[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [cdt-dev] Breaking the Makefile APIs
- From: "Alain Magloire" <alain@xxxxxxx>
- Date: Wed, 29 Apr 2009 15:04:48 -0400
- Delivered-to: email@example.com
- Thread-index: AcnI5mdlp+OMkJlXT+WHBAurQCsE7QAAmw5wAATIE/A=
- Thread-topic: [cdt-dev] Breaking the Makefile APIs
> From: cdt-dev-bounces@xxxxxxxxxxx [mailto:cdt-dev-bounces@xxxxxxxxxxx]
> Behalf Of Ed.Swartz@xxxxxxxxx
> Sent: Wednesday, April 29, 2009 12:48 PM
> > [alain] I am doing some work on the Makefile parser, I do not think,
> this point, folks care too much about this, so I intend to break the
> (and probably will do it again 8-).
> Yes, we care. A few of our products depend on the Makefile APIs (from
> 5.0, though) to read and modify Makefiles. These plugins do provide
> so please consider there may be a number of clients silently depending
> it. ;)
8-) ok noted.
I should have put a warning in the original code, something like "API
subject to change etc....", too late now.
> > [elena] In this case design paper on wiki would be usefull...
> I would second this. It would be good to have a heads-up about what
> of changes to expect when we migrate to newer versions of CDT.
Sigh... docs... Do you really need them...
I did not realize that this code was being use. It is now out of the
question to do any update so close to release.
I will continue to clean things up in the background and submit later.
> BTW, we didn't really encounter huge problems using the 5.x Makefile
Well... Thank you!
> though we needed some utility classes to make modifications easier
> "add a new directive after this existing directive", "add a command
> this rule", "modify the dependencies of this target", etc.).
Yes, this is interesting, the ability to write from the API, it was not
part of the original plan since it was considered to be read-only.
> What kinds
> of changes are involved in your rewrite?
Better Support for GNU makefile syntax and working example of POSIX Make