[Scons-dev] catching up

[anatoly techtonik]

On Sun, Aug 17, 2014 at 9:54 PM, Gary Oberbrunner <garyo at oberbrunner.com> wrote:
> There's probably not much point in my trying to respond to many of the
> threads that have been running here in the last couple of weeks; if there
> are things that should be addressed please bring them up.

https://bitbucket.org/scons/scons/pull-request/169/once-more-suppress-warning-about-missing/diff
Thank you. =)

> The other project is toolchain; I've started working on this and would like
> to spend my time on that if possible; it's where I can probably bring the
> most value to the project.  I actually don't think it'll take all that long
> to hammer out the design and first test implementation; there'll be a bunch
> of tool-porting after that but we can split that up.

Tool porting? I think you've chosen a bad strategy. Let's split that abstract
toolchain concept into something that could be discussed.

1. Before making any significant changes, I want to make sure that all
    tests pass.
2. Before making any significant changes, I want to make sure that
    running tests is a viable task. Right now a single run is 15+ minutes,
    which suxx. We need to identify the bottleneck. I thought that it is
    disk access, so I've set up a RAM drive for running them, but there
    are other things that should be done - smart DefaultEnvironment
    initialization, for example.
3. Before making anything from scratch (which is a process we all enjoy)
    the more valuable activity would be to do a code review and comparison
    for Parts. Once it is published, implementation won't be a problem.

Nevertheless, actions are more important than talks. As for Mercurial
history, I still look for a chance to cut the Docbook toolchain bloat from
repository, so if you really don't care much about history.. )