* Steve M Robbins [EMAIL PROTECTED] wrote:
Hi,
What I am proposing is a source package that generates *both* a
"main" and a "contrib" .deb.
This is not allowed.
One source package can only build packages for one section.
See, the structure on the FTP sites reflects this:
dist - main-
* Corrin Lakeland [EMAIL PROTECTED] wrote:
I'm trying to package a program (gnubg) that generates a binary file
during compilation. However this upsets dpkg-buildpackage which
can't diff between the current version and the .orig
This means, that you have to fix "make clean" to remove this
* Chris Danis [EMAIL PROTECTED] wrote:
If you want, maybe add a note echo'd to the user in the postinst
No, this is no good.
1. the messages scroll by to fast to notice, and please don't even
think of adding a "press RETURN to continue" thingy. Thing unintended
installation.
2. this is just
* Steve M Robbins [EMAIL PROTECTED] wrote:
Hi,
What I am proposing is a source package that generates *both* a
main and a contrib .deb.
This is not allowed.
One source package can only build packages for one section.
See, the structure on the FTP sites reflects this:
dist - main- i386
* Corrin Lakeland [EMAIL PROTECTED] wrote:
I'm trying to package a program (gnubg) that generates a binary file
during compilation. However this upsets dpkg-buildpackage which
can't diff between the current version and the .orig
This means, that you have to fix make clean to remove this
* Chris Danis [EMAIL PROTECTED] wrote:
If you want, maybe add a note echo'd to the user in the postinst
No, this is no good.
1. the messages scroll by to fast to notice, and please don't even
think of adding a press RETURN to continue thingy. Thing unintended
installation.
2. this is just a
* "Martin" == Martin Albert [EMAIL PROTECTED] wrote:
Hi,
So recipient has to use patch -p1 (or rename his archive) right?
Do you tell them or just let'em figure it out?
You can expect that every developer knows how to use patch and that he
will first inspect the file you sent him before
* Martin == Martin Albert [EMAIL PROTECTED] wrote:
Hi,
So recipient has to use patch -p1 (or rename his archive) right?
Do you tell them or just let'em figure it out?
You can expect that every developer knows how to use patch and that he
will first inspect the file you sent him before
* "Britton" == Britton [EMAIL PROTECTED] wrote:
Britton I just went to check for bugs, and I notice that the upload I
Britton made for the last couple seems to be reported as having been
Britton an NMU upload.
It would be more useful, if you gave some of the bug numbers where
this happened.
* "Drew" == Drew Parsons [EMAIL PROTECTED] wrote:
Drew What's the "best" way of maintaining or creating a man page?
I create a perl pod file and translate it to to a man page. The syntax
is very easy, so I prefer this. See perldoc perlpod and the equivs
package for an example.
Ciao,
* Drew == Drew Parsons [EMAIL PROTECTED] wrote:
Drew What's the best way of maintaining or creating a man page?
I create a perl pod file and translate it to to a man page. The syntax
is very easy, so I prefer this. See perldoc perlpod and the equivs
package for an example.
Ciao,
Martin
* "Thomas" == Thomas Lange [EMAIL PROTECTED] wrote:
[ Multiple binaries from one source ]
Thomas Do I have to make two source packages for that, or can I
Thomas create two single binary packages with different version
Thomas numbers from one source ?
One source is OK. I do this with the
* Mariusz Przygodzki [EMAIL PROTECTED] wrote:
Upstream sources have no icon for some application. How I can add
an icon to debian/ in other format than xpm ?
uuencode it.
I used to place a file in debian/, then did a (cd debian uudecode
swirl-icon.uue) in the build target and rm the decoded
* Mariusz Przygodzki [EMAIL PROTECTED] wrote:
Upstream sources have no icon for some application. How I can add
an icon to debian/ in other format than xpm ?
uuencode it.
I used to place a file in debian/, then did a (cd debian uudecode
swirl-icon.uue) in the build target and rm the
* Roberto Suarez Soto [EMAIL PROTECTED] wrote:
Hi,
Also, should I ask the user at install time if he/she wants to have
this script installed?
In contrast to Josip, I say install it unconditionally.
You get very little gain for making your package install interatively,
so it is not worth
* Roberto Suarez Soto [EMAIL PROTECTED] wrote:
Hi,
Also, should I ask the user at install time if he/she wants to have
this script installed?
In contrast to Josip, I say install it unconditionally.
You get very little gain for making your package install interatively,
so it is not worth it.
Hi,
maybe you want to use debconf to show the message? This has the
advantage of being non-interactive, if the admin chooses so (he will
get a mail with the warning in this case).
See the gmc package for an example.
Ciao,
Martin
Hi,
I have a problem with gnome-users-guide-en
in the current version 1.2-1, there is a symlink
/usr/share/doc/gnome-users-guide-en/html -
/usr/share/gnome/help/users-guide/C
dwww won't follow this symlink, so I turned it around in 1.2-2 to read
/usr/share/gnome/help/users-guide/C -
* opal == opal [EMAIL PROTECTED] wrote:
Hi,
opal I have created a mcal package and I want someone to test and/or
opal analyze it if I am doing something wrong. If it is not it might
opal be included in debian sometime. :)
You didn't run lintian on these packages.
Get rid of the -doc package
* Jimmy == Jimmy O'Regan [EMAIL PROTECTED] wrote:
Jimmy name: browser-history
Jimmy desc: Unified browser history logger for several browsers
Jimmy license: X11
Already in Debian. But IIRC Karl was looking for someone to adopt it.
Ciao,
Martin
* Jordi == Jordi [EMAIL PROTECTED] wrote:
Jordi Also, I see that vim, elvis and nvi provide an alternative for
Jordi vi, (in vim's case, priority 20). I think it's not against
Jordi the policy if I provide an alternative for pico?
This does not work. For alternatives, all affected packages
* Paul == Paul Serice [EMAIL PROTECTED] wrote:
Paul What's the drill whenever I have binary changes to the
Paul upstream tarball? I ask because there are some chess images
Paul (gif format) that make for a nice html-based cmoputer annotation
Paul of games that aren't distributed with the main
* Domenico == Domenico Andreoli [EMAIL PROTECTED] wrote:
Domenico here is my control file for curl source tree:
[...]
Is this the complete file? Is curl-ssl in a seperate source tree?
Domenico it looks fine, doesn't it?
Yes. If I feed equivs-build (package equivs, builds dummy packages)
with
* peter == peter karlsson [EMAIL PROTECTED] wrote:
peter Ah! dh_fixperms did indeed fiddle with the permissions, after
peter moving the chmod downwards it works just fine. Thanks!
I use (for mc) the commands in this sequence
chmod ...
dh_suidregister
dh_fixperms
This way, the binary is
* Julian == Julian Gilbey [EMAIL PROTECTED] wrote:
Julian No policy yet. But they are intended as metapackages which
Julian contain no code,
This is not a strict thing. Some of the meta-packages I maintain have
additional info in the README.Debian, and I thought about a setup tool
in one of
* Kurt == Kurt D Starsinic [EMAIL PROTECTED] wrote:
Kurt I would call it a _very_ wide interpretation of glue. For
Kurt example, task-devel-common doesn't depend on anything; it only
Kurt suggests and recommends.
I haven't looked at it yet (time...) but I have to. This won't work
with apt.
* Ramakrishnan == Ramakrishnan M [EMAIL PROTECTED] wrote:
Ramakrishnan pkgname-doc.docs
Ramakrishnan pkgname-doc.file
Ramakrishnan ex.doc-base.package
Ramakrishnan Can you please help me what to put in these files. Is
Ramakrishnan there ant documantation available on these files( Or
* tony == tony mancill [EMAIL PROTECTED] wrote:
tony I was able to circumnavigate the problem (perhaps I should have
tony played with it a little longer before posting), but it still
tony baffles me. I was doing my package build with:
tony dpkg-buildpackage -rfakeroot
tony and it was failing.
* Jozef == Jozef Hitzinger [EMAIL PROTECTED] wrote:
Jozef I've managed to do simple package, but now I need more .. what
Jozef should I do, apart from adding package section to control and
Jozef creating package.init etc for each bin package? Should I change
Jozef something in rules or
Peter == Peter Makholm [EMAIL PROTECTED] wrote:
Peter It would be nice if every user of debian solved as many bugs as
Peter they could. Probally adding [FIX] to the subject would
Peter increase the chance that somebody takes care of the patches if
Peter the maintainer doesn't.
Changing the
David == David Coe [EMAIL PROTECTED] wrote:
David I'm working on some of the open bugs, and would like to contact
David one (or more) of the bug submitters.
This is great.
David Should I write directly to the bug submitters and cc: the
David relevant bug [EMAIL PROTECTED], or is it better to
Leon == Leon Breedt [EMAIL PROTECTED] wrote:
Leon In the latest upstream release of a package, the authors put a
Leon debian/ directory into the source that is totally stuffed :(.
The author knows that you maintain this programm for Debian, right? Or
did you just start? Then you should
Stephan == Stephan A Suerken [EMAIL PROTECTED] writes:
Stephan I hope the original poster is aware of our in-depth analysis
Stephan of the problem? We might consider packaging this thread
Stephan itself, to make it available for the whole debian community
Stephan in a convenient way.
I am
Stephan == Stephan A Suerken [EMAIL PROTECTED] writes:
Stephan Sure he did, but what I meant is how could he know the next
Stephan version would be (imho wrongly) be versioned 1.6? Normally,
Stephan the version following 1.52 would be 1.53.
[...]
Stephan You are completely right, yet it seems
Stephan == Stephan A Suerken [EMAIL PROTECTED] writes:
Stephan Martin Bialasinski [EMAIL PROTECTED] writes:
Because 1.6 1.52, as 6 52
He did the right thing.
Stephan Sure he did, but what I meant is how could he know the next
Stephan version would be (imho wrongly) be versioned 1.6
Stephan == Stephan A Suerken [EMAIL PROTECTED] writes:
Stephan Eduardo Fernandez Corrales [EMAIL PROTECTED] writes:
I packaged a program that had 1.52 as upstream version number. Debian
version number was 1.5.2-1
Stephan Hmm, not knowing that the next upstream would be versioned 1.6, why
John == John Hasler [EMAIL PROTECTED] writes:
John I wish I could find out where this is documented. The nearest
John thing I can find is this line in /etc/init.d/rcS:
It was not formalized properly. I thing I saw something about it in
the last summary posting about the discussions in -policy
RC == Randolph Chung [EMAIL PROTECTED] writes:
RC what's the easiest way to write a manpage if i don't want to learn
RC troff? :)
You could use a .pod file (this is the perl documentation format),
which you can convert into a man page.
pod2man --section=1 --release=`date +'%d %b %Y'`\
CW == Chris Waters [EMAIL PROTECTED] writes:
CW The only tricky bit I've found with the whole cvs-buildpackage
CW system is that you have to commit test builds and hand export them
CW for testing purposes if you don't want to constantly overwrite
CW tags in the cvs repository.
Why this? You
RC == Randolph Chung [EMAIL PROTECTED] writes:
RC The GnuDIP package has a server component that is supposed to run
RC as a non-root user. As such, I've created a gnudip user/group for
RC it.
What rights should this user have? Maybe you could use the existing
daemon or nobody user.
Though I
MM == Mauro Mazzieri [EMAIL PROTECTED] writes:
MM 1) The program must be started as:
MM # udplog level
MM where level is a number. No default is read from is config file. Is
MM 2) Non all the user would like to start the command from init. So
MM there must a postinst script that ask the user
CL == Chris Leishman [EMAIL PROTECTED] writes:
CL On Sun, Jan 17, 1999 at 07:37:05PM -0800, R Garth Wood wrote:
On Mon, 18 Jan 1999, Chris Leishman wrote:
The main problem, however, is that it needs to validate the
people connecting, thus needs to check username/password (like
pop).
Hi,
I got a request to change a Depends: fakeroot to Depends: libtricks.
I just need the functionality of fakeroot, not the extended things in
libtricks.
Forcing the user to install libtricks seems too harsch, so: is it OK
to use Depends: libtricks | fakeroot ?
The problem is, that fakeroot
Hi,
Joost wrote he will put a Provides: fakeroot in the next libtricks (he
is working on some rewrite right now), so until then, I will use Depends:
libtricks | fakeroot
Thanks for your suggestions.
Ciao,
Martin
SC == Stephen Carpenter [EMAIL PROTECTED] writes:
SC I upgraded libc6 and libstdc++2.9 both to the latest...however
SC I STILL get the error!
Strange. What is the output of dpkg -l libc6-dev libstdc++2.9-dev ?
Mine is:
ii libc6-dev 2.0.7u-7.1 The GNU C library version 2
SJC == Stephen J Carpenter [EMAIL PROTECTED] writes:
SJC and I ran lintian on the xfstt that I uploaded yesterday and it
SJC gave a couple of errors that I will be fixing for unstable soon
SJC but.,.. it said compiled with broken libc
This must be the infamous frame-info thing.
SJC Is this a
Hi,
I am doing a rewrite of the equivs package, and have two questions.
The package itself is GPLed, and I want to include some code from
other GPLed packages (mainly the doc-base parser). Now I want to give
the doc-base authors proper credit.
What is the best way to do so? Include an entry in
PSG == Peter S Galbraith [EMAIL PROTECTED] writes:
PSG Thanks. `admin' would be my first guess as to where to look for
PSG such a package, but I didn't see any other monitoring tools
PSG there,
The xlogmaster package is in admin.
Ciao,
Martin
ZEH == Zephaniah E, Hull [EMAIL PROTECTED] writes:
ZEH 3: I've got a package which can be compiled with or without X
ZEH support, However the X support does not work without changing a
ZEH few settings in the X config, as I don't use X I'm not the best
ZEH one to try and figure out what to do..
ZEH == Zephaniah E, Hull [EMAIL PROTECTED] writes:
ZEH On Tue, Dec 01, 1998 at 10:51:09AM +0100, Martin Bialasinski wrote:
Could you tell us, which program you are packageing ?
ZEH Repackaging really, tleds..
Thats fine. I use it to monitor the ippp0 ISDN device
Which files / what changes
Hi,
here is a quick one:
will dpkg remove a conffile on a upgrade, if I remove it from the
package?
If not, then I have to remove it in postrm, if it is called as
postrm upgrade
postrm purge
Right?
Ciao,
Martin
--
from a 1996 Microshit ad campaign:
The less you know about
MS == Martin Schulze [EMAIL PROTECTED] writes:
MS Martin Bialasinski wrote:
will dpkg remove a conffile on a upgrade, if I remove it from the
package?
MS I guess so. Why don't you try it on your system? Make a backup
MS before.
I tried, and it didn't. I don't know if this is intentional
Hi,
I want to make a programm version 2.0, but I want to release some
betas before this. What version scheme should I use for the betas, so
that dpkg recognises the 2.0 release to be newer than the betas?
Ciao,
Martin
JS == Jamey Sharp [EMAIL PROTECTED] writes:
JS OK, after moving the manpage from /usr/man to /usr/X11R6/man lintian
JS stopped complaining. But I was using debstd to place the man page, why
JS didn't it put it in the right place?
Because debstd is not up-to-date with the standard. The
JS == Jamey Sharp [EMAIL PROTECTED] writes:
JS through, but I'd like to verify that I packaged oneliner correctly and
JS haven't left anything out.
Use lintian to check the package.
Ciao,
Martin
k == kolds [EMAIL PROTECTED] writes:
k I have installed Debian 2.0 on my Packard Bell 386 (4MB RAM, 124MB
k harddrive). Everything seems to have gone well, but when I boot the
k system it doesn't prompt me for a super-user password and I get the
k following error:
k bash: fork:
S == Shaleh [EMAIL PROTECTED] writes:
S With the new Enlightenment package I have a small snag. I have the main
S E package w/ docs and the E binary. I then have an enlightenment-theme
S package. E depends on e-theme. E-theme contains a /usr/doc symlink to
S E. So e-theme *should* depend
DJ == Dan Jacobowitz [EMAIL PROTECTED] writes:
DJ On Sun, Jul 05, 1998 at 06:05:25PM -0700, Joey Hess wrote:
if [ ! -e /etc/X11/xaw-wrappers.conf -a -e /etc/xaw-wrappers.conf ]; then
# Move conffile. I think it's safest to use cat here, becuase
# the user may have done something funky like
APH == Adam P Harris [EMAIL PROTECTED] writes:
Martin Bialasinski [EMAIL PROTECTED] writes:
MB Ok, I've made a perlscript to handle this. Seems to work as
MB expected. Maybe someone else finds it useful.
MB #!/usr/bin/perl
MB
MB use File::Copy;
APH Is File::Copy part of perl-base
JC == Joseph Carter [EMAIL PROTECTED] writes:
JC /usr/doc/debhelper/from-debstd
JC It raised another though: The result of following these instructions
showed
JC me a LOT of things to add to debian/rules. Most of them aren't used
in this
JC package. How do I know which are and which aren't
Hi,
I am squashing the bugs in the xisp package. Report 12773 asks to move a
conffile from /etc/options.xisp to /etc/ppp/options.xisp.
This sounds reasonable. To handle upgrading from an old version, is it enough
to move the file in preinst?
Ciao,
Martin
--
To UNSUBSCRIBE, email to
61 matches
Mail list logo
