Boost logo

Boost :

Subject: Re: [boost] [test] Looking for co-developer/maintainer
From: Alexander Lamaison (awl03_at_[hidden])
Date: 2014-01-22 05:12:37


"Vicente J. Botet Escriba" <vicente.botet_at_[hidden]> writes:

> Le 20/01/14 20:40, Richard a écrit :
>> [Please do not mail me a copy of your followup]
>>
>> I can take over the library.
> Hi Richard,
>
> it is clear that you and Gennadiy will not work together on the same
> library.
>
> What about forking the Boost.Test library into a new Boost.XTest with
> the contents of the Boost.Test release and your documentation and
> taking care of the issues associated to the release branch?
>
> What the Boost community thinks about this proposition? Will
> competition between two libraries be a good thing for Boost?

While, on the one hand, it would make things so much easier, I'm not
keen on this idea (though I'd enthusiatically join in if it went ahead).

Gennadiy has created a technically excellent library. I'm even really
excited about the new API that is waiting in the wings. The problem is
not the code, but, rather, that Gennadiy is too close to his 'baby' to
see it objectively from the perspective of its users.

The excellent code is let down by:
- documentation written for the implementor, not for the user
- unfixed bugs (as Andrey pointed out, users couldn't care less if
  something is fixed in trunk)
- communication - questions on the lists go unanswered for months

Forking Boost.Test with different maintainers would resolve these flaws,
but the new fork would be the same code + some bug fixes + readable
documentation. That seems silly. Forks are usually for when developers
want to take the _code_ in two incompatible directions.

Alex

-- 
Swish - Easy SFTP for Windows Explorer (http://www.swish-sftp.org)

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk