On Sun, 14 Mar 2021 21:19:06 +0000
Stuart Henderson <s...@spacehopper.org>:
> On 2021/03/14 22:05, Florian Viehweger wrote:
> > Hi ports,
> >
> > attached is a port of Vger, a gemini server supporting chroot,
> > virtualhosts, CGI, default language choice, redirections and MIME types
> > detection.
> >
> > Initial work was done by prx, with permission I'm taking maintainership.
> >
> > I've created a dedicated user and group '_gemini_user' both with ID 865.
>
> users are normally named after the port i.e. this would be better using
> _vger than _gemini_user
>
> > 'make test' works on amd64.
> >
> > 'portcheck -N' complains about
> > "hardcoded paths detected in pkg/README, consider using SUBST_VARS and
> > TRUEPREFIX/LOCALBASE/LOCALSTATEDIR/VARBASE"
> > but this is for an example inetd.conf file.
> >
> > Comments? OK?
> >
> > Thanks!
> >
> > --
> > greetings,
> >
> > Florian Viehweger
>
> inetd, really, in 2021?!
>
> 127.0.0.1:11965 stream tcp nowait _gemini_user /usr/local/bin/vger vger
>
> - this should use ${TRUEPREFIX} not hardcoded /usr/local.
>
> README should follow the formatting used in
> ports/infrastructure/README.template and every other pkg-readme file.
>
I'm vger author. What's wrong with inetd? In this case it acts
behind a program that will manage TLS connection and pass it to
local inetd. Limit rate can be done on the server behind to
prevent inetd to spawn too many processes (that's the only issue
I'm aware of)
