On 1 June 2007 at 15:16, Steve Langasek wrote:
| reassign 419742 quantlib-swig
| found 419742 0.3.14-1
| thanks
|
| Dirk,
|
| This reassignment of an RC bug to the toolchain has blocked all of lenny for
| another day, because it happened right when glibc was ready to enter
| testing. Please consider when reassigning bugs to the toolchain whether
| these bugs are RC for the toolchain, as distinct from being RC for your
| package.
Aie. I really want glibc to migrate so I'm going to be mad at myself
here. Sorry.
| I'm reassigning this bug back to quantlib-swig, because this *is* the
| package with the RC bug; as Thiemo noted, the build failure happens as a
| consequence of the special-case use of -O0 on mips* in the package's build
| rules. The build failure is of this package, and it can be resolved without
| any changes to the toolchain.
Hm, I must have overlooked this in his mail. Building under the default -O2
will succeed on mips as it used to? I shall prepare a fixed package rather
quickly then.
| The toolchain, further, works fine on mips for almost all other packages.
| There are limits to the number of symbols the toolchain supports on mips,
| yes, but even mozilla is able to work within those limits -- this simply
| does not make the toolchain unreleasable (we've released the toolchain with
| this limitation for as long as we've been releasing mips).
Ok, it also used to build QL. It could just be that because of the
templating, additional code may have lead to an increase over the limit --
but let's worry about that if and when we see it. First I'll do a new
package.
| The eaccess issue in glibc might be considered RC, though I haven't seen
| anything before now to indicate people consider it a bug in glibc instead of
| bugs in the packages with conflicting declarations for eaccess(). The
| number of packages affected by this issue is quite small (compared to the
| number of packages that are waiting on glibc right now for testing
| propagation, definitely!), and in all such cases involves unnecessary
| shadowing of a system declaration, so unless someone shows that the new
| glibc declaration is just plain *wrong*, I don't see any reason yet to
| consider it a glibc bug.
Agreed.
On to Brown Bag mode and preparing a fresh ql-swig.
Thanks, Dirk
--
Hell, there are no rules here - we're trying to accomplish something.
-- Thomas A. Edison
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
