Le mar 14/10/2003 à 19:44, Daniel Kobras a écrit :
> Still, this whole prelink issue is tangent to the main
> point: There are valid reasons for static linking, and I oppose the
> blanket statement that we should deprecate this method.
And there are also good reasons to change our policy regarding
On Tue, Oct 14, 2003 at 07:44:05PM +0200, Daniel Kobras wrote:
> On Tue, Oct 14, 2003 at 12:48:28PM -0400, Daniel Jacobowitz wrote:
> > Often worse, due to the dramatically increased amount of data which
> > must be loaded from disk in a cold-cache situation. Another 800K of
> > glibc you've got t
Daniel Kobras <[EMAIL PROTECTED]> wrote:
> On Tue, Oct 14, 2003 at 08:09:32AM -0500, Steve Greenland wrote:
>> Which doesn't, in any way, promote the idea that we should keep the .la
>> files. People who need/want a statically linked binary often want to
>> control exactly *which* libraries are sta
On Tue, Oct 14, 2003 at 12:48:28PM -0400, Daniel Jacobowitz wrote:
> Often worse, due to the dramatically increased amount of data which
> must be loaded from disk in a cold-cache situation. Another 800K of
> glibc you've got to read in. The memory usage sucks too.
That's glibc. It's already in
* Daniel Kobras
| If you ever tried to get, say, ten static libs in the right order
| for a medium-sized application, you know what tedious task I'm
| taking about. From my personal point of view, removal of .la files
| would significantly degrade Debian's usability as a build platform.
You can
On Tue, 2003-10-14 at 04:38, Daniel Kobras wrote:
> To name but a few. Just because there's little incentive to use static
> linkage when building Debian packages doesn't mean that we should
> deprecate it. Unless you're willing to convince the admin of the
> beowulf cluster next door to install li
On Tue, Oct 14, 2003 at 06:35:38PM +0200, Daniel Kobras wrote:
> On Tue, Oct 14, 2003 at 11:57:40AM -0400, Daniel Jacobowitz wrote:
> > On Tue, Oct 14, 2003 at 10:38:49AM +0200, Daniel Kobras wrote:
> > > On Tue, Oct 14, 2003 at 09:52:28AM +0200, Josselin Mouette wrote:
> > > > I really feel we sho
On Tue, Oct 14, 2003 at 11:57:40AM -0400, Daniel Jacobowitz wrote:
> On Tue, Oct 14, 2003 at 10:38:49AM +0200, Daniel Kobras wrote:
> > On Tue, Oct 14, 2003 at 09:52:28AM +0200, Josselin Mouette wrote:
> > > I really feel we should get rid of all these static libraries. Who uses
> > > static linkin
On Tue, Oct 14, 2003 at 08:09:32AM -0500, Steve Greenland wrote:
> Which doesn't, in any way, promote the idea that we should keep the .la
> files. People who need/want a statically linked binary often want to
> control exactly *which* libraries are statically linked, and will build
> the link comm
I hate to jump in but I really feel the need to correct the below. You
have a good number of points wrong.
On Tue, Oct 14, 2003 at 10:38:49AM +0200, Daniel Kobras wrote:
> On Tue, Oct 14, 2003 at 09:52:28AM +0200, Josselin Mouette wrote:
> > I really feel we should get rid of all these static lib
On 14-Oct-03, 03:38 (CDT), Daniel Kobras <[EMAIL PROTECTED]> wrote:
> On Tue, Oct 14, 2003 at 09:52:28AM +0200, Josselin Mouette wrote:
> > I really feel we should get rid of all these static libraries. Who uses
> > static linking now that even our glibc doesn't support it correctly
> > across ver
On Tue, Oct 14, 2003 at 09:52:28AM +0200, Josselin Mouette wrote:
> I really feel we should get rid of all these static libraries. Who uses
> static linking now that even our glibc doesn't support it correctly
> across versions?
People who want their binaries to run across different Linux machines
Le lun 13/10/2003 à 22:19, Steve Langasek a écrit :
> > You're right as for static linking, I thought pkg-config supported
> > --static while it doesn't. Well, maybe it is better that way; I
> > personally feel we should deprecate the whole static linking stuff.
>
> In which case, we could easily
On Mon, Oct 13, 2003 at 09:44:47PM +0200, Josselin Mouette wrote:
> Le lun 13/10/2003 à 21:36, Steve Langasek a écrit :
> > > All this additional information can be provided as well by pkg-config,
> > > which is much more flexible and doesn't cause random breakages.
> > $ pkg-config libgnomecanvas
Le lun 13/10/2003 à 21:36, Steve Langasek a écrit :
> > All this additional information can be provided as well by pkg-config,
> > which is much more flexible and doesn't cause random breakages.
>
> $ pkg-config libgnomecanvas-2.0 --libs
> -Wl,--export-dynamic -lgnomecanvas-2 -lart_lgpl_2 -lpangof
On Mon, Oct 13, 2003 at 09:02:06PM +0200, Josselin Mouette wrote:
> Le dim 12/10/2003 à 03:31, Steve Langasek a écrit :
> > If I was comfortable that .la files were completely dispensable on
> > GNU/Linux systems, I wouldn't hesitate to do that, but I'm not. They
> > *do* provide additional inform
Le dim 12/10/2003 à 03:31, Steve Langasek a écrit :
> If I was comfortable that .la files were completely dispensable on
> GNU/Linux systems, I wouldn't hesitate to do that, but I'm not. They
> *do* provide additional information that's useful to people linking
> applications statically; and while
On Sun, 2003-10-12 at 02:31, Steve Langasek wrote:
> On the shared library side, we already have the NEEDED field in ELF libs
> which is more elegant; so the .la's are redundant (and, indeed, can get
> in the way). I understand Scott is working on fixing libtool so that it
> doesn't try to redund
On Sat, Oct 11, 2003 at 03:54:08PM -0500, Branden Robinson wrote:
> On Sat, Oct 11, 2003 at 10:57:38AM -0500, Steve Langasek wrote:
> > On Sat, Oct 11, 2003 at 03:21:50AM -0400, Joey Hess wrote:
> > > I'm coming into this thread very late, with what may be a stupid
> > > question, but can anyone te
On Sat, Oct 11, 2003 at 10:57:38AM -0500, Steve Langasek wrote:
> On Sat, Oct 11, 2003 at 03:21:50AM -0400, Joey Hess wrote:
> > I'm coming into this thread very late, with what may be a stupid
> > question, but can anyone tell me if the breakage could be avoided by
> > just deleting the .la files
On Sat, Oct 11, 2003 at 06:06:11AM -0600, Marcelo E. Magallon wrote:
> > Me too. The actual problem seems to have been causing by a behavior of
> > ldconfig that we almost never see because we tend to ship library
> > packages with lib$SONAME.so.$SOVERSION symbolic links already in the
> > pac
Steve Langasek wrote:
> On Sat, Oct 11, 2003 at 03:21:50AM -0400, Joey Hess wrote:
> > I'm coming into this thread very late, with what may be a stupid
> > question, but can anyone tell me if the breakage could be avoided by
> > just deleting the .la files in question?
>
> Yes, it can. I've advoc
On Sat, Oct 11, 2003 at 03:21:50AM -0400, Joey Hess wrote:
> I'm coming into this thread very late, with what may be a stupid
> question, but can anyone tell me if the breakage could be avoided by
> just deleting the .la files in question?
Yes, it can. I've advocated this on a number of occasions
I'm coming into this thread very late, with what may be a stupid
question, but can anyone tell me if the breakage could be avoided by
just deleting the .la files in question? Would libtool then find the .so
files and do something sane, or would it break further?
--
see shy jo
signature.asc
Desc
This one time, at band camp, Branden Robinson said:
> So, let's see, I've been violated by ldconfig, libtool, and dpkg-divert
> all within the past 2 days. I feel like Tera Patrick.
That was a mental picture I _really_ didn't need. Thanks :)
--
> Me too. The actual problem seems to have been causing by a behavior of
> ldconfig that we almost never see because we tend to ship library
> packages with lib$SONAME.so.$SOVERSION symbolic links already in the
> package payload.
I don't want to nitpick, but the SONAME is what you are calli
On Wed, Oct 08, 2003 at 12:04:32PM -0500, Zed Pobre wrote:
> On Tue, Oct 07, 2003 at 03:40:18AM -0500, Chris Cheney wrote:
> > The new version of libXrender moves from /usr/X11R6/lib to /usr/lib
> > which has already started to cause build failures... :\ I am not
> > certain if Branden plans to mo
On Wed, Oct 08, 2003 at 12:48:48AM -0500, Chris Cheney wrote:
> It appears all of kde has it included as well. Does this happen to have
> anything to do with the rpath'ing issue that some of the XFree libs are
> causing as well? (iirc it was xrender)
Xrender is not an XFree86 library anymore.
htt
On Wed, Oct 08, 2003 at 06:12:17AM +0100, Scott James Remnant wrote:
> Actually the problem is somewhat lessened by the fact libtool generally
> doesn't put the .la path in dependency_libs and puts -lXrender instead.
>
> The *only* package I can see so far which has
> /usr/X11R6/lib/libXrender.la
On Tue, 2003-10-07 at 21:32, Chris Cheney wrote:
> On Tue, Oct 07, 2003 at 02:58:25PM -0500, Branden Robinson wrote:
> > I think the problem with .la files may be solvable by updating
> > Build-Depends and -dev packages' dependencies to refer to libxrender-dev
> > (>= 0.8.3-1), and/or libraries th
On Tue, Oct 07, 2003 at 02:58:25PM -0500, Branden Robinson wrote:
> I think the problem with .la files may be solvable by updating
> Build-Depends and -dev packages' dependencies to refer to libxrender-dev
> (>= 0.8.3-1), and/or libraries that are rebuilt against that version of
> libxrender-dev.
On Tue, Oct 07, 2003 at 03:40:18AM -0500, Chris Cheney wrote:
> On Tue, Oct 07, 2003 at 10:26:45AM +0200, Marcelo E. Magallon wrote:
> > On Tue, Oct 07, 2003 at 01:33:01AM -0500, Chris Cheney wrote:
> >
> > > Does anyone happen to know why .la files hardcode the paths to .la
> > > files that the
On Tue, Oct 07, 2003 at 09:30:56AM +0100, Scott James Remnant wrote:
> On Tue, 2003-10-07 at 07:33, Chris Cheney wrote:
>
> > Does anyone happen to know why .la files hardcode the paths to .la files
> > that they depend on?
> >
> To guarantee that you don't end up linking with something totally
>
On Tue, Oct 07, 2003 at 11:11:35AM +0200, Marcelo E. Magallon wrote:
> > What's more of a problem here is that libtool actually links
> > dependency libraries of dependencies ... it's something I've been
> > working on for a while.
> That's a bug, not a feature. From libtool's perspective at
On Tue, Oct 07, 2003 at 09:30:56AM +0100, Scott James Remnant wrote:
> On Tue, 2003-10-07 at 07:33, Chris Cheney wrote:
>
> > Does anyone happen to know why .la files hardcode the paths to .la files
> > that they depend on?
> >
> To guarantee that you don't end up linking with something totally
>
> What's more of a problem here is that libtool actually links
> dependency libraries of dependencies ... it's something I've been
> working on for a while.
That's a bug, not a feature. From libtool's perspective at least.
You have to keep in mind that libtool is designed to work around
ex
On Tue, Oct 07, 2003 at 10:26:45AM +0200, Marcelo E. Magallon wrote:
> On Tue, Oct 07, 2003 at 01:33:01AM -0500, Chris Cheney wrote:
>
> > Does anyone happen to know why .la files hardcode the paths to .la
> > files that they depend on?
>
> Anal-retentiveness wrt using the exact same library o
On Tue, 2003-10-07 at 07:33, Chris Cheney wrote:
> Does anyone happen to know why .la files hardcode the paths to .la files
> that they depend on?
>
To guarantee that you don't end up linking with something totally
different if the app being compiled happens to have a different search
path.
What
On Tue, Oct 07, 2003 at 01:33:01AM -0500, Chris Cheney wrote:
> Does anyone happen to know why .la files hardcode the paths to .la
> files that they depend on?
Anal-retentiveness wrt using the exact same library originaly used.
> This is about to bite Debian hard with some of the XFree86 lib
Does anyone happen to know why .la files hardcode the paths to .la files
that they depend on?
For example:
dependency_libs=' -lm -L/usr/lib /usr/lib/libogg.la'
This is about to bite Debian hard with some of the XFree86 libraries
moving to /usr/lib.
Chris Cheney
---
# libvorbis.la - a libtool
40 matches
Mail list logo
