Subject: Re: [Boost-build] [future] Implementation language(s)..
From: Stefan Seefeld (stefan_at_[hidden])
Date: 2016-10-25 10:39:51
On 25.10.2016 10:30, Klemens Morgenstern wrote:
> The reason I proposed that, is that I think it would have several
> 1. small, local extensions, e.g. some project-specific code generator
> 2. complete toolsets, i.e. plugins that should be shipped with the
> build system
> I'd prefer to have the first stuff in a small scripting language and
> the second C++. I would not be mixing the langauges in the shipped
> plugins (pure C++ there), but provide the scripting language for
> user-extensions if you will.
> If we only used python, we'd have a slow build system; if we'd only
> used C++, the extensions would be painful for the users.
Please demonstrate that Python for "small, local extensions" would yield
a slow build system. I think we can worry about that when we run into
that. Anything else is premature optimization.
(And I honestly have had enough of projects that use cryptic ad-hoc
languages that only the original implementers really understood. We
really need to focus on productivity, i.e. the time it takes for *other*
people to solve *their* problem with a given tool, which is clearly
helped by using a *common* and *well-understood* language.)
-- ...ich hab' noch einen Koffer in Berlin...
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk