> On Sun, Mar 14, 2004 at 05:52:04AM +0100, Alex de Kruijff wrote:
> > On Thu, Mar 11, 2004 at 09:34:16PM -0600, Paul Seniura wrote:
> > >
> > > Hi Alex,
> > >
> > > > Dear Paul,
> > > >
> > > > On Tue, Mar 09, 2004 at 04:43:48PM -0600, Paul Seniura wrote:
> > > > > It seems NetBSD and OpenBSD continue to include
> > > > > telnet+telnetd+tn3270 together under one subdir as part of
> > > > > /src/usr.bin -- but FreeBSD moved only the telnet[d] pieces
> > > > > to /src/contrib/telnet and eliminated the tn3270 pieces completely.
> > > > >
> > > > > (I haven't dug too deep yet in the libtelnet tree, which is one
> > > > > piece that FreeBSD does retain as other BSDs have it.
> > > > > But for right now let's stick to the command & daemon parts.)
> > > > >
> > > > > I'm seriously debating in my head whether FreeBSD should add back
> > > > > the tn3270 pieces to /src/contrib/telnet so that we can match the
> > > > > other BSDs albeit in the 'contrib' subtree.
> > > >
> > > > I don't understand the word albeit in this line. English is not my
> > > > native language, sorry.
> > >
> > > http://dictionary.reference.com/search?q=albeit
> > > :-)
> >
> > I did look at my offline dictionary, but this didn't make sence. I just
> > stated this, so that I reacted in a strange way it would be clear why
> > this was the case.

I understand ;) .
It is a way to short-cut a lot of words.  In my case, it means: 
The other BSDs still have many things that we keep in /src/contrib
that should be in /src/usr.bin -- not just tn3270 was moved. 
But I'll follow the Party Line at least this far. ;)

OT
NetBSD has recently severed ties to FreeBSD, and I'm wondering if it's
partly due to FreeBSD becoming "too different". 
Just thinking out loud. ;)

> > > > Why do you want so much for all different groups to be alike?
> > > >
> > > > It seem to me that differences are natural when you have diffenent
> > > > groups. NetBSD and OpenBSD didn't have ports when FreeBSD started with
> > > > the use of ports. And this now is very succeful. We whould have missed
> > > > this if all we did was be like the others.
> > >
> > > Please look at the history of the BSDs.
> > > The tn3270 command was never in a 'port' to begin with.
> > > It was meant to be a companion to 'telnet' the command, the
> > > daemon, and its libraries.  They are intertwined.
> >
> > What I mean by this is: that i think having it as a port is a good thing
> > recardless of what others do.
> >
> > > "The use of ports" has not been successful w/r/t tn3270 itself.
> > > It STILL will not compile correctly, even today.  It was moved in
> > > order to allow 'world' to compile without problems.  It was moved
> > > *instead* of being fixed.
> >
> > Then moving it back will not fix it either. That its not being fixed has
> > nothing to do with being a port or not.

It is part of the problem as it exists right now. 
Not the _entire_ problem, _part_ of the problem.

When the telnet code is changed, no one will see that it
causes problems with tn3270, and no one fixes it. 
So it compounds the problems.

But the #1 problem right now is it being written in very old
and dated C language.  No one has touched it for 'that' long.

> > It has to do with to few people
> > who use the port. If it not fixed then that because no one with the
> > skill to do so is interesed in fixing it.

Someone could've patched tn3270.c in its proper place with
compiler statements "#ifdef false"/"#endif" surrounding the
entire module very easily, thus it becomes a 'good' compile
and will prevent contaminating 'world' until it is fixed. ;) 
When someone sees that the command has either disappeared
or Does Nothing, then the related PR could be cited.

> > > But the other BSDs have seemingly fixed it, and they left it
> > > inside /src/usr.bin/telnet where it belongs -- looking right now
> > > today at their CVS trees.  (Yes I will do 'diff' between theirs
> > > and ours.)
> >
> > You could become the port maintainer. ;-)

<sigh>
I mentioned it in the very first message of this thread. 
Can you see who the maintainer is listed in the Makefile for net/tn3270? 
(the patch for it came out a couple months ago already)
I _am_ the (new) maintainer!

But I do not have nor do I want 'committer' status. 
Whenever I get things working right, I'll open a PR and do
it that way. ;)

I hope it makes sense now, why I am complaining about how
tn3270 was treated during all these years? 
Esp. when other BSDs are seemingly not having our problems.

> > > > I personal feel that getting stuff out of the base system and in to the
> > > > port tree is a good thing if it is posible. This is more modulair and
> > > > thus is much more flexible. I see this as a good thing.
> > > >
> > > > Why should this ports be in the base system? Just because NetBSD or
> > > > OpenBSD do? FreeBSD is not OpenBSD or NetBSD
> > >
> > > I hope this is not too technical:
> > > All BSDs (except for this 'Free' one) presently have tn3270.c
> > > together with telnet.c for reasons that become apparent when
> > > studying how it works and how it is compiled etc.  As they are
> > > presently written, all of telnet[d].c and tn3270.c must use the
> > > same macros and other source files from the same 'level' subdir.
> > > That means if something changes for the sake of 'telnet', it had
> > > better work with 'tn3270' also.
>
> Yes, this is too technical. I would have to study the file, which i'm
> not going to. If you say that it couldn't posibly be a port, like perl5
> can, then i will take your word for this.
>
> > > Putting tn3270 over into a port is a 'Free'BSD-only KLUDGE: look
> > > at its Makefile under /src/ports/net/tn3270.  It was moved from
> > > where it belongs, instead of fixing it to compile and work
> > > properly per current specs -- and today it STILL will not compile
> > > correctly.  Moving it only acted to permit the rest of the base
> > > system ('world') to compile without problems.  Hence I call it a
> > > 'kludge' in its present 'Free'BSD-only form.
> >
> > But it is posible to have it as a port and working. It just needs some
> > one to put in the time to fix the port as it is. You wrongly put the
> > cause in it being a port now.

If it stayed where it belongs, it could've been patched as I
mention above to help 'world' continue building.  It did
not need to be moved.  Its problems would've been more
'visible'.  The telnet stack has had changes that break
tn3270 now.  This is on top of the problems showing at bento. 
When its C-language bugs are fixed, and it compiles
properly, I expect tn3270 to still cause a coredump
(re: an old PR still open) -- something bento doesn't test for.

> > > I was not 'here' back in 1999 when this decision was made. (See
> > > my reply to Kris, too, please; I show the 'commits' there.)  In
> > > 1999, we were using OS/2 which had a fully functional basic
> > > PCom/3270 provided with the o.s. for 'free'. ;)  Now I am trying
> > > to show TPTB how 'free' o.s.+software can be used, and ran into
> > > this stupid kludge almost 5 years too late. :(
> > >
> > > By your logic, let's move all of /src/contrib to the appropriate
> > > subdirs under /src/ports and not have a built-in telnet or any
> > > other such command! ;)
> >
> > At least consider it.
> >
> > Well perl use to be part of the base system and that was moved in
> > FreeBSD 5. Perl works because it was moved and there was efford put in
> > so that the port doesn't work.

Perl doesn't directly copy files from the /src tree in order for
it to be compiled. 
I mean .c files -- things that should be in a library function
if such code is to be shared like that.
The way it is now, tn3270 is an end-user app that copies /src
files over!  This is a real mess...

 > Alex

  --  thx, Paul Seniura.

_______________________________________________
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"

Reply via email to