On Sun, Mar 30, 2014 at 4:21 AM, Jeff Allen <ja.py@farowl.co.uk> wrote:

Congratulations to Frank on getting this far. I'm sure this will help our pace.

I have installed with "java -jar jython-installer-2.7-b2.jar", on Windows 7, including source and documentation. This went smoothly.

I ran the regression tests. With Java 7 I get 22 failures on Windows, which is up quite a bit from what I have come to expect (about 12) in the development environment. Many are familiar, including the float ones I'm working on at present, and some are cascades (due to non-deletion of files), but some look like they may be problems with building the installer.

In particular some relate to missing idna codecs I believe Jim added. I wonder if the related jars are not getting distributed, or onto the class path?
Weird - we should definitely look into that.

I'll send the output of "jython -m test.regrtest -e -w" (193K) separately to Frank and anyone who asks.
That's good to know - thanks for doing this investigation!

Should we have failing tests in (even) a beta release? I'd feel happier about inserting skips if I knew there was a "FIXME thermometer" somewhere that would move me to treat the underlying causes.
We really shouldn't have failing tests in beta releases - but since it's been a year since beta1 and there are so many improvements I thought it was a good idea to just do it and fix things up by beta3 (I hope).