Will Estes

If we're going to provide an implementation of a function, we should use the libobj mechanism. On Sunday, 24 June 2018, 3:39 am -0700, "Kang-Che Sung (宋岡哲)" wrote: > @GitMensch...

libcob? On Sunday, 24 June 2018, 6:29 am -0700, "Kang-Che Sung (宋岡哲)" wrote: > @westes I think the licensing problem (it's LGPL) already prevents us from including libcob's header directly...

libobj as handled by automake. It's the way to provide portability functions when absent from the system at compile time. On Sunday, 24 June 2018, 6:36 am -0700, "Kang-Che Sung...

Yes. When autotools are run from autogen.sh, tableopts.am needs to be present. The only way to manage that is to track tableopts.am in version control. On Friday, 14 September 2018,...

I have some vague memory that there was an issue with that, but I can't recall anything specific enough to raise an objection. Certainly try it and see what happens....

Yes, if it's going to work, it would fit best in autogen.sh On Sunday, 16 September 2018, 12:23 am -0700, Simon Sobisch wrote: > @Explorer09 it will be done during...

It makes sense to switch to github actions. For now, I'd say ignore the mswindows and osx builds since they consume minutes at higher rates.

Good catch. The file itself contains the generic installation information, so there's not much there that is different about flex. That being said, yes, either reference to it should be...

the ABOUT-NLS file is included in the flex distribution. (For the record, it's the standard GNU gettext created file).

In terms of how to build flex: - From a flex source distribution (e.g. a file with a name like flex-2.6.4.tar.gz), the standard "./configure && make" will build flex for...