Hi Ken,
mhshow: Can't convert unicode-1-1-utf-7 to UTF-8
mhshow: unable to convert character set of part 1 to unicode-1-1-utf-7,
continuing...
I thought that charset was mega-bogus, but it turns out that's wrong.
It's defined in RFC 2152. We simply pass those strings directly to
The good folks in charge of PHPMailer fixed the issue I reported with
an invalid CTE for multipart MIME parts (and rather promptly!):
Congratulations, Ken. Quite a big source of email sent too.
One down... https://xkcd.com/386/ :-)
Cheers, Ralph.
Hi Ken,
The extra semicolon ... okay, Eric makes a good point in that why are
we complaining about that? There's no confusion there on what to do.
Perhaps rather than the generator just being confused being separators
and terminators, there's a missing parameter after the final semicolon.
Hi epg,
But I'll be honest; I have never in my life seen a message like
that.
Looks like these are mostly from mailchimp.
From what I hear from some folks that share an office with them, they're
pretty nice guys; customersupp...@mailchimp.com would probably garner a
good response. If
Hi David,
Another thought: it'd be nice if the test suite could optionally use
ISO-8859-1 instead of UTF-8. Even better, the tests that care could
do both. That would have caught this problem.
Rather than have the test author decide if it should do both, and be
prone to human error, could
Hi Paul,
here's the place where the new containsnul is used, for CT_TEXT and
CT_APPLICATION.
...
case CT_TEXT:
if (contains8bit !containsnul !linelen !linespace
!checksw)
ct-c_encoding = CE_8BIT;
else if (contains8bit || containsnul ||
Hi Paul and Ken,
Paul wrote:
but if we don't, should the conversion use 1024 (2.4M) or 1000
(2.6M) as the divisor? and of course that leads to MB vs. MiB.
and how many digits should there be after the decimal? sigh.
If it's base 2 then it must be Ki rather than K, etc. I don't think the
B
Hi Ken,
FWIW, I looked at mutt and it's fallback if the locale is US-ASCII but
the characters are 8bit is to output unknown-8bit as the charset.
Seems like a reasonable idea to me.
Isn't that just punting the problem downstream, forcing others to
investigate unknown-8bit, probably with
Hi Ken,
The question I put to all of you: who gets to decide what is right?
The RFCs. They define interoperability. :-)
I agree with kre; from looking at FLOSS MUAs and mail-producing
libraries, one often gets the impression that they didn't open an RFC,
or if they did, they backed off at
Hi Paul,
it's clearly (to me, at least) not too late to change units to
kilo. kibi can be added separately.
Ken said the code used base 2, 10, so the shipped files using the new
`units' should switch to kibi? Or perhaps there aren't any? (I thought
the functions were being added so marker
Hi Paul,
i used the most convenient method of forwarding an entire mail message
that nmh provides -- the method that i've been encouraged to switch to
over writing build directives
Are you aware of `forw -mime'?
Cheers, Ralph.
___
Nmh-workers
Hi Ken,
Or if you're not comfortable
with that, a make clean make make install will also work.
BTW, make(1) accepts a list of targets, e.g. `make clean all install',
`all' being the conventional name for the first, default, target. It
only builds `all' if `clean' succeeds, and so on, unless
Hi Ken,
I know what you mean ... it seems like we need a new tool. mhshow
really seems to be more about display this content, mhstore is more
about store this content. We don't really have something that's in
the middle: direct this content to stdout.
mhstore(1) reads stdin with `-file -'
Hi Paul,
mhlist still uses 1024-based math, but prints K/M/G. barring a more
complete rewrite to make mhlist's output tuneable, i guess i should
change it to either print Ki/Mi/Gi or use 1000-based math, to at least
be self-consistent. thoughts?
Changing the suffix has the least impact on
Hi Paul,
Sorry for the delay in replying, moving house...
case CT_TEXT:
if (contains8bit !containsnul !linelen !linespace
!checksw)
ct-c_encoding = CE_8BIT;
else if (contains8bit || containsnul || linelen || linespace
||
Hi Ken,
Earl wrote:
I am unclear as to the right solution. Obviously adjusting the MIME
parsing routines is wrong. Making sure that the last character of
text/plain parts on output contains a newline might be better, but
I'm not sure that's technically right either.
One could take the
Hi Ken,
The question I put to all of you: who gets to decide what is
right?
The RFCs. They define interoperability. :-)
Sure, but that's not really the problem we're talking about here.
...
Eric Gillespie said last month:
Is nmh primarily trying to help users file bugs in other
Hi Norm,
That puts multiple addresses on the same line, delimited by commas.
But commas can also occur in comments, so I would have to go to some
trouble to parse the output. Is there a variant that would put each
address on on a separate line? I tried sprinkling '\n's in your
format
abandoned draft behind, I can't figure our where. Certainly not
in my Draft-folder.
A cut-down manual version gives
$ repl -cc all -query -editor false
Reply to Norman Shapiro n...@dad.org? n
Reply to Ralph Corderoy ra...@inputplus.co.uk? n
Reply to nmh-workers@nongnu.org? n
Reply
Hi Norm,
$ repl -cc all -query -editor false
Reply to Norman Shapiro n...@dad.org? n
Reply to Ralph Corderoy ra...@inputplus.co.uk? n
Reply to nmh-workers@nongnu.org? n
Reply to n...@dad.org? n
Reply to ralph? n
whatnow: problems with edit--draft left in /home/ralph/mail/drafts/,41
Hi Paul,
The command would operate on one or more messages. Its arguments
would be like the arguments of almost all the nmh commands that
operate on one more more messages, with the default being cur. It
would output, to its stdout, all the addressees, including comments,
of all the
Hi David,
Jón wrote:
Background: I’m using mhstore in a script to automatically extract
images and text from messages. Since the messages come from and
results go to naïve users, I’d like to be able to use the original
file basenames so that they recognise what they sent.
...
For
Hi,
In producing Norm's test HTML I found this ancient mhbuild duplicates
charset when it's us-ascii.
$ for f in foo iso-8869-1 utf-8 us-ascii; do
printf '\n#text/plain; charset=%s /dev/null\n' $f |
mhbuild -;
done
MIME-Version: 1.0
Content-Type: text/plain;
Hi Norm,
Would you, please send me an Email containing an html part, short
enough so that I can take a screen shot of the mhshow output, and send
it to you, in order to confirm that mhshow is working for me the way
you think it should, but long enough to be non-trivial. My monitor is
Hi,
Paul wrote:
i'm not convinced that piping yes n to a clearly interactive command
turns it into a scriptable interface. it's more akin to screen
scraping. there's no particular reason for Reply to address? to
change any time soon, but also no reason for us to promise that it
won't. if
Hi Ken,
Ken, by sanitizing, do you mean not using the filenames that
mhstore considers insecure? Or, the basename that Jón wants? I
thought it was the latter, in which case I agree that it should be
handled by the user.
I thought Jón wanted both. I'm kinda torn on the first item; if
Hi David,
Norm wrote:
How do I get mhshow to use my current browser for html parts?
...
If you want to use a graphical browser, this should work:
mhshow: -noconcat
mhshow-show-text/html: firefox %F
Can't recall if it's been discussed before, but feeding HTML into a text
browser
Hi Norm,
I note that iconv(1) at the command line gives an offset. That might
help you track it down.
$ printf 'foo\xe2\xe1\xf2\n' | iconv -f utf-8
fooiconv: illegal input sequence at position 3
$
Cheers, Ralph.
___
Nmh-workers mailing
Hi Norm,
I don't know how to change locales. man -k locale does not suggest
anything. And if I did, I would not know what locale to change to:
'locale --all-locales' gives 735 options. 'locale --charmaps' gives
230 options.
As Ken said, the default for modern Linux distributions is UTF-8
Hi Paul,
Ken wrote:
I had to experiment a little bit for this; it turns out a profile
entry in specified in MHSHOW overrides things in mhn.defaults, but
NOT things in your .mh_profile.
So the order is .mh_profile, $MHSHOW, mhn.defaults.
this paragraph is cumbersome -- perhaps someone
Hi David,
Just curious, is LANG set to that as well? It is for me on Linux and
I don't explicitly set it (that I know of).
You can walk up /proc to see where it may be set or changed, e.g.
$ sudo true; p=$$; while :; do
cd /proc/$p pwd
sudo sh -c 'tr \\0 \\n environ'
Hi Ken,
Paul wrote:
the fact that the $MHSHOW/$MHBUILD etc variables come in between
makes no sense -- one would usually view an environment variable as
a high-priority override, but not in this case.
I think some of this might be simple oversight.
Could it be more that .mh_profile has
Hi Paul,
The first occurrence of a component is used,
e.g.\ .mh_profile's trumps $MHSHOW's.
...
is '\' a hard space?
No, it's a non-printing, zero-width character. See 4.1 in Kernighan's
http://troff.org/54.pdf. troff rightly believes in a bigger space at
the end of a sentence
Hi David,
- I'll echo Norm's request for better repl to MIME messages. Maybe
replyfilter would do the job. But I tried once and couldn't use it
due to Perl madness, and haven't gone in to figure it out. Also, I'm
little wary of depending on par in a UTF-8 world: the patch to par to
deal
Hi Alexander,
e.g. all manpages to go in as whatever.1mh instead of whatever.1
OOI, why's that? And is that a recentish change on Debian as this old
Ubuntu doesn't do that. Does that mean co(1) is co.1rcs, etc. too?
Cheers, Ralph.
___
Nmh-workers
Hi David,
I am not sure about that shell syntax; someone else will have to confirm
if it's portable enough.
! isn't portable. Neither is egrep -q.
What kind of portability are we aiming at? ! is in POSIX;
Hi Lyndon,
I agree with David's comments about Heirloom shell and posh.
Just redirect stdout to /dev/null and you can avoid using either of -s
or -q.
In most cases. Sometimes it's used to wait until a match is seen on an
infinite stream and then quit. As if `sed /foo/q'. ;-)
Cheers,
Hi,
On this old nmh, if cur is 42 and I `show 314' then running a command
from within less, e.g. `scan .', has 314 be cur. Good. Do similar with
mhshow and 42 is still cur. On exiting mhshow, 314 is cur.
I think the inconsistency trains one to err, e.g. `!rmm', and cur should
be set early on,
Hi Ken,
int
main(int argc, char *argv[])
{
if (argc != 2) {
fprintf(stderr, Usage: %s locale-name\n, argv[0]);
exit(2);
}
Cheers, Ralph.
___
Nmh-workers mailing list
Nmh-workers@nongnu.org
Hi David,
I think the only true solution here is to create a test program.
I noticed that the return value of setlocale() isn't checked by nmh
programs. Should it be? I think so.
Me too.
I got there by wondering if we could use mhparam as the test program.
I was also digging for a
Hi Norm,
So you are saying that normal unix commands, such as grep, wc, tr
etc, do or someday the GNU versions will, know about UTF-8, at least
for file contents,
Yes, they do, today. And have done for quite a while. You need your
environment variables set up properly so `locale' reports
Hi Jererad,
Putting the specified messages into the specified order is trivial,
but where do they go in the overall order of the folder?
As Norm's clarified said, things are just being moved within the same
message numbers.
If they are to be inserted into one another's slots,
I see no sane
Hello Ken,
The Unix kernel stores filenames as a run of bytes, not including
`/' and NUL.
That's not universally true anymore. Some newer filesystems are
mandating that filenames are UTF-8 and enforcing normalization rules
(MacOS X and Solaris are two notable examples).
Thanks, I didn't
Hi Norm,
Here's where each has to end up.
$ paste (echo $l) (echo $l | sort -n) | awk '$1 != $2'
25 2
4 3
3 4
31 9
29 10
41 13
9 20
13 23
2 25
23 29
20 31
Hi Norm,
This request is for sortm to have a '-program path-name' feature
path-name would name a command which would be invoked by sortm. It
would be given 2 arguments, full path names of messages. Its standard
output would either
begin with +, meaning that the first argument is to be
Hi Ken,
By implication, the comparison program is buggy if that doesn't
hold? sortm(1) punts to qsort(3) for the hard graft and that
demands consistency; I think I'd like sortm to protect me from a
buggy comparison program. [...]
You know what? Forget I said anything :-)
Sorry,
Hi Ken,
We won't find standards support for it: 465 for SMTPS was revoked
by IANA (in 1998!) and reassigned to another service. But as long
as ATT (I know), Verizon (I think), and whoever else require it,
we're stuck with it.
No we're not. Nor should we support it. The
Hi Ken,
I am unsure if there is a standards-base API that lets us detect
invalid characters (I'm not interested in something like Recode for
this). I wonder if mbtowc() and friends would throw errors if they
encounter an invalid character in the current locale.
The mbtowc(3posix) here says
Hi Earl,
I have the habit of always converting HTML emails I received into
plain text when I reply
Me too! :-)
I did not know Pythogorous wrote in English. If in ASCII, it would
be:
Ignoring the malformed octal... ;-)
$ egrep '^( +[0-9]{3})+$' ~/mail/inbox/11661 |
tr -s ' '
Hi,
Norm wrote:
I am not at all secure about how the standard GNU utilities will
handle non-ascii characters. For example, 'wc -c', just counts
bytes.
Christian has pointed out -c has remained bytes, --bytes is a synonym,
because otherwise too many things would break, and that -m has been
Hi Jerrad,
It understands more structured information as well, but seemingly only
in American style dates and without support for 24-hour times e.g;
% google calendar add 12/31 at 10:30-12:00AM Partay!
Does the underlying API used by this command-line program support ICS
file import as I
Hi Ken,
Right now a call to the MIME parsing routines end up slurping in the
whole message, but that's not desirable for a lot of programs (scan,
pick). It seems like parsing all of the messages headers is generally
worthwhile; that (usually) fits within a single stdio buffer, so doing
Hi Ken,
If we're having lazy evaluation of MIME parts, which is good, can it
also cover the headers? `pick --list-id f...@bar.com' isn't
concerned with decoding Subject and all those Received headers. It
may not sound like much, but we have folders with tens of thousands
of emails.
Hi Ken,
Well, I guess we could make it work both ways. Right now it's not
really decoded before it hits the format engine. We could keep with
that logic. Or if you wanted to convert it to ASCII ... well, I don't
see a better option than converting it and substituting an appropriate
Hi Ken,
I suppose in theory we should replace that with specific tests rather
than using isspace()
Might be easier to tie it in with the RFCs if it used macro names
borrowed from it.
we don't handle the case where a locale says 0x00-0x7F does not match
ASCII (are there such locales?).
Are
Hi Earl,
Would be much nicer if it travelled internally as UTF-8 and then
popped into the US-ASCII draft as
Subject: Re: =3D?utf-8?B?wqE=3D?=3DHola world!
and into the UTF-8 draft as Re: =C2=A1Hola world!.
But this behavior does not require the data to be stored in UTF-8.
No,
, no subject; hope that's ok
$ fgrep From: /var/spool/mail/ralph
From: ra...@inputplus.co.uk (Ralph Corderoy)
$
I seem to vaguely recall being a bit surprised when seeing the `Foo Bar
f...@bar.com' instead and wondering how legal it was. ;-) So there
must have been a transition
Ken wrote:
% scan -width -format '%{x-spam-status}' messages
Aside, it would be nice if scan's -width grew an infinite value as
is a little kludgy. I was thinking `0', but that's not an error at
the moment, it seems to mean `The default is to use the width of the
terminal' --
Hi Ken,
Paul wrote:
Ken wrote:
And it continues to this day.
$ readlink -e `which mail`
/usr/bin/heirloom-mailx
$ mail ralph $N
No message, no subject; hope that's ok
$ fgrep From: /var/spool/mail/ralph
From: ra...@inputplus.co.uk (Ralph Corderoy
Hi Ken,
ATT Message Center Version 1 (Apr 11 2006)
ELM [version 2.4 PL22]
Mailer::1.0 [Perl?]
slrn/0.9.5.7 (UNIX) [Usenet, I know]
[Some MTA error reports.]
XFMail 1.3 [p0] on Linux
So what percentage is that out of all the email you have? Is there a
tapering
Hi Lyndon,
We were optimizing for response time across hundreds of transactions
per second. I can't imagine how anyone using MH from the command line
would ever notice this sort of tune-up.
Hundreds? My tens of thousands of emails are sitting in RAM, and I'm
waiting a few seconds for pick.
Hi Jón,
I’m with Norm on the general principle.
Not to take away from Norm's and your point, but I often find GNU grep's
-o option handy as a tokeniser.
$ echo 'foo+bar9/3.14*.42' |
grep -Eo '[a-z][a-z0-9_]*|[0-9]*\.[0-9]+|[0-9]+|.'
foo
+
bar9
/
3.14
*
.42
Hi David,
Ken's pointed out the meta wasn't in the original message; were you
seeing that in Chrome's view of the source? It might add it based on
the default it chose. Ken said the HTML default was Windows-1252; it
seems surprising Windows was ever a default. I thought it was
ISO-8859-1 up
Hi Norm,
How is it decided that text/html is a better format than text/plain?
The RFCs that define MIME say that alternatives parts representing the
same content, e.g. text/plain, text/rtf, and text/html, shall be in the
email in order worst to best when judged by their ability to represent
the
Hi David,
Something else added it along the way. As Ken noted later,
it's not in the message in the raw mailing list archive.
Yes, I think that's what I said. :-) I was wondering if that something
was Chrome.
Ken said the HTML default was Windows-1252; it seems surprising
Windows was
Hi Jerrad,
This allows a user reading the email in an MUA that's not
MIME-capabable to come across the simplest format first, e.g.
text/plain.
That's the theory, but many senders use a text/plain part akin to If
you cannot read this, get a better MUA
Agreed, but I was trying to explain
as Anthony pointed out, the default is really ISO-8859-1, but everyone
treats that as windows-1252 (until we get to HTML5, of course).
Everyone bar at least links(1). :-)
-html-assume-codepage codepage
Use the given codepage when the webpage did not specify its
codepage.
Hi David,
2) Optionally, the user could move to the new behavior of not
counting the trailing newline.
Like Python's `from future import sanity'.
A problem with ~/.mh_profile calling for sanity is that it might be
wanted generally but the user can't be sure no scripts they use want
Hi Anthony,
$ for w in {0..9}; do
echo $w:`scan -width $w -format '%(decode{subject})' .`
done
0:Po£nds.
1:
2:P
3:Po
4:Po�
5:Po£
6:Po£n
7:Po£nd
8:Po£nds
9:Po£nds.
$
Perhaps a new -runes that counts
Hi David,
OK, so we're normally after columnar output. How about `-columns
C', where the string is output, as measured by wcwidth(), until
either the string ends or C would be exceeded.
How about -outsize, to be consistent with fmttest(1):
The -outsize switch controls the maximum
Hi David,
$ for w in {0..9}; do
echo $w:`scan -width $w -format '%(decode{subject})' .`
done
0:Po£nds.
1:
2:P
3:Po
4:Po£
5:Po£n
6:Po£nd
7:Po£nds
8:Po£nds.
9:Po£nds.
$ scan -version
scan -- nmh-1.6+dev [compiled on atto.localdomain at Sat Jul 26 07:57:00 CDT
2014]
Tsk! I
Hi David,
The functionality sounds OK, though it needs to know `Ｆoo' is
2+1+1; there's no mention of wider-than-one runes above AFAICS.
$ fmttest -outsize 2 -raw -format '%{text}' 'Ｆo'
produces no output but
$ fmttest -outsize 3 -raw -format '%{text}' 'Ｆo | od -x
000
Hi David,
I'm looking now at replacing the use of fixed buffers with dynamic
allocation in the format engine so that we could interpret -width 0
(or -width max?) as infinite.
`-width ∞', surely. ;-)
Cheers, Ralph.
___
Nmh-workers mailing list
Hi Ken,
I know there are few nmh programmers that haven't coded assember,
but -outsize is something only a programmer could like. :-)
Sigh. fmttest(1) really is designed to be able to exercise the
details of the format engine.
Oh, I think it's fine for that, just not suitable to be
Hi David,
but when I look inside the message, the first alternative is
text/plain and the second is text/html - mhlist is lying to me.
mhlist isn't lying. You define the order of alternatives differently
than it does. You don't have to like it, but it's not lying.
I'd argue mhlist's
Hi Norm,
Can anybody suggest a course of study so that I can learn enough about
MIME to able to participate, meaningfully, in most of these
discussions?
The link to Jerry's book given earlier is good. There's also
http://www.hunnysoft.com/mime/mime-guide.html as a high-level overview.
Hi Ken,
So ... what would break if the newline _wasn't_ counted? I am leaning
towards going the wcwidth() route and simply not counting things like
that as printable. I will note that isprint('\n') returns 0.
There's been many detours along the way so I could be confused, but I
think
Hi Ken,
I am not the expert here ... but it looks like to me (after some
experimentation with exactly 1 data point, an xterm), if you print a
character in the last column but follow it with a newline, you don't
get two newlines.
That's on `xenl', AKA `eat_newline_glitch', terminals, AIUI,
Hi Ken,
http://www.nongnu.org/nmh/rfc.html
I don't know who started that list, but I've been adding stuff to that
based on what I use during nmh development. I hadn't needed to use
RFC 2049, so that's why I didn't add it. But it probably makes sense
to add it at some point.
I find the
Hi Ken,
As an aside ... I do wonder if there are actual working versions of
most of the terminals in the termcap/terminfo database.
Clearly what's needed is a terminal emulator that has switches to
properly emulate all of these termcap/terminfo capabilities, with canned
configurations for
Hi Ken,
Across different rules, where does the CFWS end up at when a reduction
happens? However, it really doesn't matter where CFWS ends up since
we ignore it (except for the case of the old-style addresses).
Rather than settle all the shift/reduce conflicts with %prec, etc.,
could it be
Hi Ken,
The advantage here is that yacc/bison (I'm only targeting bison) is
pretty standard in terms of basic Unix development tools. Another
tool isn't necessarily so widespread.
That's true, though http://piumarta.com/software/peg/peg.1.html looks
quite fun. Will nmh.tar ship bison's C
Hi David,
I suspect you're running afowl of the 'linespace' test.
Yeah, I do that a lot because body:component= turns a blank line
in replied text into one with a trailing space.
Yes, that's really icky. Would be nice if mhl supported some kind of
`strip' to prune those off afterwards.
Hi Norm,
The problem was that, for reasons I do not remember, Chart.pj had
Microsoft line endings and your blanks, above includes '\r', and
probably other white space. Confession: It took me an hour to figure
that out.
Useful ways to look for odd bytes in a file ./foo. One approach is to
Hi David,
Yeah, I do that a lot because body:component= turns a blank
line in replied text into one with a trailing space.
Yes, that's really icky. Would be nice if mhl supported some kind
of `strip' to prune those off afterwards.
...
Is there any need to keep the trailing
Hi Norm,
tr -d '\012 -~' foo | od -c
I have a request (a request, not a demand!). When a command has both a
'-' option and a '--' option for the same feature, use the '--'
version in examples.
I take your point, but the problem is... I don't know the long versions
and my fingers certainly
Hi,
http://git.savannah.gnu.org/cgit/nmh.git/about/ looks a little odd; it
just says foo.
Cheers, Ralph.
___
Nmh-workers mailing list
Nmh-workers@nongnu.org
https://lists.nongnu.org/mailman/listinfo/nmh-workers
Hi Lyndon,
My only concern is with where the manpages get installed. Currently
we put them under .../share/man/. That seems to be the autoconf
default location. But for most of the systems I have access to,
versions of man that adapt their search path based on $PATH search on
../man for
Hi Ken,
%{delivered-to}%(match k...@pobox.com)
From: Ken Hornstein k...@pobox.com
%(void(num 1))%|(void(num))%%
%(zero)\
From: Ken Hornstein work@address
%
Assuming Jon's addresses are jon-...@hisdomain.com then he might find
%(amatch) useful to match the start of the address for `jon-'
Hi David,
This behavior of file(1) has been noted in the past:
$ file --brief --mime-type .mh_profile
message/news
Could file(1) be taught about an .mh_profile, --mime-type or not?
Cheers, Ralph.
___
Nmh-workers mailing list
Hi Ken,
But if you read the rest of the note, in terms of Unix utilities all
that is really mentioned is using your favorite text editor (e and
Word Perfect are mentioned!) to compose messages.
Wouldn't sed and grep examples be a bit off-topic once it's been made
clear they're text files?
Hi Lyndon,
Or perhaps the time has come to switch stores to one with a similar
structure to upasfs's?
Upas' store is a single flat mbox file. The magic is that upasfs is
a file server that interprets the contents of the mbox file on the fly
and presents the file's contents in an
Hi Ken,
I understand that some people find that useful, but I have not yet
been persuaded that it is nmh's job to provide an interface for
generic Unix text processing tools.
That was there, whether by design or accident, from the start. And I
suspect it was clearly by design. Regardless,
Hi Ken,
If, however, you want an alias to be expanded when you do a 'repl',
that's not possible.
Wondering about the case of the draft having email addresses turned back
into aliases, I played with whom(1).
$ echo To: me draft
$ whom draft
-- Network Recipients --
ralph at
Hi Bill,
What has likely changed are expectations. Users are used to HTML
messages with inline images and links. As others as said, some sort of
GUI makes nmh useful to a broader range of users.
If that GUI were started today, would it be HTML served up by a local
nmh server rather than Tk,
Hi Lyndon,
Trying to keep everything in sync in the face of errors, without
transactions, would require some horribly complicated code.
You'd work with a temporary directory and atomically rename it to
mail/inbox/42.
you still face the problem of non-MH commands directly modifying
Hi Lyndon,
You have an fsck-like that can report inconsistencies, this is
simply the unpack followed by a diff, and would likely be the same
command that refreshes. Just as now, if the user edits
mail/inbox/42 then he's expected to know how to miss shooting his
foot. This is Unix.
Hi,
I've just seen http://www.pldaniels.com/ripmime/ make its way into
Debian. Thought it might be of general interest to the list. There's
already recode(1), base64(1), uudecode(1), etc., but this works at the
whole email level and some Microsoft oddities.
Cheers, Ralph.
Hi David,
Nmh-mhbuild-TYPE: [ARGSTRING] file://FILE
What is the URL style file:// syntax needed for? Is it sometimes
http or something?
No, while it looks familiar, it's only used here as a delimiter for
the filename. It's very unlikely to be part of a filename. Any other
Hi David,
Swap the order, /path/first,
Then the arguments couldn't contain file://. (And I didn't
want to require quoting of the filename because Attach: doesn't
and that works out well.)
Think I'm missing something. I thought
nmh-mhbuild-image/png: /home/ralph/mail/inbox/42.3.png
301 - 400 of 1415 matches
Mail list logo
