Re: [Boost-testing] RC 1.34 build broken since at least 20070315-072956 UTC

From: David Abrahams (dave_at_[hidden])
Date: 2007-03-16 12:04:14

on Fri Mar 16 2007, Martin Wille <mw8329-AT-yahoo.com.au> wrote:

> David Abrahams wrote:
>> on Fri Mar 16 2007, David Abrahams <dave-AT-boost-consulting.com> wrote:
>> 
>>> on Fri Mar 16 2007, Martin Wille <mw8329-AT-yahoo.com.au> wrote:
>>>
>>>> Markus Schöpflin wrote:
>>>>> I get the following error message in my builds:
>>>> That could explain why my x86 test runs were so lightning fast.
>>>>
>>>>
>>>> I get the following on Linux x86_64 (the test continues, though):
>>>>
>>>> error: No best alternative for /python-2.4
>>>>      next alternative: required properties: (empty)
>>>>          matched
>>>>      next alternative: required properties: (empty)
>>>>          matched
>>>
>>> Can you please post your user-config.jam (and site-config.jam if you
>>> use one) and the initial output of running bjam with
>>> --debug-configuration?
>> 
>> Also, please respond to my comments at
>> https://zigzag.cs.msu.su:7813/boost.build/ticket/119.  You might add
>> yourself to the Cc list.
>
>
> After changing my firefall settings and ignoring the certificate 
> problems, I am moderately confused now.
>
>
> I have no user-config.jam and no site-config.jam. My test-config.jam

I've never heard of test-config.jam.  Is that some new thing?

> is on the page you ask me to respond to. I have no idea how to
> respond to that web page.

You log in to the trac (see https://zigzag.cs.msu.su:7813/boost.build)
and then you can edit it.

> Apparently, my email address is already on that page 
> where it is free for the bots to harvest. 

Sorry, I didn't know that Volodya's Trac exposed those without
mangling.  I'll remove it.

> Boost.Build has been changed since the time I reported that problem
> (in order to fix that problem).  So I doubt adding
> --debug-configuration now will help debugging the problem from that
> time.

Is the problem fixed?

> Adding --debug-configuration now results in:

That bjam invocation clearly didn't have toolset=... or
--toolset=... in the command-line (was not done using toolset
autoconfiguration) so it's not really a test of the same problem.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com