Re: broken after libtoolswitch on

[Thomas Walter <walter at pctc dot chemie dot uni-erlangen dot de>]

>>>>> "Dave" == Dave Morrison <dave@bnl.gov> writes:

    Dave> Hi-

    Dave> Isn't the situation easier to handle if you do both?  I mean, install a
    Dave> slew of separate gsl libraries and a conglomerate "convenience"
    Dave> library?  No fancy conditionals, just plain vanilla automake.  I made a
    Dave> very, very simple package to try this and it seems to work OK.  Grab it
    Dave> at

    Dave> http://www.phenix.bnl.gov/~dave/multilib-1.0.tar.gz

    Dave> if you'd like to see what I did.

    Dave> Steve ROBBINS wrote:

    >> Is having a choice something that you want to support indefinitely?  Or is
    >> it just an issue during a transition from one model to the other?

Hello,
I solved/tracked down the problem.

You are correct if you say it is best to build both.  And 'libtool'
does that if you change 'libtoolswitch.pl' to translate
     pkglib_LIBRARIES
into
     pkglib_LTLIBRARIES
Then you have the shared libraries and the static libraries in a
subdirectory '.libs' or '_libs' (depends on your system).

Now a missing feature of 'libtool' shows up:
It can add the contents of a static only library into another static
library.  For this 'libtool' extracts the contents from the static
library located in for example 'utils' (this library is never build as
shared, following the comments in the Makefile) by reading
'libgslutils.al'.

The missing feature of 'libtool' is now:
    It has no rules to do the same by using the information located in
    for example 'vector/libgslvector.la'.

If this feature will be there then you would get:
    - shared libraries for each sub-directory
    - static libraries for each sub-directory
    - a static library 'libgsl.a' in the top-level.

Bye
Thomas

PS:
I have the impression the script 'gsl-config' is totally broken.
I suggest to take 'gtk-config' as template for a new version.

-- 
Warum hat eine Leiterbahn keine 90 Grad Winkel sondern 2 mal 45 Grad?
Weil die Elektronen sonst immer gegen die Wand laufen 8-)))

----------------------------------------------
Dipl. Phys. Thomas Walter
Inst. f. Physiklische Chemie II
Egerlandstr. 3				Tel.: ++9131-85 27326 / 27330
91058 Erlangen, Germany			email: walter@pctc.chemie.uni-erlangen.de