Re: Uncoordinated quantlib transition
On 24 May 2012 at 01:34, Cyril Brulebois wrote:
| Hi Dirk,
| it seems like we have an uncoordinated transition, from libquantlib-1.1
| to libquantlib-1.2. That can be seen on the excuses page, and that
My packages (listed below) are the only users of libquantlib. In the past (as
I've maintained this for a decade) I simply use 'libquantlib' but I was told
that versionsed libraries would be "better". I am not convinced that that is
| explains why your package isn't migrating:
| | quantlib (1.1-2 to 1.2-2)
| | Maintainer: Dirk Eddelbuettel
| | 69 days old (needed 10 days)
| | out of date on armel: libquantlib-1.1 (from 1.1-2)
| | out of date on mipsel: libquantlib-1.1 (from 1.1-2)
| | out of date on s390: libquantlib-1.1 (from 1.1-2)
| | out of date on armhf: libquantlib-1.1 (from 1.1-2) (but armhf isn't keeping up, so nevermind)
| | Not considered
| See: http://release.debian.org/britney/update_excuses.html#quantlib
| Trying to remove the old binary packages would lead to breaking the
| following packages, which therefore need binNMUs, hoping they won't
| | # Broken Depends:
| | quantlib-swig: quantlib-python [mipsel s390]
| | quantlib-ruby [mipsel s390]
As can be seen on https://buildd.debian.org/status/package.php?p=quantlib-swig
both builds simply die of resource starvation. There is /a lot/ of C++ here
in a single large Swig generated file, and it needs sufficient memory to
Someone needs to schedule this with sufficient build resources. We had this
issue before on other arches.
| | rquantlib: r-cran-rquantlib [armel armhf]
That failure (which is echoed on a lot of other r-cran-* packages) is a
complete mystery to me, and to upstream.
The build process is /highly/ automated at the R side too, and works on
hundreds of other source packages across a dozen other arches.
As can be seen on https://buildd.debian.org/status/package.php?p=rquantlib
two variables simply do not expand.
But I do not know why it fails here, and I have no suggestion.
| Next time, please coordinate with us, we have been trying to get the
| message across during the past few years through messages to dda@, and
| we have documentation on the process:
| xapplication/pgp-signature [Click mouse-2 to save to a file]
Dirk Eddelbuettel | firstname.lastname@example.org | http://dirk.eddelbuettel.com