Hubert Figuière wrote on Mon, 30 Apr 2012 11:59:19 -0700:
>> +AC_ARG_WITH([standard_symbols_l],
>> + [AS_HELP_STRING([--with-standard-symbols-l], [use font family 'Standard Symbols L' for Symbol])],
>> +[
>> + abi_cv_standard_symbols_l="$withval"
>> +])
>> +if test "$abi_cv_standard_symbols_l" = "yes"; then
>> + AC_DEFINE([WITH_STANDARD_SYMBOLS_L], [1], [Use font family 'Standard Symbols L' for Symbol])
>> +fi
>> +
> This is probably as bad. Because you again assume a runtime condition at
> compile time. And this is not good at all.
I don't assume, I request.
Well, maybe I should have used AC_ARG_ENABLE instead of AC_ARG_WITH for
reasons of clarity and comprehensibility, but by giving this option you are
building with the request to use 'Standard Symbols L' for Symbol. What's bad
about it? Isn't that the same like a build with request to use a library that
isn't available at run time, or a GTK3 build for a GTK2 machine?
Ingo
Received on Mon Apr 30 21:22:49 2012
This archive was generated by hypermail 2.1.8 : Mon Apr 30 2012 - 21:22:49 CEST