[RFC] solutions for maintaining the install target, will affect developers work-flow

Skip to first unread message

Mike Mattie

Mar 19, 2007, 8:35:07 AM3/19/07
to parrot ML

I have been working on a series of cleanup patches that lay the ground towards improving the install target,
and feature enhancing .include, and load_bytecode.

The patches have been accepted and up to this point do not change parrot's behavior for existing code. I have laid out
a two-step of proposal, in in rt # 41908 ( please read the ticket for the proposals ). This mail is a pointer to that
RT. link to a text/plain of the message body below:


The first proposal regards an environment variable PARROT_LOAD_PREFER that can be accepted or rejected independent of
the second which is a much larger scale change, and slightly more intrusive.

I would like feedback at this point before moving forward on the proposals. I am not welded to these proposals, and
even for my own learning I would like to learn how other dev's would approach the problems I am working on.

a closing section expands on what is needed to insulate the src/library.c API properly, and fix a minor race. This
insulation is not linked to the proposals technically. My knowledge of parrot is still very limited,
to limited to gauge the impact across the rest of the tree, so I will hold fire before making up patches.

Mike Mattie (coder...@gmail.com)

Reply all
Reply to author
0 new messages