[Bug 428567] Review Request: ETL-devel - Extended Template Library

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug report.

Summary: Review Request: ETL-devel - Extended Template Library


https://bugzilla.redhat.com/show_bug.cgi?id=428567





------- Additional Comments From rc040203 at freenet.de  2008-01-23 23:49 EST -------
(In reply to comment #16)
> Defines: not yet, help welcome.
Provided how this package is designed, fixing this is close to impossible.

One way to work around this would be to rename all defines and to use a second,
private, manually written autoheader, which would be installed into $(libdir)
instead of $(includedir). Packages using ETL then would be required to use
-I$(libdir)/ETL to pickup this header (c.f. how glib2 and gtk2 install their
headers)

> ETL-config is generated from ETL-config.in at configure time.
This is pretty easy to fix. The key is not to store any configuration info in
files which are not in multilib'ed dirs. I.e. not to store @libdir@ in
$(bindir)/ETL-config, but to let $(bindir)/ETL-config pick up this info from a
different location.

A common trick is to implement $(bindir)/*-config as wrappers around pkgconfig.
(*.pc's are being stored into $(libdir)/pkgconfig (libdir is multilib'ed. e.g.
*/lib on ix32 and */lib64 on ix64)

> Removing it
> upstream would be OK, since synfig/studio don't use it as far as I can see.
Exactly this is the cause of your problems with the #defines. 
If they were, you could move these headers into $(libdir) and let ETL.pc use
Cflags= -I$(libdir)/ETL or similar.


-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.