less5 - 2 fixes and an enhancement...

[Mark Nudelman]

In article <4300004 at hpcvca.HP.COM>, charles at hpcvca.HP.COM (Charles Brown) writes:
> At least some of these fixes have been sent to the author several
> times over the past 3 years.  It WOULD be appropriate for the author to
> implement the fixes.  So far he has not.
> 
> In particular the incorrect behaviour of less with HP terminals has
> been reported several times.  Because this has to be patched each time
> the sources come out, I stopped upgrading less at this site back in
> 1986 with version 61.
> 
> To the author of less I say "Thank you for writing it".  But I am a
> little frustrated that these fixes are repeatedly ignored.  When I
> sent the same set of fixes (independently created 2 years ago) my mail
> was not even agknowledged.
> --
> 	Charles Brown		charles%hpcvca at hplabs.hp.com
> 	Not representing my employer.

As the author of "less", I have received literally hundreds of mail 
items concerning the program over the past few years.  I've gotten 
everything from glowing praise to nasty-grams because someone's 
favorite feature hasn't been included.  While I can understand your 
frustration, I'd like to make a few points.  

A large fraction of the mail I've received has included suggested code 
changes.  Evaluating and integrating these changes is a time-consuming, 
altho sometimes enjoyable, task.  I have, despite the large volume of 
mail, attempted to answer every note I've received.  If I didn't 
respond to your mail, I almost certainly never received it.  I'm
surprised that you took a lack of response as a lack of interest,
rather than the more obvious lack of receipt.

Second, concerning the specific problem you mention with HP terminals, 
I do not have access to an HP terminal.  I have to evaluate the changes 
concerning such terminals based on what other people tell me.  When I 
posted my very first version of less, over three years ago, I got 
various complaints about it not working on different kinds of 
terminals.  I also got several (incompatible) suggestions on how to fix 
it for each particular type of terminal.  In regard to the problem with 
terminals with off-screen memory, I was told that the termcap "ti" 
string should set up the terminal so that the on-screen memory was the 
only accessible memory, in effect making these terminals appear like 
"normal" terminals; and that people who complained had a broken termcap 
description.  I nevertheless tried to integrate the fixes, without a 
way to test the result.  After my first attempt at a fix, I got 
complaints that less no longer worked on some terminals; so I tried 
again.  Now you say it still doesn't work on some terminal.  Frankly, 
I'm not surprised.  But I'm a bit at a loss as to how to proceed.  Can 
you guarantee that your change will make it work on every terminal, and 
won't break anything? If so, then I'll gladly incorporate your change. 
If not, I'll probably incorporate it anyway; I've tried to be very 
generous in acceptance of new code, which is one of the reasons less is 
getting as large and baroque as it is.  

I am, by the way, evaluating and integrating the recent postings on the 
net.  The ones which are clearly bug fixes will of course go in.  My
current thinking is that the decrypting/unpacking stuff just doesn't
belong there; pipes suffice.  The others I am still evaluating.

Finally, you're welcome.

Mark Nudelman 
{sun,decwrl,hplabs}!pyramid!ctnews!UNIX386!markn