GaryScott wrote:
> On May 14, 9:47 am, nos...@[EMAIL PROTECTED]
(Richard Maine) wrote:
....
>> The Moving Finger writes; and, having writ,
>> Moves on: nor all your Piety nor Wit
>> Shall lure it back to cancel half a Line,
>> Nor all your Tears wash out a Word of it
>> -- Omar Khayyam
>>
>> Even the "half a line" bit seems perfect.
>
> Unfortunate. Even if the partial line had been flushed, I can't think
> of any output device that can't achieve the desired result. If there
> were one, it would be a rare case. One of my main desires for
> advance=no is to be able to overlay as CVF originally sup****ted,
> although I've converted nearly everything to GUIs now, so this type of
> processing is less im****tant for me.
I can think of several devices where you can't go back again.
Terninals that physically print would have been a common one
back when the "left tab limit" rule was written. There are,
however, terminals that can't rewrite already displayed data
even today. And there are terminals (and emulations thereof)
that can't do it without lots of very obscure embedded directives
that the Fortran I/O sup****t library can hardly be expected to
know.
Further, there are "devices" which actually consist of other
processes. Redirection on plain-vanilla unix can do that.
The data already consumed (and acted upon) by that external
process can hardly be recalled without expecting the application
programmer who wrote that process to provide all kinds of
additional capabilities to his (her) code.
If you are on a file which can be BACKSPACEd, that's how you
should rewrite data that was processed by previous I/O statements.
Some files can't be backspaced. Coincidentally, they're the same
ones that would be so problematical about tabbng backwards
across them.
--
J. Giles
"I conclude that there are two ways of constructing a software
design: One way is to make it so simple that there are obviously
no deficiencies and the other way is to make it so complicated
that there are no obvious deficiencies." -- C. A. R. Hoare
|
