|
Boost : |
From: Braden McDaniel (braden.n.mcdaniel_at_[hidden])
Date: 2002-04-11 14:45:12
On Tue, 09 Apr 2002 09:25:55 -0400, David Abrahams wrote:
> Here's what I sent Dan'l that prompted this posting. Thanks again to
> Dan'l for his efforts.
>
> -Dave
> --------
>
> Hi Dan'l,
>
> This stuff looks really great. The installing.html page is a little
> complicated (so many options, rather technically described!), though.
> Why don't you put this in the files area and post an announcement to the
> list? I would prefer it if we could get some other commentary and
> perhaps go through a short refinement cycle if needed.
Dan'l, thanks for getting the ball rolling again. First, one minor
correction: the autotools comprise autoconf, automake, and libtool.
I'm not sure how fond I am of the terms "build-to-develop" and
"build-to-install". I am interested in what you call "build-to-install",
but I want to install for development purposes. Where I develop,
deployment is a normal responsibility of the build process. The
descriptions of these terms suggest that one categorizes users and the
other developers, which isn't accurate at all. As I see it, the
difference to users is simply that an autotools build offers a feature
lacking in the Jam build: deployment. This is practical with the
autotools because for POSIX systems, there are better understood
notions of how software should be installed; and as a result of this,
the autotools provide support for deploying software per those
conventions. But deployment is a perfectly reasonable feature to want
regardless of whether you're developing software or just installing it.
I'm inclined to agree with Dave about the installing page. I think its
complexity is derived from being excessively pedantic. I think most users
will know what they want/expect re. building and installation. Thus it
should be adequate simply to describe what is available, rather than
attempt to exhaustively categorize Boost users.
***
I made some effort toward an autotools build, but haven't had time to
follow up and get it into a working state. There's the issue that I
don't need this myself quite yet, so I'm not as motivated as I might be
otherwise.
Braden
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk