distribution rule for & and |.
The end result looks complicated but has the same meaning as
kage so the upgrade from potato shouldn't
> have all this mess. all this work could be a waste of time, since some
> upgrade problems in woody are ok since it is "testing".
Then I don't think it's worth the complication.
Richard Braakman
kage so the upgrade from potato shouldn't
> have all this mess. all this work could be a waste of time, since some
> upgrade problems in woody are ok since it is "testing".
Then I don't think it's worth the complication.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ow much of Debian this would affect.
You could have satisfied your curiosity by running lintian -I on
the archive :) IIRC, the archive scanner was originally set up this way.
Richard Braakman
know how much of Debian this would affect.
You could have satisfied your curiosity by running lintian -I on
the archive :) IIRC, the archive scanner was originally set up this way.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
7;s an info message instead of
a warning.
I tend to use install -s. With some packages I use
strip --remove-section=.note --remove-section=.comment
and with some packages, where I use the upstream makefile to install,
I just leave them in. It's about 1% of the binary size, IIRC.
Richard Braakman
7;s an info message instead of
a warning.
I tend to use install -s. With some packages I use
strip --remove-section=.note --remove-section=.comment
and with some packages, where I use the upstream makefile to install,
I just leave them in. It's about 1% of the binary size, IIRC.
Richar
n the first place.
Richard Braakman
n the first place.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
lt;= 0.1.25)
Note that it is rarely a good idea to do this. If libfoo 0.1.26 is not
compatible with 0.1.25, then it should have a new major number. Otherwise,
upgrading libfoo is going to get really icky. The svgalib experiment
showed this.
Richard Braakman
t the problem with the
permissions check was fixed).
> Do I simply ignore the errors?
We can add another override for powstatd, or teach Lintian about the
specialness of ups-monitor. In fact, it could then also check that
the package Provides and Conflicts with "ups-monitor".
Richard Braakman
ery easy to fix.
They are, they're byte-order dependent.
> But how can I
> determine whether ispell is i386-specific? (In
> theory it should not be.) Should I just upload
> one with "Architecture: All" and see what fails?
"Any", not "all".
Please re-read section 4.2.3 of the Packaging Manual.
Richard Braakman
>
> I wish to know whether I can pakcage this xacursor. and if I can,
> whether it should be main or non-free. Thanks.
No, without a license statement we cannot legally distribute it.
I see that the author has provided a number of ways to contact him.
Perhaps you can ask the author for
can not be allowed to overwrite any of the A* packets but
> I get many bugreports now that claims this behaviour as bugs :-(
Just add a lot of conflicts :-)
Richard Braakman
n answer?
Yes, that is a reasonable course. But remember that mail sent to the
bug number does *not* reach the submitter. Are you sure that you
have contacted them?
([EMAIL PROTECTED] is a handy, though undocumented, way to
reach a bug's submitter.)
Richard Braakman
prefix.
>
> You should call ldconfig from the postinst.
You should call ldconfig from the postinst, *AND* install the symlink
it wants to make. Otherwise, you leave dangling symlinks when the package
is removed.
Richard Braakman
Having a package Conflict with itself is only useful if its name doubles
as a virtual package name. Since no package Provides xpm4g-dev, the
Conflict has no effect. Just remove it.
Richard Braakman
Randolph Chung wrote:
> what's the easiest way to write a manpage if i don't want to learn troff? :)
Grab a manpage from a similar utility and fill in your own stuff :-)
Learning troff is not hard, though. man(7) tells you all you need to know.
Richard Braakman
VERRIDE it...
You can't. The diff file format cannot represent this.
What Adrian probably meant is that you rm the symbolic link in the
clean target, and recreate it in the build target.
Of course, this is going to be an icky solution either way, because
it means your package can't autobuild.
Richard Braakman
you get:
Depends: lynx | wget | libwww-perl, lynx | wget | libhtml-parser-perl
Richard Braakman
be separated from the program, which the
license doesn't allow :)
It may be possible to have a Postilion with graphics in non-free,
and one without graphics in main.
Richard Braakman
Stephen J. Carpenter wrote:
> Not sure what to do since all it says is "Broken libc" it doesn't say why it
> is broken or what version is a known not broken one.
It looks like I'll have to plaster notices about lintian's -i option in
yet more places. Where did you
d to generate the manpage package from another
source instead.
Richard Braakman
roblem, and how it was solved?
ps used to have this problem... and it was solved by fixing the package.
I don't think there's any other way. At the least, the program should
detect when a config file is an old version, and offer to delete it.
Richard Braakman
lots and lots of swap. Like 128 MB.
I found that building the static-Motif version took a lot more memory
than any other step, so if you're building dynamic only (lesstif),
you might need a less room than I did. Or did Mozilla leave Motif
behind entirely?
Richard Braakman
et
a lot of replies, and see how they differ from yours.
Richard Braakman
to discuss it on debian-policy.)
In this case, it was probably a mistake. I don't consciously remember
changing it. utils fits as well as anything, though perhaps admin
would be better. I changed it back to utils yesterday.
Richard Braakman
ve a .reason file with the
reason for the rejection.
Richard Braakman
, not a validator!
Richard Braakman
oth a libc5 and libc6 version.
This is probably a bug in libcompface. It should refrain from building
the libc5 packages on architectures that don't want them.
Richard Braakman
fixed in lintian 0.9.4.
Richard Braakman
Chris Waters wrote:
> if [ "$(echo * .*)" = "* . .." ] ; then echo empty dir; fi
Brilliant. :-)
It has only one flaw. *cough*
% mkdir emptydir; cd emptydir
% touch *
% if [ "$(echo * .*)" = "* . .." ] ; then echo empty dir; fi
empty dir
*evil grin*
Richard
you want it in frozen, please upload a new version with
distribution "frozen unstable".
Richard Braakman
|| true
The rmdir will fail if it's not empty :)
(You can use its return code if you need to check if it succeeded)
Richard Braakman
greports that were never even replied to.
I say go for it :)
Richard Braakman
nately
it's not yet implemented. See bug#17575 for more information.
Richard Braakman
"man 7 man" is a good start. And pick a manpage for a program similar
to yours, to see how it's structured :)
(There's no one true style for manpages, but it's better to use an
existing one than to invent your own.)
Richard Braakman
It's updated daily, from the keyring in
/debian2/debian/doc/debian-keyring.tar.gz
This is done just before the daily install run, so if your key is
in that tarfile it should be ok.
Richard Braakman
".
That makes it very clear what the relationship between the distributions is.
Richard Braakman
of the source package in the debian changelog
In your case, dpkg-gencontrol is complaining about a mismatch between
the last two. I think dpkg-source gives a warning if the directory
name does not match.
Richard Braakman
That's quite a lot of detail :)
Perhaps this should all be written down somewhere and become the
Python Subpolicy? We already have such subpolicies for emacs and
menu files and probably some that I forgot. It could be distributed
in the python package and referenced from debian-policy.
Ri
he form used by dpkg:
libdbd-mysql-perl | libdbd-msql-perl, libdbd-mysql-perl | msql, mysql |
libdbd-msql-perl, mysql | msql
It's not pretty, I'll admit :)
Richard Braakman
. It should be less than two days in any case.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
file;
that has to be done by hand.
> There is no way to change the override files by myself, isn't it?
Indeed not, you'll have to wait. Fortunately the main archive
now has three maintainers, so it shouldn't take long.
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED
alled. Right, folks?
Only if you check that the argument is "configure". The postinst
may be called in other situations with other arguments, and dependencies
are guaranteed for those. (See section 6.2 of the Packaging manual)
Richard Braakman
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
org also have some extra fields including the description
> architecture, md5 sums and such.
You are writing the changes file by hand? You should use dpkg-buildpackage
to compile Debian packages, and it creates the changes file for you.
Richard Braakman
--
To UNSUBSCRIBE, email to [EM
s notwithstanding.
This also makes it non-free in my book, but the DFSG is not clear about it.
As far as I'm concerned it makes the entire license meaningless; there's
nothing you can do with the code that they cannot revoke whenever they
feel like it.
> It looks free, right?
No
g "tar tv" output at that point. (It doesn't use
ls). Do you have a strange tar?
Please submit a bug report for lintian. Do you get this behaviour with
other packages as well, or just tcpquota?
Richard Braakman
--
E-mail the word "unsubscribe" to [EMAIL PROTECTED]
TO UNSUBSCRIBE FROM THIS MAILING LIST. Trouble? E-mail to [EMAIL PROTECTED]
at's another reason
to be fast with even small bugfixes.
Richard Braakman
, otherwise you will have to remember
to update it manually each time there is a new upstream release.
Richard Braakman
f PGP'ing)
> and my email address in the control file?
This currently has no effect, but the automatic bug closing feature
that's in the works will probably view such uploads as non-maintainer
uploads.
It would be a good idea to either change your maintainer address to
match one of the userids on your PGP key, or add the address as a
new userid.
Richard Braakman
Ulf Fredriksson wrote:
> [EMAIL PROTECTED] (Richard Braakman) writes:
>
> > Ulf Fredriksson wrote:
> > > tcpquota (1.6.10-1) unstable contrib; urgency=low
> >
> > You need to upload it to unstable, with section contrib/admin.
>
> Isn't that what &
to rectify the #17595 bug... That
> version have not yet been installed, so... ? Now what?
>
You need to upload it to unstable, with section contrib/admin.
Then it will sit in Incoming until Guy edits the override files,
because packages can't move from one distribution to another without
manual intervention.
Richard Braakman
93 matches
Mail list logo
