[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: build system work

From: Phil Holmes
Subject: Re: build system work
Date: Fri, 4 Mar 2011 12:34:36 -0000

----- Original Message ----- From: "Graham Percival" <address@hidden>
To: "Lily devel" <address@hidden>
Cc: "Phil Holmes" <address@hidden>
Sent: Friday, March 04, 2011 12:16 PM
Subject: build system work

There is widespread dissatisfaction with our current build system.
Replacing it is estimated to take at least 200 hours, but that's a
total shot in the dark.  However, a few people have expressed interest
in working on this.

IMO, such a project should begin by figuring out what the current
build system does.  Absolutely nobody understands the entire thing,
and it's not particularly documented.  I'd estimate that we should
spend about *50 hours* documenting what the current system does,
before even beginning any serious discussion about making a new one.

To that end, I've just added a new CG 12 Build system notes.  The goal
is to have a "wiki-like" dump of whatever info we suspect about the
build system.  If you think you have some info that you can add, just
add it.  Don't worry about whether it's correct or not.  If you see
some existing material in there that you think is incorrect, then
we'll talk about it and either correct it, or just dump your alternate
theories about what's happening, so that the reader can see both

My suggestion is to spend no more than *1 hour* at once.  In my
experience, it's really easy to get into trouble with the build
system, and to start going around in circles and stuff.  There is no
time pressure on this project, so if you start to get stuck, just stop
working and come back to it in a few days.  And/or send an email to
the other people working on this, so that somebody else can look at it
with a fresh viewpoint.

If you're interested, take a look at the new CG 12 chapter (it's only
in git at the moment, not only).

- Graham

I'm wondering whether we should actually start with a real Wiki - like the one used for the regtests. Doing this would mean that any work done would be immediately visible, rather than having to pull git and make docs, which would save me oodles of time (a technical phrase).

What do you think?

Phil Holmes

reply via email to

[Prev in Thread] Current Thread [Next in Thread]