AW: TeX Live license issue

[Roland Hohler]

Dear Karl Berry,

thank you for the quick and detailed reply.

I've taken the software versions from TeX Live 2021. Also in the current Ghostscript version the license issue still exists. Therefore I'm going to create a bug ticket in Ghostscript Bugzilla.

Kind regards
Roland Hohler 

-----Ursprüngliche Nachricht-----
Von: Karl Berry <karl at freefriends.org> 
Gesendet: Samstag, 12. Februar 2022 00:29
An: Roland Hohler <Roland.Hohler at vitronic.com>
Cc: tex-live at tug.org
Betreff: Re: TeX Live license issue

    In ghostscript-9.53.3

FYI, the current version is 9.55.0.

    gdevmd2k.c gdevalps.c

1) Please report this to Ghostscript. We don't maintain gs.
https://bugs.ghostscript.com/

2) It appears that these are contributed printer drivers that are not normally compiled. Thus, unless you actually need support for the "ALPS MD" printer (guessing not :), you (and we) could/should just remove those files. They don't affect the license of anything else ("mere aggregation").

    In avantgar.r31835
    uagb8a.afm uagbi8a.afm uagr8a.afm uagri8a.afm

Oops. Ok, I updated those stale license statements to say the GPL, like the other afms from URW. In fact they long ago gave permission for their versions of the 35 basic PS fonts under both the GPL and LPPL, but not going to bother delving into that.

    May I assume that the package license is the valid one if nothing
    else is explicitly mentioned in e.g. a Readme file?

It's not my place to say what licensing assumptions you should or should not make.  TeX Live is a compilation of work by thousands of people over several decades. I'm just one of the volunteers who help put it together.

We do our best to ensure that everything in it is free software, but we're only human, and as you've seen, things slip through, especially from, e.g., 30 years ago when people weren't especially careful about such things. When we hear about problems, we do our best to address them.

So ... ultimately, although the package license indicated (at CTAN upload time, normally) is presumably what the author intended, it's up to you whether to take that as sufficient. Clearly it would be best/safest to write package authors when there is no explicit license statement.  --best, karl.