asy linked to libGLX.so : what is that ?

[Ken Moffat]

On Wed, Mar 22, 2023 at 02:52:55PM -0600, Karl Berry wrote:
>     > > Looking at past builds, I don't think I've ever had a libGLX.so.0
>     > > and yet the binary has worked in the past[1.].  
> 
> FWIW, libGLX.so.0 was included in our x86_64-linux asy binaries for at
> least the last few years ...
> 
> $ ldd /usr/local/texlive/2022/bin/x86_64-linux/asy|fgrep libGLX
> libGLX.so.0 => /lib64/libGLX.so.0 (0x00007fd9e5a2b000)
> $ ldd /usr/local/texlive/2021/bin/x86_64-linux/asy|fgrep libGLX
> libGLX.so.0 => /lib64/libGLX.so.0 (0x00007f041cf99000)
> $ ldd /usr/local/texlive/2020/bin/x86_64-linux/asy|fgrep libGLX
> libGLX.so.0 => /lib64/libGLX.so.0 (0x00007f6bf0dd6000)

Well, I certainly installed the 20200406 binary but it looks as if I
only did that to measure the size of a full install and in later
years I have let someone else test a minimal install of the binary.

Thanks, this proves my assertion to a colleague that none of our
users try to run asy from install-tl-unx - and I suspect that the
only user of asy in BLFS is me (to test it).

It seems my past explorations of pretest were in earlier years, when
it was all after I'd finished my testing for a release of the BLFS
book.

In March 2020 I had tested source builds of 2.82 and 2.83 on a different
machine, and after that I tested source builds on that same machine.

Anyway, I've now been pointed to where libGLX comes from, and with a
straight meson/ninja install (letting it overwrite various libs from
mesa with newer versions, this is only a development machine) the 2D
test works.

ĸen
-- 
Often there is no more than a little plaque to reveal that, against
all gynaecological probability, someone very famous was born halfway
up a wall.                 -- Equal Rites