[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: lynx-dev Personal mail address
From: |
Klaus Weide |
Subject: |
Re: lynx-dev Personal mail address |
Date: |
Thu, 13 Jan 2000 10:23:13 -0600 (CST) |
On Wed, 12 Jan 2000, Doug Kaufman wrote:
> I had forgotten about that discussion. Thanks for the references.
> Since the default changed,I have been commenting out the
> NO_ANONYMOUS_EMAIL define in all the DOS binaries that I have
> distributed. I agree with changing the default back. Since this is one
> of the few items that can't be specified with the configure script,
> I would hope that someone on the list more familiar than I am with
> configure can put this in the configure script. We can then put the
> documentation of whatever configure option name is chosen (e.g.,
> "--anonymous-email") in the INSTALLATION file.
I don't think it should be given that much prominence. Frankly, I don't
think the NO_ANONYMOUS_EMAIL "feature" is fit to be used - and that in
spite of the fact that it *has* been the default for quite a while now.
If turned on -
1. Lynx behavior contradicts the documentation.
2. A non-functional (as far as the main purpose goes) option appears
on the Options Menu.
3. It doesn't apply to *all* the ways in which mail sending can be
initiated. This shows that, AFAICS & IMO, the NO_ANONYMOUS_EMAIL
feature was never well thought-out or comprehensive. It doesn't
guarantee what it seems to promise: that the personal_mail_address
entered by the user will never be used for mail. (It *may* achieve that
if lynx is configured / the user is restricted in such a way that
none of the ways of sending mail, other than "mailto:" handling, can
apply. But it isn't documented anywhere how exactly do do that, as
far as I know.)
4. It is misnamed - NO_PSEUDONYMOUS_EMAIL would be a better description
of what it does.
5. A basic underlying assumption made by this feature is: the system's
MTA is configured correctly and reliably knows the "real" and valid
address of the user; the user should be mistrusted. But in fact, these
days the opposite is often true: the user is interested to enter a valid
return address, but the system is misconfigured.
6. What *does* it do for non-VMS and non-UNIXlike systems? Can lynx even
be compiled with NO_ANONYMOUS_EMAIL?
Then there's this, in LYMail.c
#ifdef VMS
if (isPMDF || !body) {
.....
#ifdef NO_ANONYMOUS_MAIL
if (!isPMDF) {
fprintf(fd, "\n");
}
#endif /* NO_ANONYMOUS_MAIL */
}
Huh? there's an E missing, this seems to have been an undetected misspelling
since at least 2.7.1 days, what exactly is going on here? Someone on VMS
ougt to figure it out...
So, unless someone wants to clean all this up and turn it into a really
useful thing (in more than the most restricted guest situations) - let
it languish in userdefs.h at best.
If someone *really* uses NO_ANONYMOUS_EMAIL as a useful and necessary
option, in any sort of situation, let's hear it.
Klaus