Re: [Nmh-workers] things still to fix for this release?

[pmaydell]

Harald Geyer wrote:
>>  > You didn't include it in your list, but we perhaps ought to consider bug
>>  > 9228:
>>  > https://savannah.nongnu.org/bugs/index.php?func=detailitem&item_id=9228
>>  > 
>>  > Anyone know the history of $mheditor? Can we just remove the reference
>>  > in the man page?
>> 
>> i only checked 1.1, but mheditor sure looks like it's honored to me.
>> i think the bug report is bogus.
>
>I once forwarded this bug from the Debian BTS. I believe I did some
>tests then and came to the same conclusion as the original submitter.
>However I agree, that I can't reproduce this anymore. It seems to work
>as expected in the CVS version and the Debian version.

I just wrote this in the savannah BTS:

I think the idea is that environment variables in uppercase are
user-settable, and lowercase ones are set only to communicate information
from one part of nmh to another. The explanatory text generally says 'this
is set [by dist/comp/etc]' rather than 'you can specify'.

It's clear from the code that whatnow does check the envariable; I assume
the original reporter found setting it had no effect because it was
overridden by whatever other bit of nmh was invoking whatnow.

My suggestion is that we should add something to the mh-profile manpage,
explaining this convention and that lowercase envariables aren't
user-settable. We could probably also put something more of a separator
between them and the user-settable envariables.

Or am I completely off-beam?

-- PMM