lynx-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: lynx-dev lynx2-8-1/Linux patch for new VTs


From: Klaus Weide
Subject: Re: lynx-dev lynx2-8-1/Linux patch for new VTs
Date: Sun, 28 Mar 1999 12:22:29 -0600 (CST)

On Sun, 28 Mar 1999, Philip Webb wrote:

> 990328 Tony Garnock-Jones wrote:
> > I miss the capability to open a new browser window to explore sublinks.
> > As I run Lynx mostly on console on a Linux box,
> > I decided to play with implementing a "open new browser window" feature
> > using the Linux VT_??? ioctl()s.
> 
> as JS pointed out, one basic principle of Lynx development is
> to keep it fully compatible across all platforms,
> so if you want your patch accepted,
> it will have to work on other UNIX's, VMS, Win95 etc.

Don't elevate this to an absolute requirement.  Lynx is not "fully
compatible across all platforms", in the sense that a feature that works
on one platform works on all.  I suspect it has never been since it
started supporting more than one platform.

LYNXCGI works only on Unix.  NSL_FORK works only on Unix.  There's
most likely more like that.  DEFAULT_VIRTUAL_MEMORY_SIZE works only on
VMS.  Some things (still many?) don't work under DOS.  And
EXP_CHARTRANS_AUTOSWITCH works only for the Linux console.

Many of these features are not bound in principle to one platform.  The
functionality could be ported to other platforms, or a (sometimes inferior)
equivalent provided, if there were someone interested (and knowlegeable)
enough to do it.  The existing implementation (for platforms that have it)
can be used as hook or template for new implementation (for platforms that
don't yet have it).

It does make no sense, IMO, to require a new contributor to implement
his stuff for all possible platforms.  If the stuff is useful for
those who can use it (a different question), but the implementation is
obvious only for some platforms, and there are good reasons why it
should be implemented *in* lynx (which is not a given in *this* case),
and it doesn't interfere with other platforms (#ifdef's etc.), then I
see no reason to exclude it.

> > There's the possibility there's already a way of doing something similar
> > without having to resort to the VT ioctls
> > (which are heavily festooned with warnings in the manpages).
> 
> well, on a UNIX system -- incl Linux -- you can easily run Screen,
> which gives you a lot of other capabilities:
> if you don't know it, goto  www.math.fu-berlin.de/~guckes/screen .
>
> > <shrug>. It works for me :-)
> 
> great, but that's not going to be of interest to anyone else ... (smile).

It's of potential interest to everyone who finds this functionality useful.
Implementing it for e.g. linux may lead to implemetation for more OSs.

   Klaus

reply via email to

[Prev in Thread] Current Thread [Next in Thread]