Path: utzoo!attcan!uunet!cs.utexas.edu!samsung!uakari.primate.wisc.edu!uflorida!reef.cis.ufl.edu!bb
From: bb@reef.cis.ufl.edu (Brian Bartholomew)
Newsgroups: comp.sys.hp
Subject: Re: standards for windowing software?
Message-ID: <25121@uflorida.cis.ufl.EDU>
Date: 26 Oct 90 05:41:18 GMT
References: <1990Oct20.123109.7287@cs.dal.ca>
Sender: news@uflorida.cis.ufl.EDU
Distribution: na
Organization: UF CIS Dept.
Lines: 57


In article <1990Oct20.123109.7287@cs.dal.ca> you write:
>engineers, and they claim that by doing this HP is maintaining some kind
>of standard.  Software that adjusts to the actual window size is in
>their opinion "non-standard".
>
>Can anyone explain this to me?  Since vi and more get the screen size
>from termcap or terminfo when they load, what is nonstandard about
>getting the size of a window as well, which is after all a virtual
>terminal?  I can understand not adjusting to a resized window, but when
>you load a program it seems to me it should use the window you are in.

It's very simple - you are being screwed by HP.  The salesman is making
up ridiculous things, to get you to buy.  Imagine, a window system that
didn't let you window things!  Save your sanity, buy a Sun SPARC or a
NeXT.  There are far more applications available for a SPARC, anyway.

Actually, the technical excuse for this behavior is as follows:  On
Suns, when the window size changes, it is stored in the same area that
other terminal-related things like baud rate, kill characters, and
in/output processing are stored.  (The same data structure that stty
manipulates, i.e. the tty driver data area, i.e. the termio struct).
All applications may get to this information with a minimum of hassle -
and when the window size changes, a SIGWINCH signal is passed to all
applications using that window.  Apps that choose to pay attention to
window size may choose to define an action for that signal.  For
instance, the EMACS editor, even when running in a terminal (instead of
the more usual case of its own X Window), will adapt to the changed
size window.  Under SunOS, vi will not adapt to a window size change
while it is running - but it reads the window size upon startup, so
running it fresh in a size changed window works like you would want.
This behavior is not perfect, but useful, as it will let you run vi
with more than 24 lines displayed.

On the other hand, HP's have a more SysV idea of the termio struct.
The /etc/termcap file is considered unchanging system-wide information;
and, how would you modify one terminal type entry ("hpterm", say), to
reflect the sizes of all umpty-zillion windows on the screen?  I
believe that there are two environmental variables that termcap
applications respect - ROWS and COLUMNS.  However, they aren't used.
In fact, there are no provisions made for recording of the window size,
in anyplace that termcap applications can get to.

In conclusion, with termcap applications on HP 3xx/8xx computers, even
while running a modern, network-cooperative window system like X, you
are stuck with 80x24 windows.  Ridiculous, isn't it?


"Any sufficiently advanced technology is indistinguishable from a rigged demo."
-------------------------------------------------------------------------------
Brian Bartholomew	UUCP:       ...gatech!uflorida!matrix.math.ufl.edu!bb
University of Florida	Internet:   bb@matrix.math.ufl.edu
--
"Any sufficiently advanced technology is indistinguishable from a rigged demo."
-------------------------------------------------------------------------------
Brian Bartholomew	UUCP:       ...gatech!uflorida!matrix.math.ufl.edu!bb
University of Florida	Internet:   bb@matrix.math.ufl.edu