On Tue, Aug 02, 2005 at 08:58:14PM +0300, Alin Nastac <[EMAIL PROTECTED]> wrote:
> It also controls whether a cron script that use native squid log
> rotation is installed or not. You cannot select your preferred rotation
> mechanism (logrotate or cron job) through other way than useflags.
If it's
On Tue, 02 Aug 2005 10:52:44 -0700 Donnie Berkholz
<[EMAIL PROTECTED]> wrote:
| Ciaran McCreesh wrote:
| | Files in /etc are expensive in terms of sysadmin time. The only
| | things in /etc should be things that are both necessary and likely
| | to be modified by a sysadmin.
|
| So we should stop
On Tue, 02 Aug 2005 13:48:17 -0400 Alec Joseph Warner
<[EMAIL PROTECTED]> wrote:
| > Files in /etc are expensive in terms of sysadmin time. The only
| > things in /etc should be things that are both necessary and likely
| > to be modified by a sysadmin.
| >
| And who makes that call, shouldn't the
Donnie Berkholz wrote:
> Tom Martin wrote:
> | Hi list,
> |
> | Bug 97447 wants a logrotate USE flag, which is used by about five
> | packages locally. Unless there are any objections, I'll globalify it
> | later today.
>
> I think this flag is a bad idea. Why should I have to recompile a
> packag
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Ciaran McCreesh wrote:
| Files in /etc are expensive in terms of sysadmin time. The only things
| in /etc should be things that are both necessary and likely to be
| modified by a sysadmin.
So we should stop installing example files, too?
-BEGIN
Ciaran McCreesh wrote:
On Tue, 02 Aug 2005 09:48:25 -0700 Donnie Berkholz
<[EMAIL PROTECTED]> wrote:
| Tom Martin wrote:
| | Hi list,
| |
| | Bug 97447 wants a logrotate USE flag, which is used by about five
| | packages locally. Unless there are any objections, I'll globalify it
| | later today.
On Tue, 02 Aug 2005 09:48:25 -0700 Donnie Berkholz
<[EMAIL PROTECTED]> wrote:
| Tom Martin wrote:
| | Hi list,
| |
| | Bug 97447 wants a logrotate USE flag, which is used by about five
| | packages locally. Unless there are any objections, I'll globalify it
| | later today.
|
| I think this flag i
On Tue, 2005-08-02 at 18:04 +0100, Tom Martin wrote:
> On Tue, Aug 02, 2005 at 08:17:08AM -0400, Chris Gianelloni <[EMAIL
> PROTECTED]> wrote:
> > If you make the flag global, can you make sure that all of the packages
> > that use this flag exhibit the same behavior when using it. If they all
>
On Tue, Aug 02, 2005 at 08:17:08AM -0400, Chris Gianelloni <[EMAIL PROTECTED]>
wrote:
> If you make the flag global, can you make sure that all of the packages
> that use this flag exhibit the same behavior when using it. If they all
> do their own thing, then I would suggest leaving it local.
>
Donnie Berkholz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Tom Martin wrote:
| Hi list,
|
| Bug 97447 wants a logrotate USE flag, which is used by about five
| packages locally. Unless there are any objections, I'll globalify it
| later today.
I think this flag is a bad idea. Why sho
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Tom Martin wrote:
| Hi list,
|
| Bug 97447 wants a logrotate USE flag, which is used by about five
| packages locally. Unless there are any objections, I'll globalify it
| later today.
I think this flag is a bad idea. Why should I have to recompile a
On Tue, Aug 02, 2005 at 10:22:01AM +0100, Tom Martin wrote:
> Hi list,
>
> Bug 97447 wants a logrotate USE flag, which is used by about five
> packages locally. Unless there are any objections, I'll globalify it
> later today.
>
You may want to read over this thread from April when the logrotate
On Tue, 2005-08-02 at 10:22 +0100, Tom Martin wrote:
> Hi list,
>
> Bug 97447 wants a logrotate USE flag, which is used by about five
> packages locally. Unless there are any objections, I'll globalify it
> later today.
If you make the flag global, can you make sure that all of the packages
that
On Tuesday, August 02, 2005 2:22 am, Tom Martin wrote:
> Bug 97447 wants a logrotate USE flag, which is used by about five
> packages locally. Unless there are any objections, I'll globalify it
> later today.
I'd welcome this, almost as much as a global do_not_install_cron_tasks flag.
--
Anthon
Hi list,
Bug 97447 wants a logrotate USE flag, which is used by about five
packages locally. Unless there are any objections, I'll globalify it
later today.
Tom
--
Tom Martin, http://dev.gentoo.org/~slarti
AMD64, net-mail, shell-tools, vim, recruiters
Gentoo Linux
pgpBtgHxX9V5k.pgp
Descriptio
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Robin H. Johnson wrote:
> On Fri, Apr 29, 2005 at 09:36:54AM +0900, Jason Stubbs wrote:
>
>>What about the unused `ebuild [ebuild] config`? Isn't that the perfect place
>>for this sort of stuff? The only package that I know that uses this feature
>>
On Fri, 2005-04-29 at 10:44 -0700, Robin H. Johnson wrote:
> On Fri, Apr 29, 2005 at 09:36:54AM +0900, Jason Stubbs wrote:
> > What about the unused `ebuild [ebuild] config`? Isn't that the perfect
> > place
> > for this sort of stuff? The only package that I know that uses this feature
> > is m
On Saturday 30 April 2005 02:44, Robin H. Johnson wrote:
> On Fri, Apr 29, 2005 at 09:36:54AM +0900, Jason Stubbs wrote:
> > What about the unused `ebuild [ebuild] config`? Isn't that the perfect
> > place for this sort of stuff? The only package that I know that uses this
> > feature is mysql. The
On Saturday 30 April 2005 02:46, Chris Gianelloni wrote:
> On Sat, 2005-04-30 at 02:25 +0900, Jason Stubbs wrote:
> > # emerge config db --ask
> >
> > Please select a package to configure:
> > 1) sys-libs/db-4.2.52_p2
> > 2) sys-libs/db-1.85-r2
> > X) Cancel
> > Selection? [1/2/X] 1
> >
> > Ready t
On Sat, 2005-04-30 at 02:25 +0900, Jason Stubbs wrote:
> On Saturday 30 April 2005 00:52, Chris Gianelloni wrote:
> > On Fri, 2005-04-29 at 22:55 +0900, Jason Stubbs wrote:
> > > The only thing I think is necessary to improve for this to work is the
> > > method in which a package's config method i
On Fri, Apr 29, 2005 at 09:36:54AM +0900, Jason Stubbs wrote:
> What about the unused `ebuild [ebuild] config`? Isn't that the perfect place
> for this sort of stuff? The only package that I know that uses this feature
> is mysql. There are way more possibilities.
No, ebuild config shouldn't be a
On Sat, 2005-04-30 at 01:17 +0900, Jason Stubbs wrote:
> On Saturday 30 April 2005 00:52, Chris Gianelloni wrote:
> > On Fri, 2005-04-29 at 22:55 +0900, Jason Stubbs wrote:
> > > The only thing I think is necessary to improve for this to work is the
> > > method in which a package's config method i
On Saturday 30 April 2005 00:52, Chris Gianelloni wrote:
> On Fri, 2005-04-29 at 22:55 +0900, Jason Stubbs wrote:
> > The only thing I think is necessary to improve for this to work is the
> > method in which a package's config method is invoked. Doing it via emerge
> > would be much more intuitive
On Saturday 30 April 2005 00:52, Chris Gianelloni wrote:
> On Fri, 2005-04-29 at 22:55 +0900, Jason Stubbs wrote:
> > The only thing I think is necessary to improve for this to work is the
> > method in which a package's config method is invoked. Doing it via emerge
> > would be much more intuitive
On Fri, 2005-04-29 at 22:55 +0900, Jason Stubbs wrote:
> The only thing I think is necessary to improve for this to work is the method
> in which a package's config method is invoked. Doing it via emerge would be
> much more intuitive. Something like "emerge --config dev-db/mysql"? Use
> standar
On Fri, 2005-04-29 at 09:25 -0400, Dan Meltzer wrote:
> The problem with ebuild config, at least until bug 11359 is handled,
> is if the package is emerged early on in a list of packages, there is
> a chance the person won't know to use ebuild config would it be
> possible for portage to run an
On Friday 29 April 2005 22:25, Dan Meltzer wrote:
> The problem with ebuild config, at least until bug 11359 is handled,
> is if the package is emerged early on in a list of packages, there is
> a chance the person won't know to use ebuild config would it be
> possible for portage to run an ebu
The problem with ebuild config, at least until bug 11359 is handled,
is if the package is emerged early on in a list of packages, there is
a chance the person won't know to use ebuild config would it be
possible for portage to run an ebuild config for _all_ packages that
need it after _all_ pac
On Fri, 2005-04-29 at 09:36 +0900, Jason Stubbs wrote:
> > In fact, I've thought many times about supplying "pre-packages" that are
> > no more than a collection of all the config files for a given package.
>
> What about the unused `ebuild [ebuild] config`? Isn't that the perfect place
> for thi
On Thu, Apr 28, 2005 at 05:18:07PM -0500, Grant Goodyear wrote:
> As an aside, my preference would be to install them by default, but
> allow a nologrotate equivalent to "nodoc", "noinfo", or "noman". That
> would require hacking portage, though, which seems silly for something
> so trivial. Does
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Jason Stubbs wrote:
> What about the unused `ebuild [ebuild] config`? Isn't that the perfect place
> for this sort of stuff? The only package that I know that uses this feature
> is mysql. There are way more possibilities.
Interesting idea.
> Iggy'
On Friday 29 April 2005 07:28, Donnie Berkholz wrote:
> My problem with it is that Joe User, who doesn't do ebuilds, will change
> his USE flags and recompile an entire package just to get a little
> logrotate file that has nothing to do with the source of the package
> itself. This waste of time i
On Thu, 2005-04-28 at 10:30 +0200, Henrik Brix Andersen wrote:
> Hi,
>
> I'd like to propose a new global USE flag named 'logrotate' to add
> support for app-admin/logrotate (by installing logrotate config files
> to /etc/logrotate.d/).
>
> There are currently one local USE flag named logrotate i
On Thu, 2005-04-28 at 10:30 +0200, Henrik Brix Andersen wrote:
> Hi,
>
> I'd like to propose a new global USE flag named 'logrotate' to add
> support for app-admin/logrotate (by installing logrotate config files
> to /etc/logrotate.d/).
>
> There are currently one local USE flag named logrotate i
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Grant Goodyear wrote:
> Donnie Berkholz wrote: [Thu Apr 28 2005, 03:21:50PM CDT]
>
>>Yes. Or don't install it at all, and let people create their own. This
>>is a dumb USE flag.
>
>
> I'd say that "dumb USE flag" is rather harsh. It seems clear tha
Donnie Berkholz wrote: [Thu Apr 28 2005, 03:21:50PM CDT]
> Yes. Or don't install it at all, and let people create their own. This
> is a dumb USE flag.
I'd say that "dumb USE flag" is rather harsh. It seems clear that (a)
many users would like a logrotate script installed for them, and since
it's
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Ciaran McCreesh wrote:
> On Thu, 28 Apr 2005 12:48:52 -0700 Donnie Berkholz
> <[EMAIL PROTECTED]> wrote:
> | How annoying is it for someone to remerge a package just to get a
> | logrotate file? I suggest that if you're going to install it at all,
> |
On Thu, 28 Apr 2005 12:48:52 -0700 Donnie Berkholz
<[EMAIL PROTECTED]> wrote:
| How annoying is it for someone to remerge a package just to get a
| logrotate file? I suggest that if you're going to install it at all,
| do it unconditionally.
What, and get another file in /etc that needs manual tid
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Henrik Brix Andersen wrote:
> Hi,
>
> I'd like to propose a new global USE flag named 'logrotate' to add
> support for app-admin/logrotate (by installing logrotate config files
> to /etc/logrotate.d/).
>
> There are currently one local USE flag named
On Thu, 28 Apr 2005 11:23:09 -0400
Chris Gianelloni <[EMAIL PROTECTED]> wrote:
> Maybe for embedded or livecd? Maybe for people that don't use
> logrotate?
>
I would add a more general reason which is that this kind of
"something.d" config files are much more invasive than others, and
thus shou
On Thu, 2005-04-28 at 13:41 +, Cory Visi wrote:
> Why do we even need a USE flag for this? The logrotate config files are
> max, 4k. Just include the config file with your ebuild, which is what I
> have done up until now. A user can either run logrotate or not. This
> isn't about turning on
On Thu, 28 Apr 2005 10:00:23 -0400
Mike Frysinger <[EMAIL PROTECTED]> wrote:
> would it be possible to easily script a 'make_logrotated' func
> where you pass it a few params and it'd create a file for you
> based off a common template ?
There is one in the eclass i've previously sent (attachment
On Thursday 28 April 2005 07:01 am, Henrik Brix Andersen wrote:
> On Thu, 2005-04-28 at 03:44 -0700, Robin H. Johnson wrote:
> > (rough prototype here)
> > IUSE="$IUSE logrotate"
> > RDEPEND="$RDEPEND logrotate? ( app-admin/logrotate )"
> >
> > dologrotated() {
> > use logrotate && insinto /etc
On Thu, 28 Apr 2005 13:41:22 + Cory Visi <[EMAIL PROTECTED]> wrote:
| Why do we even need a USE flag for this? The logrotate config files
| are max, 4k. Just include the config file with your ebuild, which is
| what I have done up until now. A user can either run logrotate or
| not. This isn
Cory Visi wrote:
> On Thu, Apr 28, 2005 at 10:30:03AM +0200, Henrik Brix Andersen wrote:
>
>>Hi,
>>
>>I'd like to propose a new global USE flag named 'logrotate' to add
>>support for app-admin/logrotate (by installing logrotate config files
>>to /etc/logrotate.d/).
>>
>>There are currently one loc
On Thu, Apr 28, 2005 at 10:30:03AM +0200, Henrik Brix Andersen wrote:
> Hi,
>
> I'd like to propose a new global USE flag named 'logrotate' to add
> support for app-admin/logrotate (by installing logrotate config files
> to /etc/logrotate.d/).
>
> There are currently one local USE flag named logr
maillog: 28/04/2005-13:05:10(+0100): Ciaran McCreesh types
> On Thu, 28 Apr 2005 12:34:50 +0200 Henrik Brix Andersen
> <[EMAIL PROTECTED]> wrote:
> | Do we really need an eclass for that?
>
> Take a look at bash-completion.eclass. I'm strongly in favour of
> offloading that kind of thing into ecla
On Thu, 28 Apr 2005 12:34:50 +0200 Henrik Brix Andersen
<[EMAIL PROTECTED]> wrote:
| Do we really need an eclass for that?
Take a look at bash-completion.eclass. I'm strongly in favour of
offloading that kind of thing into eclasses as far as possible, if
nothing else because it'll let us move logr
On Thu, 28 Apr 2005 03:44:44 -0700
"Robin H. Johnson" <[EMAIL PROTECTED]> wrote:
> I meant an eclass to take care of all of it together.
I had started writing one too some time ago, but i've actually
never applied it to any ebuild and then forgot about it. Anyway,
here it is, with some freshly ad
On Thu, 2005-04-28 at 03:44 -0700, Robin H. Johnson wrote:
> >I meant an eclass to take care of all of it together.
Ah, ok - sounds good to me. Perhaps we can stick into eutils.eclass?
> (rough prototype here)
> IUSE="$IUSE logrotate"
> RDEPEND="$RDEPEND logrotate? ( app-admin/logrotate )"
>
> d
On Thu, Apr 28, 2005 at 12:34:50PM +0200, Henrik Brix Andersen wrote:
> On Thu, 2005-04-28 at 01:51 -0700, Robin H. Johnson wrote:
> > +1 here. Probably best to do it via an eclass since lots of packages
> > will be using it.
> It's as simple as:
>
> if use logrotate; then
> in
On Thu, 2005-04-28 at 01:51 -0700, Robin H. Johnson wrote:
> +1 here. Probably best to do it via an eclass since lots of packages
> will be using it.
It's as simple as:
if use logrotate; then
insinto /etc/logrotate.d
newins ${FILESDIR}/${P}-logrotate ${PN}
On Thu, Apr 28, 2005 at 10:30:03AM +0200, Henrik Brix Andersen wrote:
> I'd like to propose a new global USE flag named 'logrotate' to add
> support for app-admin/logrotate (by installing logrotate config files
> to /etc/logrotate.d/).
+1 here. Probably best to do it via an eclass since lots of pac
Hi,
I'd like to propose a new global USE flag named 'logrotate' to add
support for app-admin/logrotate (by installing logrotate config files
to /etc/logrotate.d/).
There are currently one local USE flag named logrotate in
net-proxy/squid, which of course isn't enough to justify having a global
US
54 matches
Mail list logo
