Laszlo Peter wrote:
Hi Rod,
Thanks for the detailed response.
Of course I agree that it's better to avoid the problem by
carefully controlling public interface changes. However more
and more software in Solaris comes from external sources ...
So we can't break the interfaces we'd shipped but
Hi Rod,
Thanks for the detailed response.
Of course I agree that it's better to avoid the problem by
carefully controlling public interface changes. However more
and more software in Solaris comes from external sources
and while we can work with community maintainers and try
and make sure they fo
Laca wrote:
As far as I understand, the main problem with multiple versions
(apart from the increased maintenance effort) is that different
versions of the same lib may end up linked to the same executable
(through dependent libs) and it'll cause unpredictable runtime
linking and most likely a c
Darren Kenny <[EMAIL PROTECTED]> wrote:
> A Debian distro based on Open Solaris is a perfectly valid idea, and for
> some people would be perfect for them - this is exactly why people like
> RedHat and Novell/SuSE have the concept of an enterprise distro and a
> Fedora Core or Novell Linux Desk
On Fri, Jul 22, 2005 at 12:48:11AM -0700, Bob Palowoda wrote:
> > What are the core differences WRT the "way Sun has
> > traditionally done
> > sfw and the companion CD"? A list or comparison
> > would do quite a bit
> > to move this discussion from "theory/grandstanding"
> > to "a potential pla
UNIX admin wrote:
Did you ever consider that one possible reason is
that Solaris X86
simply doesn't have all the drivers - I have a
laptop, and it's less
than 2 years old, which Solaris will install on, but
there simply are no
network drivers for it. So it would take much more of
my time to
> Ferdinand O. Tempel wrote:
> > for Debian packages to work, you need to adopt the
> entire Debian way of thinking.
> > Bells and whistles and all.
>
> > But it'll be a fight against both the OpenSolaris
> community and
>
> What are the core differences WRT the "way Sun has
> traditionally done
> Joerg Schilling wrote:
>
> > Don't expect things to work if you adopt to "the
> e entire Debian way
> > of thinking".
> >
> > As long as Debian compiles software on Linux-2.4
> 4 and expects the
> > resulting binaries to work on Linux-2.2, the did
> d not yet grok how to
> > deal with evolvoi
On 7/21/05, UNIX admin <[EMAIL PROTECTED]> wrote:
> This is the exact same approach that Linux takes. The only way to patch a
> software subsystem through the OS interfaces is to do `rpm -u` which goes and
> [I]replaces the entire software subsystem[/I] in order to "update" it.
> This is very muc
--- Glynn Foster <[EMAIL PROTECTED]> wrote:
> Hey,
>
> > Yes, of course it would have been better!
> > We have clearly identified an area where there
> things could improve,
> > but that's not the main problem.
> >
> > The main problem is that people who converted are
> trying to push "what
>
Hey,
> Yes, of course it would have been better!
> We have clearly identified an area where there things could improve,
> but that's not the main problem.
>
> The main problem is that people who converted are trying to push "what
> they're used to" from Linux, thereby reinventing hot water.
Bull
"Ferdinand O. Tempel" <[EMAIL PROTECTED]> wrote:
> >Don't expect things to work if you adopt to "the entire Debian way of
> >thinking".
>
> Actually, I do. I don't care much about what the debian project "thinks" or
> "expects", and I doubt it's what you claim as:
>
> "As long as Debian compiles
Joerg Schilling wrote:
John Plocher <[EMAIL PROTECTED]> wrote:
One of the core values of Solaris is that sharable things really
should be shared.
Did you check how many versions of libz are available on Solaris?
Less than there used to be, more than there should be. It's a design
goal,
> Wouldn't it be even better to not have to make that
> judgment call? Now
> don't get me wrong, the stuff done by the likes of
> Blastwave and
> Sunfreeware were *hugely* important - I'd just like
> to see us acting as
> a combined community, working on porting the
> software, integrating the
> so
> Did you ever consider that one possible reason is
> that Solaris X86
> simply doesn't have all the drivers - I have a
> laptop, and it's less
> than 2 years old, which Solaris will install on, but
> there simply are no
> network drivers for it. So it would take much more of
> my time to get to
>Don't expect things to work if you adopt to "the entire Debian way of
>thinking".
Actually, I do. I don't care much about what the debian project "thinks" or
"expects", and I doubt it's what you claim as:
"As long as Debian compiles software on Linux-2.4 and
expects the resulting binaries to w
John Plocher <[EMAIL PROTECTED]> wrote:
> One of the core values of Solaris is that sharable things really
> should be shared.
Did you check how many versions of libz are available on Solaris?
Jörg
--
EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin
[EMAIL PROTECTED]
> when I want to install libbonobo as a dependency for
> another package, I want just that and one copy on my
> system, and let every package that needs it find it.
> company ticker appended to package names is one
> ridiculous idea, when a simple package name would do
> just fine. All, pkg-get wou
Hi John,
On Wed, 2005-07-20 at 19:31, John Plocher wrote:
> Without this ability to manage the evolution of shared components,
> it becomes extremely difficult to reuse them, because asynchronous
> development almost guarantees that a change to the shared component
> will break a consumer, and th
--- Bob Palowoda <[EMAIL PROTECTED]> wrote:
> > On Wed, 2005-07-20 at 15:17, Torrey McMahon wrote:
> > This is a real problem when multiple copies of the
> > same interface
> > get dragged into a single processes.
> >
>
> Interesting as I'm seeing this in the kde 3.4.1
> from blastwave.
> Mu
> Interesting as I'm seeing this in the kde 3.4.1 from
> m blastwave.
> Multiple copies of different versions of libstdc++
> compiled
> into some of the daemon process. I wonder if this is
> what is causing
> it to crash.
if 'ldd' shows both .5 and .6 referenced, it will lead to trouble sooner t
> On Wed, 2005-07-20 at 15:17, Torrey McMahon wrote:
> > Which is a bad thing why? (Please don't say space.)
>
> Lets take a real example which was discussed this
> week at Sun,
> the same thing can be applies to other libraries.
>
> libldap is in /usr/lib and is linked to the NSS/NSPR
> librarie
One of the core values of Solaris is that sharable things really
should be shared.
This begs the related questions of
What constraints does this place on the shared component, and
What do you do when the shared thing evolves
Sun has a proactive development methodology that looks
> Which is a bad thing why? (Please don't say space.)
if you can prove duplication is a good thing, I will be happy.
anyway, its bad because it:
1. wastes network bandwidth for downloading packages which are already
installed, working, used, tested.
2. wastes time re-installing the packages wh
when I want to install libbonobo as a dependency for another package, I want
just that and one copy on my system, and let every package that needs it find
it. company ticker appended to package names is one ridiculous idea, when a
simple package name would do just fine. All, pkg-get would check
25 matches
Mail list logo
