Re: [Nmh-workers] sortm and folder -pack preserve sequence membership

[Ken Hornstein]

>>I'm neutral on that; it seems obvious to me that the sequences would be
>>preserved;
>
>I don't know why, but the opposite was obvious to me. Perhaps for
>historical reasons.

Fair enough.  I don't know the history ... would sortm completely trash
sequences back in the day?  It would seem to me that you'd either need
to preserve them or zero them out when doing a sortm, otherwise they'd
point to the wrong messages.

The basic design of the code, at least, made it pretty easy to ensure
that they would always be preserved.  I don't have the complete MH history,
but I see sortm preserved sequences even back in MH-5 ... which was released
in 1985.  Well, let me rephrase that ... there was code that clearly was
intended to do that.  Assuming there were no bugs of course.

>> I'd only expect a notice if they were not (although, I'm not sure what
>>burst would do that it would change sequence membership).  If someone
>>wants to to propose a change, that's fine with me.
>
>Until a few weeks ago, I had been assuming that they were not.

Fair enough; might be a more appropriate entry in mh-sequences(5), though.

>>As we've seen, most of those man pages originate from your era, Norm.
>>How come you guys never mentioned that? :-)
>
>Yes, the man pages for sortm, folder and (I think) burst, originated at
>Rand, but sequences originated at U.C Irvine, after Rand released MH
>into the public domain. It's true that we had a lot of input into the
>process, but almost all the actual work was was done at Irvine.

Huh, had no idea.  Thanks for the info!

--Ken