[boost] Re: boost v 1.33 ?

From: Robert Ramey (ramey_at_[hidden])
Date: 2005-02-01 11:26:07

Aleksey Gurtovoy wrote:
> Robert Ramey writes:
>> Aleksey Gurtovoy wrote:
>>> Robert Ramey writes:
>>> [...]

>> I don't believe the number of red squares is an accurate reflection
>> of the state and how much work there is to do.
>
> I'd say it's proven to be pretty accurate (excluding the extreme
> cases when everything is red due to Boost-wide or
> environment/configuration issues).
>
>> Many of these problems have a common cause.
>
> Across all the libraries?

Boost.Test is getting upgraded.  Its "priviledged" position in the
dependency hiearchy means that an issue with this library ripples accross
all libraries.  This is very apparent when one investigates the the red
squares.

> Note that currently the reports are missing a significant share of
> (our) Win32 toolsets which will definitely add more colors to the
> picture.
>
>> In at least one case, there's a new compiler which wasn't
>> tested at 1.32 at all.
>
> Which one? If it wasn't tested with 1.32, it shouldn't appear as a
> regression.

This is labeled as darwin.  Its different that the previous one named
darwin.  Its Codewarrior 9.3 for the Mac OS X

>
>> I believe that a new release have improvements over
>> the current one and have no regressions.
>
> If you meant "should have", I agree 100%.

I believe we are much closer to this than a cursory examination of the test
matrix would suggest.

Robert Ramey