RPM Community Forums

Mailing List Message of <rpm-users>

Re: Building internal LUA fails (Was: RPM 5.1.5 released)

From: Jeff Johnson <n3npq@mac.com>
Date: Tue 14 Oct 2008 - 16:35:10 CEST
Message-id: <4C35F376-2151-490E-839C-CB23E1540893@mac.com>

On Oct 14, 2008, at 10:24 AM, Eric MSP Veith wrote:

> Hash: SHA1
> Hello,
> On Tuesday 14 October 2008, Jeff Johnson <n3npq@mac.com> wrote:
>> I'd *love* to provide a guarantee of interoperability w rpm4, but I
>> cannot
>> do that unilaterally when there is a active fork in progress.
> Well, some systems here a currently using RPM version 4.4.8 from  
> rpm5.org,
> which means that I could report on switching from rpm4 to rpm5, if  
> that
> would be appreciated. But I don't know how much that helps in  
> regard to the
> rpm.org fork.

Holler if you need help switching to rpm-5.x. I'm very interested
in letting rpm-4.4.x die gracefully, switching to rpm-5.x (or reverting
to rpm4 from rpm.org) are the only available upgrade paths, and
both choices aren't exactly drop-in replacements.

>> Personally, I think its Really Bad Craziness to link -lrpm into gdb.
> Why is it done, anyways? Debugging RPM packages? *scnr*

gdb needs to map executable/library paths back to detached symbols that
may be contained in a -debuginfo package.

While that is a perfectly reasonable usage case for gdb, there are  
easier ways to accomplish the mapping without stuffing Berkeley DB  
into gdb.

There are huge compatibility issues, like the segfault I'm seeing  
daily, that
are introduced into gdb.

But I'm quite sure that GDB packagers will see the error of their  
ways if/when
developers start seeing the same segfault that I am seeing.

73 de Jeff
Received on Tue Oct 14 16:35:41 2008
Driven by Jeff Johnson and the RPM project team.
Hosted by OpenPKG and Ralf S. Engelschall.
Powered by FreeBSD and OpenPKG.