New golden master

[Luigi Ballabio-2]

Hi all,
        new golden masters for the 0.3.1 release are available at

http://quantlib.org/gm

Play with them and tell me if there are any last-minute bugs. Otherwise,
we'll go ahead and release next week.

Dirk: could we try and run the QL-Python tests on the Debian machines after
building? Just curious to see whether they actually run...

Bye,
        Luigi

[Jens Thiel]

> Hi all,
> new golden masters for the 0.3.1 release are available at
>
> http://quantlib.org/gm
>

...and so are the release-candidates for QL.NET, split into a binary release
(containing compiled assemblies and documentation) and a source-only
installer.


Jens.

[Dirk Eddelbuettel]

On Tue, Jan 14, 2003 at 10:11:42AM +0100, Luigi Ballabio wrote:
>
> Hi all,
> new golden masters for the 0.3.1 release are available at
>
> http://quantlib.org/gm

Nitpick: The naming sucks. I called the last one 0.3.0.9 [ so that it sorts
higher than 0.3.0 but lower than the expected 0.3.1 ] and had to make this
0.3.0.9.1.  I think it is a bad idea to replace one tar.gz with another that
is different, yet sports the same version.

> Play with them and tell me if there are any last-minute bugs. Otherwise,
> we'll go ahead and release next week.

QL and QL-Python built fine and are now on the Debian master archive.

> Dirk: could we try and run the QL-Python tests on the Debian machines after
> building? Just curious to see whether they actually run...

Yes, should be fine now. Yours truly had not quite grasped the meaning of a
boolean NOT in the relevant test....

Dirk

--
Prediction is very difficult, especially about the future.
                                             -- Niels Bohr

[Luigi Ballabio-2]

At 09:03 PM 1/14/03 -0600, Dirk Eddelbuettel wrote:
>On Tue, Jan 14, 2003 at 10:11:42AM +0100, Luigi Ballabio wrote:
>Nitpick: The naming sucks. I called the last one 0.3.0.9 [ so that it sorts
>higher than 0.3.0 but lower than the expected 0.3.1 ] and had to make this
>0.3.0.9.1.  I think it is a bad idea to replace one tar.gz with another that
>is different, yet sports the same version.

Well, these are (and the previous were) golden masters---meaning that they
_are_ version 0.3.1. If the current tarballs don't show any problems, they
will be released as they are. It is unfortunate that the previous tarballs
had to be corrected, but at the time the hope was that they _were_ version
0.3.1.
However, you're right---we'll have to find a way for you to say to the
Debian system "these are 0.3.1, but not quite". A quick look at dselect
shows me that the version of gcc-3.2 installed on sarge is 3.2.1-0pre3.
Maybe this could work for next version's gm? Something like 0.3.2-0pre1,
0.3.2-0pre2 if something needs to be changed, and 0.3.2-1 when it's final?

Bye,
         Luigi