Re: status of tn5250?

On Wed, 23 Aug 2017, Terrence Enger wrote:

David Gibb's question about the need for this mailing list reminded me
that I would really like to connect from Linux to an IBM/i.  Segfaults
in tn5250 have stopped me from doing this for a long time, perhaps
since I dropped 32-bit support during an O/S upgrade.

I've been using tn5250 for years on 64 bit without problems. I have a few changes that I made a long time ago and made a tarball with them, available here:

http://www.chowhouse.com/~jimi/tn5250-0.17.5.tar.gz

Digging in tn5250-0.17.4 showed that a couple of functions were
implicitly declared, thus declared to return an int.  This truncates
the return value to 32 bits, which is a bad thing for a pointer in a
64-bit environment.  I have fixed all implicit declarations, and now
tn5250 (I have not tested the other programs) works so far as I have
used it.  So a couple questions ...

(*) Is my patch of interest to anybody?  Looking on sourceforge
    <https://sourceforge.net/p/tn5250/patches/>, I see that an open
    patch has been unchanged since 2013.  Not, mind you, that I can
    see how to submit a patch.

I haven't used SourceForge for a very long time and honestly I don't remember the status of the code there. I don't think I have the ability to post anything there anymore. Send your patch to this mailing list in unified diff format and let's take a look at it.

(*) I did all my changes with only language-level knowledge, without
    understanding what the code is actually doing.  So, a code review
    and testing of all the applications would be in order.  Who wants
    to do a review?

(*) `make check` found nothing to do.  Am I doing something wrong?

I don't think 'make check' does anything normally.

James Rich