another proposed crosstool project

[Robert P. J. Day]

  as (sort of) an adjunct to yann's "ct-ng" (crosstool next
generation) project, i'd like to propose something way over on the
other side of the spectrum that's based on what i was once calling my
"mini" crosstool script, or "mini-ct".

  as opposed to yann's project (and, in fact, dan's current version),
the point of mini-ct was/is to strip the workings of crosstool down to
its absolute basics so that it does the minimum possible work to build
a toolchain.

  the point of this is to extract the essense of what it means to
create a cross-compiler, without any fancy lights, bells or whistles,
and to keep the configuration options to a minimum as well.  that
means ripping out anything that's not absolutely essential to the
construction of a toolchain.  that would include things like locating
and downloading packages on the net -- you would be responsible for
getting everything beforehand and putting that stuff in the downloads/
directory.

  a lot of the things that are normally choices would be fixed, to
represent the idea of building the *best* choice of toolchain.  for
instance, the toolchain would be hardcoded to be "sysroot"ed since
that's just plain better than *not* being sysrooted.  in addition, it
would automatically build shared libs, wouldn't play games with make
options, be able to compile just C and C++, and so on.  you get the
idea.  KISS -- keep it simple, stupid.

  over the next few posts, i want to describe the minimum work that
would be done at each phase in the process, and see what people think.
you'll see where i'm going with this fairly quickly, i think.

rday
-- 
========================================================================
Robert P. J. Day
Linux Consulting, Training and Annoying Kernel Pedantry
Waterloo, Ontario, CANADA

http://fsdev.net/wiki/index.php?title=Main_Page
========================================================================

--
For unsubscribe information see http://sourceware.org/lists.html#faq