Mutt without Portage/in Local Overlay, for Air-Gappers

[miroR]

EDIT 2017-07-15
I have tried to fut forth the problem in a less deniable way:

http://www.croatiafidelis.hr/foss/cap/c ... entoo.webm

More following, with time (this was a lot of work). Just, can the followers of this topic guess what method I used to get the mutt deployed, IMO, properly, as seen in the last 1 and 1/2 minutes of that 5 minutes video?

Namely, most of the problems I tried to explain textually here (and now some are shown in video) are still there, and have been there for, apparently, long time.


EDIT 2014-11-01
Unfortunately, for a human memory fault, or processing fault in my head, there is one error --skip to second half of that post for the explanation of the error-- that is hard to correct now, after the entire event. So just bear in mind that gnupg-1 is perfectly installable with Mutt. But in, as far as I go, the wrong way.
I also just tried hard to correct the error regarding the manual install out-of-portage, which is viable (but the overlay-one is certainly preferable), and I couldn't correct all. The inconsistency is marked with anorher "2014-11-01" further below. Sorry!
---
EDIT 2014-10-29:
Renaming the topic to:
Mutt without Portage/in Local Overlay, for Air-Gappers
[ Because of Lack of gnupg-1 Support ]
---
Previos Title was:
Install Mutt without Portage, and Why, for Air-Gappers
---
After doing a regular "emerge mutt", and trying to use it, I got:

Code: Select all

ukrainian@mybox ~ $ mutt
Error in /home/ukrainian/.muttrc, line 11: ssl_usesystemcerts: unknown variable
source: errors in /home/ukrainian/.muttrc
Press any key to continue...
ukrainian@mybox ~ $

So I did:

Code: Select all

# emerge -C mutt

(
actually you should first see more in detail this directory:

Code: Select all

# ls -l /usr/portage/mail-client/mutt/
total 88
-rw-r--r-- 1 portage portage 56984 2014-05-15 08:31 ChangeLog
drwxr-xr-x 2 portage portage  4096 2014-03-13 22:01 files
-rw-r--r-- 1 portage portage  4409 2014-05-15 08:31 Manifest
-rw-r--r-- 1 portage portage   650 2009-10-27 22:47 metadata.xml
-rw-r--r-- 1 portage portage  8041 2014-05-14 18:31 mutt-1.5.22-r3.ebuild
-rw-r--r-- 1 portage portage  8055 2014-04-08 22:37 mutt-1.5.23-r2.ebuild
mybox ukrainian #

and if it changed, then maybe use emerge... Such as if this:

Code: Select all

# view /usr/portage/mail-client/mutt/mutt-1.5.23-r2.ebuild
...[snip]...
    gpg?     ( >=app-crypt/gpgme-0.9.0 )
...[snip]...

is not anymore the case (and other things possibly).

Why? Because you are better of trying to use gpg-1 instead --that one above is a GUI interface, goes with the gpg-2 which you have to use gpg-agent and GUI to type password; not the mutt way really
)

The mutt in portage is the unstable of March 2014 from mutt.org, so I just checked it with the signature pointed to there. It's fine. Sure portage would not accept otherwise, but I double and triple check things often.

I will use it without the gentoo patches, however. I really don't need any GUI to type in my passwords in Mutt and other such desktop poetteringware-like stuff.

So:

Code: Select all

# cd /usr/src
# tar xzvf /usr/portage/distfiles/mutt-1.5.23.tar.gz
# cd mutt-1.5.23/
# view INSTALL

and, after I unmounted all

Code: Select all

# umount /<whatever-was-mounted> #proof-reading-note: '-xdev' option to find, and not needed
# find / -name '*' > /root/FIND_mutt-1.5.23_`date +%s`_BEFORE-PRECAUTION.txt
# ./configure --prefix=/usr 2>&1 | tee /root/mutt-1.5.23_`date +%s`_configure

This is the one I'll use for diff'ing (previous is lest I forget).

Code: Select all

# find / -name '*' > /root/FIND_mutt-1.5.23_`date +%s`_BEFORE.txt
# make install 2>&1 | tee /root/mutt-1.5.23_`date +%s`_make_install
# find / -name '*' > /root/FIND_mutt-1.5.23_`date +%s`_AFTER.txt

Code: Select all

# diff /root/FIND_mutt-1.5.23_1413405
FIND_mutt-1.5.23_1413405491_BEFORE.txt  FIND_mutt-1.5.23_1413405618_AFTER.txt
# diff /root/FIND_mutt-1.5.23_1413405*

shows the diff, but to get the list of files that belong to the newly
installed mutt more is needed.

This is how I got the list:

EDIT: 2014-10-29 I added this substring:

Code: Select all

|\/usr\/src\/

into the string for egrep below, and I can't check it, because I now have my local overlay modified ebuild which i installed Mutt with. But that is the principle which should work in some similar case of quick fix via installing outside portage... With that string, therer is much less output, that Navar, further below, noticed something was wrong with...

Code: Select all

diff /root/FIND_mutt-1.5.23_1413405* | grep '> ' | sed 's/> \//\//' | egrep -v '\/proc\/|\/root\/|\/usr\/src\/'
/usr/share/man/man1/smime_keys.1
/usr/share/man/man1/pgpring.1
/usr/share/man/man1/flea.1
/usr/share/man/man1/pgpewrap.1
/usr/share/man/man1/muttbug.1
/usr/share/man/man1/mutt.1
/usr/share/man/man5/muttrc.5
/usr/share/man/man5/mbox.5
/usr/share/man/man5/mmdf.5
/usr/share/doc/mutt
/usr/share/doc/mutt/applying-patches.txt
/usr/share/doc/mutt/devel-notes.txt
/usr/share/doc/mutt/security.html
/usr/share/doc/mutt/INSTALL
/usr/share/doc/mutt/README.SSL
/usr/share/doc/mutt/COPYRIGHT
/usr/share/doc/mutt/samples
/usr/share/doc/mutt/samples/Mush.rc
/usr/share/doc/mutt/samples/pgp2.rc
/usr/share/doc/mutt/samples/mutt_xtitle
/usr/share/doc/mutt/samples/sample.mailcap
/usr/share/doc/mutt/samples/smime_keys_test.pl
/usr/share/doc/mutt/samples/iconv
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.6.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.6-cjk.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-4.1.5.rc
/usr/share/doc/mutt/samples/iconv/iconv.irix-6.5.rc
/usr/share/doc/mutt/samples/iconv/iconv.hpux-10.20.rc
/usr/share/doc/mutt/samples/iconv/iconv.hpux-10.01.rc
/usr/share/doc/mutt/samples/iconv/iconv.osf1-4.0d.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-3.2.5.rc
/usr/share/doc/mutt/samples/iconv/iconv.osf1-4.0a.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.4.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.5.1.rc
/usr/share/doc/mutt/samples/iconv/iconv.glibc-2.1.3.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-4.3.2.rc
/usr/share/doc/mutt/samples/iconv/iconv.glibc-2.1.90.rc
/usr/share/doc/mutt/samples/iconv/iconv.freebsd-3.3.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.7.rc
/usr/share/doc/mutt/samples/iconv/iconv.hpux-11.00.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-4.2.0.rc
/usr/share/doc/mutt/samples/gpg.rc
/usr/share/doc/mutt/samples/colors.linux
/usr/share/doc/mutt/samples/sample.muttrc
/usr/share/doc/mutt/samples/ca-bundle.crt
/usr/share/doc/mutt/samples/pgp6.rc
/usr/share/doc/mutt/samples/smime.rc
/usr/share/doc/mutt/samples/sample.muttrc-tlr
/usr/share/doc/mutt/samples/colors.default
/usr/share/doc/mutt/samples/pgp5.rc
/usr/share/doc/mutt/samples/Tin.rc
/usr/share/doc/mutt/samples/Pine.rc
/usr/share/doc/mutt/PGP-Notes.txt
/usr/share/doc/mutt/ChangeLog
/usr/share/doc/mutt/smime-notes.txt
/usr/share/doc/mutt/tuning.html
/usr/share/doc/mutt/advancedusage.html
/usr/share/doc/mutt/reference.html
/usr/share/doc/mutt/configuration.html
/usr/share/doc/mutt/mimesupport.html
/usr/share/doc/mutt/manual.html
/usr/share/doc/mutt/miscellany.html
/usr/share/doc/mutt/README
/usr/share/doc/mutt/README.SECURITY
/usr/share/doc/mutt/GPL
/usr/share/doc/mutt/gettingstarted.html
/usr/share/doc/mutt/optionalfeatures.html
/usr/share/doc/mutt/TODO
/usr/share/doc/mutt/NEWS
/usr/share/doc/mutt/intro.html
/usr/share/doc/mutt/index.html
/usr/share/doc/mutt/patch-notes.txt
/usr/share/doc/mutt/manual.txt
/usr/share/locale/id/LC_MESSAGES/mutt.mo
/usr/share/locale/de/LC_MESSAGES/mutt.mo
/usr/share/locale/bg/LC_MESSAGES/mutt.mo
/usr/share/locale/it/LC_MESSAGES/mutt.mo
/usr/share/locale/lt/LC_MESSAGES/mutt.mo
/usr/share/locale/eu/LC_MESSAGES/mutt.mo
/usr/share/locale/cs/LC_MESSAGES/mutt.mo
/usr/share/locale/zh_TW/LC_MESSAGES/mutt.mo
/usr/share/locale/pl/LC_MESSAGES/mutt.mo
/usr/share/locale/tr/LC_MESSAGES/mutt.mo
/usr/share/locale/fr/LC_MESSAGES/mutt.mo
/usr/share/locale/nl/LC_MESSAGES/mutt.mo
/usr/share/locale/ko/LC_MESSAGES/mutt.mo
/usr/share/locale/ga/LC_MESSAGES/mutt.mo
/usr/share/locale/et/LC_MESSAGES/mutt.mo
/usr/share/locale/ca/LC_MESSAGES/mutt.mo
/usr/share/locale/zh_CN/LC_MESSAGES/mutt.mo
/usr/share/locale/gl/LC_MESSAGES/mutt.mo
/usr/share/locale/uk/LC_MESSAGES/mutt.mo
/usr/share/locale/ru/LC_MESSAGES/mutt.mo
/usr/share/locale/sv/LC_MESSAGES/mutt.mo
/usr/share/locale/hu/LC_MESSAGES/mutt.mo
/usr/share/locale/sk/LC_MESSAGES/mutt.mo
/usr/share/locale/pt_BR/LC_MESSAGES/mutt.mo
/usr/share/locale/el/LC_MESSAGES/mutt.mo
/usr/share/locale/ja/LC_MESSAGES/mutt.mo
/usr/share/locale/es/LC_MESSAGES/mutt.mo
/usr/share/locale/da/LC_MESSAGES/mutt.mo
/usr/share/locale/eo/LC_MESSAGES/mutt.mo
/usr/bin/mutt
/usr/bin/pgpring
/usr/bin/flea
/usr/bin/smime_keys
/usr/bin/muttbug
/usr/bin/pgpewrap
/usr/etc
/usr/etc/mime.types.dist
/usr/etc/mime.types
/usr/etc/Muttrc.dist
/usr/etc/Muttrc

Sure enough, I really need to do:

Code: Select all

diff /root/FIND_mutt-1.5.23_1413405* | grep '> ' | sed 's/> \//\//' | egrep -v '\/proc\/|\/root\/|\/usr\/src\/' > /root/mutt-1.5.23_`date +%s`_FILES

to stow that in a file (and stow it somewhere more appropriate than /root. This way or some othe, but some way is surely needed in case you want to remove these some day. Portage keeps no track whatsoever of files installed this way.

I just stow away three files:

Code: Select all

# ls -ltr /root/mutt-1.5.23_141340*
-rw-r--r-- 1 root root 10802 2014-10-15 22:35 /root/mutt-1.5.23_1413405295_configure
-rw-r--r-- 1 root root 44661 2014-10-15 22:40 /root/mutt-1.5.23_1413405572_make_install
-rw-r--r-- 1 root root  7490 2014-10-15 22:53 /root/mutt-1.5.23_1413406411_FILES

of which the last one (..._FILES) is the most important to have, if I want to remove the package.

Actually I just need it, because judging by not being advisory to give the portage's mutt the gpg flag (and get the whole GUI silly --for security aware people-- stuff for gpg), I now see that I, with the source compilation should have followed different route regarding ssl, another thing I use FOSS Linux for. I can't log into my dovecot server because no ssl support...

I know that because on trying to launch mutt I got these errors:

Code: Select all

$ mutt
Error in /home/ukrainian/.muttrc, line 8: ssl_starttls: unknown variable
Error in /home/ukrainian/.muttrc, line 9: ssl_force_tls: unknown variable
Error in /home/ukrainian/.muttrc, line 10: ssl_force_tls: unknown variable
Error in /home/ukrainian/.muttrc, line 11: ssl_usesystemcerts: unknown variable
Error in /home/ukrainian/.muttrc, line 12: imap_passive: unknown variable
source: errors in /home/ukrainian/.muttrc
$

(I'm not Ukrainian, but Croat, a related nation, but I support Ukrainians and wish Russians to be humane and leave other people's lands alone.)

So:

Code: Select all

for i in `cat /root/mutt-1.5.23_1413406411_FILES`; do ls -ld $i ; read FAKE ; done ;

as I first want to see all the files to uninstall scroll before me, and stop for closer look if I decide.

"read FAKE" is for you to press Enter to go on, or keep pressed for it to go all the way.

Now the first run:

Code: Select all

for i in `cat /root/mutt-1.5.23_1413406411_FILES`; do if [ -f "$i" ] ; then rm -v $i ; fi ; done ;

That removed files. And the second run:

Code: Select all

for i in `cat /root/mutt-1.5.23_1413406411_FILES`; do if [ -d "$i" ] ; then rmdir -v $i ; fi ; done ;

And that removed directories.

And I can start over (I have to, unfortunately, because I need the ssl flag of the mutt configure) with system not dirtied with files not-known neither by portage nor by me which would have been the case had I not had logged what I was installing.

This must work fine:

Code: Select all

 ./configure --prefix=/usr --with-homespool=Maildir --enable-pop -enable-imap --with-ssl --with-gnutls --with-sasl 2>&1 | tee /root/mutt-1.5.23_`date +%s

And I repeated the other steps.

EDIT 2014-11-01 Upon trying to correct here, I introduced inconsistencies. The method is viable, but the demonstration is not. Below is the diff inconsistent with the commands given.

Code: Select all

mybox mutt-1.5.23 # diff /root/mutt-1.5.23_*_FILES
mutt-1.5.23_1413406411_FILES  mutt-1.5.23_1413408034_FILES  
mybox mutt-1.5.23 # diff /root/mutt-1.5.23_*_FILES

143
> /usr/src/mutt-1.5.23/pop_auth.o
144a146
> /usr/src/mutt-1.5.23/pop.o
151a154
> /usr/src/mutt-1.5.23/mutt_ssl.o
166a170
> /usr/src/mutt-1.5.23/pop_lib.o
167a172
> /usr/src/mutt-1.5.23/account.o
191a197
> /usr/src/mutt-1.5.23/mutt_tunnel.o
192a199
> /usr/src/mutt-1.5.23/mutt_socket.o
202a210
> /usr/src/mutt-1.5.23/bcache.o
211a220,229
> /usr/src/mutt-1.5.23/imap/util.o
> /usr/src/mutt-1.5.23/imap/message.o
> /usr/src/mutt-1.5.23/imap/auth_sasl.o
> /usr/src/mutt-1.5.23/imap/imap.o
> /usr/src/mutt-1.5.23/imap/auth_login.o
> /usr/src/mutt-1.5.23/imap/libimap.a
> /usr/src/mutt-1.5.23/imap/command.o
> /usr/src/mutt-1.5.23/imap/browse.o
> /usr/src/mutt-1.5.23/imap/auth.o
> /usr/src/mutt-1.5.23/imap/utf7.o
212a231
> /usr/src/mutt-1.5.23/mutt_sasl.o

That is the diff btwn the previous, deleted and this install, which is to stay.

And, upon:

Code: Select all

$ mutt # That is "$", meaning as regular user, not anymore root

I get this page (well I also have other things set up, but no time to explain)

Code: Select all

q:Exit  ?:Help
This certificate belongs to:
   gitb-ukrainian
   Unknown
   Croatia Fidelis
   Officium Centralis
   Zagreb

This certificate was issued by:
   gitb-ukrainian
   Unknown
   Croatia Fidelis
   Officium Centralis
   Zagreb

This certificate is valid
   from Aug 29 08:52:03 2014 GMT
     to Aug 29 08:52:03 2015 GMT

Fingerprint: 071A 2DC0 F91C 7E76 0020 BDEF 948B E119

...[snip]...

-- Mutt: SSL Certificate check (certificate 0 of 0 in chain)
(r)eject, accept (o)nce, (a)ccept always

only nicer that I can describe here.

Now at least two reasons why I didn't file a bug for the error in portage's
current mutt package.

I'm not yet very familiar with reporting bugs, but that's half a reason, To
make it complete first reason, I work far too slowly to find time for it.

On the other side, I do understand that this is the better way to install mutt
currently, for people who tend to like air-gapping, and there are many of such
people (amoong whom I belong), ending up on Gentoo, as the last bastion of
still some-scale free computing (potteringware is not, as M$ and apple is
not), it is simply wrong to add all those GUIs for just gnupg to have
passwords. I wrote about it:

Air-Gapped Gentoo Install, Tentative
https://forums.gentoo.org/viewtopic-t-9 ... ml#7551458

( and a little more on what I chose for my mail setup is one of the issues here:

Postfix smtp-tls-wrapper, Bkp/Cloning Mthd, A Zerk Provider
https://forums.gentoo.org/viewtopic-t-999436.html

but, to understand all more easily, the most important link, referred below by Navar as sorely missing , pls. look up this topic:

how to disable (sanitize) gpg2 GUI features (pinentry)?
http://forums.gentoo.org/viewtopic-t-63 ... ml#7534996
)

and since it's mainstream, it would be much more work talking maintainers to
do it than to tell people how to do it the right way (as air-gappers like me
consider it to be). That is the second reason.

I just tried and sent a note that I posted this tip on Gentoo Forums to ago (Agostino Sarubbo, in charge of mutt in Gentoo, and to grobian, Fabian Groffen). Maybe that should go some way for the devs.

[Ant P.]

After doing a regular "emerge mutt", and trying to use it, I got:

Code: Select all

ukrainian@mybox ~ $ mutt
Error in /home/ukrainian/.muttrc, line 11: ssl_usesystemcerts: unknown variable
source: errors in /home/ukrainian/.muttrc
Press any key to continue...
ukrainian@mybox ~ $

So I did:

Code: Select all

# emerge -C mutt

So instead of paying attention to the output of emerge, or trying to actually diagnose the problem, your first reaction is to panic, rip it out and compile it by hand. Just a cargo-cult, ricer, knee-jerk reaction. Okay.

(
actually you should first see more in detail this directory:

Code: Select all

# ls -l /usr/portage/mail-client/mutt/
total 88
-rw-r--r-- 1 portage portage 56984 2014-05-15 08:31 ChangeLog
drwxr-xr-x 2 portage portage  4096 2014-03-13 22:01 files
-rw-r--r-- 1 portage portage  4409 2014-05-15 08:31 Manifest
-rw-r--r-- 1 portage portage   650 2009-10-27 22:47 metadata.xml
-rw-r--r-- 1 portage portage  8041 2014-05-14 18:31 mutt-1.5.22-r3.ebuild
-rw-r--r-- 1 portage portage  8055 2014-04-08 22:37 mutt-1.5.23-r2.ebuild
mybox ukrainian #

and if it changed, then maybe use emerge...

What a waste of CPU time.

Such as if this:

Code: Select all

# view /usr/portage/mail-client/mutt/mutt-1.5.23-r2.ebuild
...[snip]...
    gpg?     ( >=app-crypt/gpgme-0.9.0 )
...[snip]...

is not anymore the case (and other things possibly).

Why? Because you are better of trying to use gpg-1 instead --that one above is a GUI interface, goes with the gpg-2 which you have to use gpg-agent and GUI to type password; not the mutt way really
)

The mutt in portage is the unstable of March 2014 from mutt.org, so I just checked it with the signature pointed to there. It's fine. Sure portage would not accept otherwise, but I double and triple check things often.

I will use it without the gentoo patches, however. I really don't need any GUI to type in my passwords in Mutt and other such desktop poetteringware-like stuff.

Or you could use `eselect pinentry ncurses` instead of screaming about "poetteringware" like a raving lunatic. God forbid you actually learn to use the damn software you're running!

So:

Code: Select all

# cd /usr/src
# tar xzvf /usr/portage/distfiles/mutt-1.5.23.tar.gz
# cd mutt-1.5.23/
# view INSTALL

and, after I unmounted all

Code: Select all

# umount /<whatever-was-mounted> #proof-reading-note: '-xdev' option to find, and not needed
# find / -name '*' > /root/FIND_mutt-1.5.23_`date +%s`_BEFORE-PRECAUTION.txt
# ./configure --prefix=/usr 2>&1 | tee /root/mutt-1.5.23_`date +%s`_configure

This is the one I'll use for diff'ing (previous is lest I forget).

Code: Select all

# find / -name '*' > /root/FIND_mutt-1.5.23_`date +%s`_BEFORE.txt
# make install 2>&1 | tee /root/mutt-1.5.23_`date +%s`_make_install
# find / -name '*' > /root/FIND_mutt-1.5.23_`date +%s`_AFTER.txt

Code: Select all

# diff /root/FIND_mutt-1.5.23_1413405
FIND_mutt-1.5.23_1413405491_BEFORE.txt  FIND_mutt-1.5.23_1413405618_AFTER.txt
# diff /root/FIND_mutt-1.5.23_1413405*

shows the diff, but to get the list of files that belong to the newly
installed mutt more is needed.

This is how I got the list:

Code: Select all

diff /root/FIND_mutt-1.5.23_1413405* | grep '> ' | sed 's/> \//\//' | egrep -v '\/proc\/|\/root\/'
/usr/share/man/man1/smime_keys.1
/usr/share/man/man1/pgpring.1
/usr/share/man/man1/flea.1
/usr/share/man/man1/pgpewrap.1
/usr/share/man/man1/muttbug.1
/usr/share/man/man1/mutt.1
/usr/share/man/man5/muttrc.5
/usr/share/man/man5/mbox.5
/usr/share/man/man5/mmdf.5
/usr/share/doc/mutt
/usr/share/doc/mutt/applying-patches.txt
/usr/share/doc/mutt/devel-notes.txt
/usr/share/doc/mutt/security.html
/usr/share/doc/mutt/INSTALL
/usr/share/doc/mutt/README.SSL
/usr/share/doc/mutt/COPYRIGHT
/usr/share/doc/mutt/samples
/usr/share/doc/mutt/samples/Mush.rc
/usr/share/doc/mutt/samples/pgp2.rc
/usr/share/doc/mutt/samples/mutt_xtitle
/usr/share/doc/mutt/samples/sample.mailcap
/usr/share/doc/mutt/samples/smime_keys_test.pl
/usr/share/doc/mutt/samples/iconv
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.6.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.6-cjk.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-4.1.5.rc
/usr/share/doc/mutt/samples/iconv/iconv.irix-6.5.rc
/usr/share/doc/mutt/samples/iconv/iconv.hpux-10.20.rc
/usr/share/doc/mutt/samples/iconv/iconv.hpux-10.01.rc
/usr/share/doc/mutt/samples/iconv/iconv.osf1-4.0d.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-3.2.5.rc
/usr/share/doc/mutt/samples/iconv/iconv.osf1-4.0a.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.4.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.5.1.rc
/usr/share/doc/mutt/samples/iconv/iconv.glibc-2.1.3.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-4.3.2.rc
/usr/share/doc/mutt/samples/iconv/iconv.glibc-2.1.90.rc
/usr/share/doc/mutt/samples/iconv/iconv.freebsd-3.3.rc
/usr/share/doc/mutt/samples/iconv/iconv.solaris-2.7.rc
/usr/share/doc/mutt/samples/iconv/iconv.hpux-11.00.rc
/usr/share/doc/mutt/samples/iconv/iconv.aix-4.2.0.rc
/usr/share/doc/mutt/samples/gpg.rc
/usr/share/doc/mutt/samples/colors.linux
/usr/share/doc/mutt/samples/sample.muttrc
/usr/share/doc/mutt/samples/ca-bundle.crt
/usr/share/doc/mutt/samples/pgp6.rc
/usr/share/doc/mutt/samples/smime.rc
/usr/share/doc/mutt/samples/sample.muttrc-tlr
/usr/share/doc/mutt/samples/colors.default
/usr/share/doc/mutt/samples/pgp5.rc
/usr/share/doc/mutt/samples/Tin.rc
/usr/share/doc/mutt/samples/Pine.rc
/usr/share/doc/mutt/PGP-Notes.txt
/usr/share/doc/mutt/ChangeLog
/usr/share/doc/mutt/smime-notes.txt
/usr/share/doc/mutt/tuning.html
/usr/share/doc/mutt/advancedusage.html
/usr/share/doc/mutt/reference.html
/usr/share/doc/mutt/configuration.html
/usr/share/doc/mutt/mimesupport.html
/usr/share/doc/mutt/manual.html
/usr/share/doc/mutt/miscellany.html
/usr/share/doc/mutt/README
/usr/share/doc/mutt/README.SECURITY
/usr/share/doc/mutt/GPL
/usr/share/doc/mutt/gettingstarted.html
/usr/share/doc/mutt/optionalfeatures.html
/usr/share/doc/mutt/TODO
/usr/share/doc/mutt/NEWS
/usr/share/doc/mutt/intro.html
/usr/share/doc/mutt/index.html
/usr/share/doc/mutt/patch-notes.txt
/usr/share/doc/mutt/manual.txt
/usr/share/locale/id/LC_MESSAGES/mutt.mo
/usr/share/locale/de/LC_MESSAGES/mutt.mo
/usr/share/locale/bg/LC_MESSAGES/mutt.mo
/usr/share/locale/it/LC_MESSAGES/mutt.mo
/usr/share/locale/lt/LC_MESSAGES/mutt.mo
/usr/share/locale/eu/LC_MESSAGES/mutt.mo
/usr/share/locale/cs/LC_MESSAGES/mutt.mo
/usr/share/locale/zh_TW/LC_MESSAGES/mutt.mo
/usr/share/locale/pl/LC_MESSAGES/mutt.mo
/usr/share/locale/tr/LC_MESSAGES/mutt.mo
/usr/share/locale/fr/LC_MESSAGES/mutt.mo
/usr/share/locale/nl/LC_MESSAGES/mutt.mo
/usr/share/locale/ko/LC_MESSAGES/mutt.mo
/usr/share/locale/ga/LC_MESSAGES/mutt.mo
/usr/share/locale/et/LC_MESSAGES/mutt.mo
/usr/share/locale/ca/LC_MESSAGES/mutt.mo
/usr/share/locale/zh_CN/LC_MESSAGES/mutt.mo
/usr/share/locale/gl/LC_MESSAGES/mutt.mo
/usr/share/locale/uk/LC_MESSAGES/mutt.mo
/usr/share/locale/ru/LC_MESSAGES/mutt.mo
/usr/share/locale/sv/LC_MESSAGES/mutt.mo
/usr/share/locale/hu/LC_MESSAGES/mutt.mo
/usr/share/locale/sk/LC_MESSAGES/mutt.mo
/usr/share/locale/pt_BR/LC_MESSAGES/mutt.mo
/usr/share/locale/el/LC_MESSAGES/mutt.mo
/usr/share/locale/ja/LC_MESSAGES/mutt.mo
/usr/share/locale/es/LC_MESSAGES/mutt.mo
/usr/share/locale/da/LC_MESSAGES/mutt.mo
/usr/share/locale/eo/LC_MESSAGES/mutt.mo
/usr/bin/mutt
/usr/bin/pgpring
/usr/bin/flea
/usr/bin/smime_keys
/usr/bin/muttbug
/usr/bin/pgpewrap
/usr/src/mutt-1.5.23/rfc1524.o
/usr/src/mutt-1.5.23/score.o
/usr/src/mutt-1.5.23/pgp.o
/usr/src/mutt-1.5.23/cryptglue.o
/usr/src/mutt-1.5.23/curs_main.o
/usr/src/mutt-1.5.23/mbox.o
/usr/src/mutt-1.5.23/crypt-mod.o
/usr/src/mutt-1.5.23/parse.o
/usr/src/mutt-1.5.23/pgpewrap.o
/usr/src/mutt-1.5.23/crypt.o
/usr/src/mutt-1.5.23/crypt-mod-pgp-classic.o
/usr/src/mutt-1.5.23/getdomain.o
/usr/src/mutt-1.5.23/resize.o
/usr/src/mutt-1.5.23/pgpkey.o
/usr/src/mutt-1.5.23/pgpmicalg.o
/usr/src/mutt-1.5.23/mutt
/usr/src/mutt-1.5.23/handler.o
/usr/src/mutt-1.5.23/pgpring
/usr/src/mutt-1.5.23/dotlock.o
/usr/src/mutt-1.5.23/enter.o
/usr/src/mutt-1.5.23/mbyte.o
/usr/src/mutt-1.5.23/history.o
/usr/src/mutt-1.5.23/conststrings.o
/usr/src/mutt-1.5.23/patchlist.c
/usr/src/mutt-1.5.23/charset.o
/usr/src/mutt-1.5.23/hash.o
/usr/src/mutt-1.5.23/sort.o
/usr/src/mutt-1.5.23/commands.o
/usr/src/mutt-1.5.23/mh.o
/usr/src/mutt-1.5.23/ascii.o
/usr/src/mutt-1.5.23/pager.o
/usr/src/mutt-1.5.23/compose.o
/usr/src/mutt-1.5.23/txt2c
/usr/src/mutt-1.5.23/flags.o
/usr/src/mutt-1.5.23/gnupgparse.o
/usr/src/mutt-1.5.23/menu.o
/usr/src/mutt-1.5.23/pgppubring.o
/usr/src/mutt-1.5.23/addrbook.o
/usr/src/mutt-1.5.23/crypt-mod-smime-classic.o
/usr/src/mutt-1.5.23/alias.o
/usr/src/mutt-1.5.23/complete.o
/usr/src/mutt-1.5.23/lib.o
/usr/src/mutt-1.5.23/keymap.o
/usr/src/mutt-1.5.23/send.o
/usr/src/mutt-1.5.23/hdrline.o
/usr/src/mutt-1.5.23/sendlib.o
/usr/src/mutt-1.5.23/postpone.o
/usr/src/mutt-1.5.23/status.o
/usr/src/mutt-1.5.23/patchlist.o
/usr/src/mutt-1.5.23/group.o
/usr/src/mutt-1.5.23/browser.o
/usr/src/mutt-1.5.23/url.o
/usr/src/mutt-1.5.23/pgppacket.o
/usr/src/mutt-1.5.23/muttlib.o
/usr/src/mutt-1.5.23/help.o
/usr/src/mutt-1.5.23/mx.o
/usr/src/mutt-1.5.23/reldate.h
/usr/src/mutt-1.5.23/flea
/usr/src/mutt-1.5.23/from.o
/usr/src/mutt-1.5.23/md5.o
/usr/src/mutt-1.5.23/sha1.o
/usr/src/mutt-1.5.23/recvcmd.o
/usr/src/mutt-1.5.23/extlib.o
/usr/src/mutt-1.5.23/copy.o
/usr/src/mutt-1.5.23/system.o
/usr/src/mutt-1.5.23/thread.o
/usr/src/mutt-1.5.23/filter.o
/usr/src/mutt-1.5.23/rfc2047.o
/usr/src/mutt-1.5.23/main.o
/usr/src/mutt-1.5.23/hook.o
/usr/src/mutt-1.5.23/headers.o
/usr/src/mutt-1.5.23/pattern.o
/usr/src/mutt-1.5.23/doc/smime_keys.1
/usr/src/mutt-1.5.23/doc/makedoc
/usr/src/mutt-1.5.23/doc/muttrc.man
/usr/src/mutt-1.5.23/doc/instdoc
/usr/src/mutt-1.5.23/doc/pgpring.1
/usr/src/mutt-1.5.23/doc/makedoc.o
/usr/src/mutt-1.5.23/doc/pgpewrap.1
/usr/src/mutt-1.5.23/doc/mutt.1
/usr/src/mutt-1.5.23/doc/manual.xml
/usr/src/mutt-1.5.23/edit.o
/usr/src/mutt-1.5.23/smime.o
/usr/src/mutt-1.5.23/rfc2231.o
/usr/src/mutt-1.5.23/attach.o
/usr/src/mutt-1.5.23/curs_lib.o
/usr/src/mutt-1.5.23/init.o
/usr/src/mutt-1.5.23/date.o
/usr/src/mutt-1.5.23/color.o
/usr/src/mutt-1.5.23/base64.o
/usr/src/mutt-1.5.23/pgplib.o
/usr/src/mutt-1.5.23/recvattach.o
/usr/src/mutt-1.5.23/rfc822.o
/usr/src/mutt-1.5.23/editmsg.o
/usr/src/mutt-1.5.23/query.o
/usr/src/mutt-1.5.23/smime_keys
/usr/src/mutt-1.5.23/safe_asprintf.o
/usr/src/mutt-1.5.23/rfc3676.o
/usr/src/mutt-1.5.23/mutt_idna.o
/usr/src/mutt-1.5.23/buffy.o
/usr/src/mutt-1.5.23/pgpinvoke.o
/usr/src/mutt-1.5.23/signal.o
/usr/src/mutt-1.5.23/keymap_defs.h
/usr/src/mutt-1.5.23/conststrings.c
/usr/src/mutt-1.5.23/pgpewrap
/usr/etc
/usr/etc/mime.types.dist
/usr/etc/mime.types
/usr/etc/Muttrc.dist
/usr/etc/Muttrc

Sure enough, I really need to do:

Code: Select all

diff /root/FIND_mutt-1.5.23_1413405* | grep '> ' | sed 's/> \//\//' | egrep -v '\/proc\/|\/root\/' > /root/mutt-1.5.23_`date +%s`_FILES

to stow that in a file (and stow it somewhere more appropriate than /root. This way or some othe, but some way is surely needed in case you want to remove these some day. Portage keeps no track whatsoever of files installed this way.

This is fucking ridiculous. Not only that, you're flat out wrong and didn't RTFM at any point. /var/db/pkg/*/*/CONTENTS

I just stow away three files:

Code: Select all

# ls -ltr /root/mutt-1.5.23_141340*
-rw-r--r-- 1 root root 10802 2014-10-15 22:35 /root/mutt-1.5.23_1413405295_configure
-rw-r--r-- 1 root root 44661 2014-10-15 22:40 /root/mutt-1.5.23_1413405572_make_install
-rw-r--r-- 1 root root  7490 2014-10-15 22:53 /root/mutt-1.5.23_1413406411_FILES

of which the last one (..._FILES) is the most important to have, if I want to remove the package.

And how do you think portage knows which files emerge -C should remove? /dev/random?

Actually I just need it, because judging by not being advisory to give the portage's mutt the gpg flag (and get the whole GUI silly --for security aware people-- stuff for gpg), I now see that I, with the source compilation should have followed different route regarding ssl, another thing I use FOSS Linux for. I can't log into my dovecot server because no ssl support...

I know that because on trying to launch mutt I got these errors:

Code: Select all

$ mutt
Error in /home/ukrainian/.muttrc, line 8: ssl_starttls: unknown variable
Error in /home/ukrainian/.muttrc, line 9: ssl_force_tls: unknown variable
Error in /home/ukrainian/.muttrc, line 10: ssl_force_tls: unknown variable
Error in /home/ukrainian/.muttrc, line 11: ssl_usesystemcerts: unknown variable
Error in /home/ukrainian/.muttrc, line 12: imap_passive: unknown variable
source: errors in /home/ukrainian/.muttrc
$

So after all that, your insane shotgun debugging approach screwed up your system even worse than it was to begin with! Bravo.

And you have the nerve to put this crap under Documentation?!

I'm not even going to bother rebuking any of the parts after that, it descended into outright political trolling.

Stop flooding our forums with this inane crap ffs. You don't need technical help, you need psychiatric help.

[steveL]

This is fucking ridiculous.

I agree; WTF is wrong with you? Why are you quoting massive stacks of file output?

And stop swearing so directly; you're not in OTW.

Stop flooding our forums with this inane crap ffs. You don't need technical help, you need psychiatric help.

This goes too far as well, imo; I heartily concur that miror has not found the right length for his posts yet.

But hopefully he'll get better over time.

What's your excuse?

[i92guboj]

As someone said, this is not GC nor it is OTW, so we won't tolerate any direct attacks of any kind here (not that they should be tolerated in OTW either, anyway). This is a support forum, not a place for name calling and/or witch hunting.

If you have any technical arguments against a how-to that's posted here just express them in a civil manner so that people around are aware of the downsides of procedure posted by the OP.

If you think the OP is either lying or doing something that violates any forum policy, please, just report it.

[miroR]

I just niow read all the three posts above. I need time to see which of them need my reply, and what I will be able to reply.

[miroR]

...
Or you could use `eselect pinentry ncurses` instead...
...

No, the point is, thanks to khayyam
(
how to disable (sanitize) gpg2 GUI features (pinentry)?
http://forums.gentoo.org/viewtopic-t-63 ... ml#7534996
)
, as can be read in some of the links pointed to in this tip, who fixed it, there is gnupg-1 in portage.
Mutt should be fine with gnupg-1 without stupid frills.
And I thanked khayyam for fixing it back then; since I really want a gnupg that doesn't need assistance to just read a password. That is unreasonable to force on users.

...
Not only that, you're flat out wrong and didn't RTFM at any point. /var/db/pkg/*/*/CONTENTS
...

I see. It is there to be found. Reading so much all the time, Missed that info somewhere.

Thanks for warning the fellow, steveL and i92guboj!

[miroR]

...
Not only that, you're flat out wrong and didn't RTFM at any point. /var/db/pkg/*/*/CONTENTS
...

I see. It is there to be found. Reading so much all the time, Missed that info somewhere.
...

Hey, I may have not known before about that database in /var, but the tip is still just fine and completely correct!

Because that database holds only what portage installs. And I don't see another way to check on what's installed than what I suggested, apart from, sure, safekeeping the 'make install''s log.

Because it is all in make install's STDOUT, and only needs to be kept. Portage will never know about it.

Surely the best thing is if the Mutt maintainers decide to keep with what Michael Elkins, the Mutt Dude, preaches, and not the, sorry to have to say, GUI crap on top of Mutt just for the sake of reading a password from user. That is crap!

And that is why I sent the mail to two of the Mutt maintainers.

[steveL]

I may have not known before about that database in /var, but the tip is still just fine and completely correct!

Because that database holds only what portage installs. And I don't see another way to check on what's installed than what I suggested, apart from, sure, safekeeping the 'make install''s log.

Because it is all in make install's STDOUT, and only needs to be kept. Portage will never know about it.

This doesn't make sense to me, as it's based on a misconception. The point is that portage will install whatever the makefile installs to DESTDIR (which is a root image.)

The most that you'd have to "worry" about is the ebuild doing something customised for Gentoo. And that's all transparently in front of you in the ebuild. Hell you can just -r bump it and rework it if you want, so it's hardly obfuscated.

So, I'm not sure what problem you're trying to solve by acting as if portage is an untrusted black box, instead of just building on top of what it provides.

[miroR]

I may have not known before about that database in /var, but the tip is still just fine and completely correct!

Because that database holds only what portage installs. And I don't see another way to check on what's installed than what I suggested, apart from, sure, safekeeping the 'make install''s log.

Because it is all in make install's STDOUT, and only needs to be kept. Portage will never know about it.

This doesn't make sense to me, as it's based on a misconception. The point is that portage will install whatever the makefile installs to DESTDIR (which is a root image.)

The most that you'd have to "worry" about is the ebuild doing something customised for Gentoo. And that's all transparently in front of you in the ebuild. Hell you can just -r bump it and rework it if you want, so it's hardly obfuscated.

So, I'm not sure what problem you're trying to solve by acting as if portage is an untrusted black box, instead of just building on top of what it provides.

I said it in the first post of this topic you are reading.

I don't want to be forced to install gnupg-2, because it introduces GUI based gpgme.

I don't even have ssh-agent in Gentoo, because I have not had the need for it yet (I have it in Debian, and that one is not the problem. That one, the ssh-agent (author Daniel Robbins, the initiator of Gentoo, but missing for me in Gentoo... maybe by his own fault, really don't know)...

[That one, the ssh-agent] was previously functionally taken in by dbus in my Debian, and is so by default, with all its "remote call" non-free functionality as you pointed out in various posts of yours in the "Why is Gentoo not switching to systemd?" topic. Now that was a problem, it was, because it is now solved for me in Debian, via mirabilos' MirDebian WTF packages)...

Why is that so hard to understand? When, among other reasons, also Michael Elkins, the Mutt Dude, is, as the maintainers of Mutt in Gentoo say somewhere in the ebuilds, [paraphrasing, from memory] "uncooperative" and even "doesn't reply" to their "mails"... ?

Mutt is not meant to work with GUI programs like that, IMO.

And those introduce vulnerabilities, don't they? Three or more programming layers when one was enough, to just take in a password! Forking of something, as the discussion went about it, but on the links pointed at and explained in my "Air-Gapped Gentoo Install, Tentative" topic.

Why, oh why is that so hard to grasp? At least the last argument, how the Mutt Dude does not accept patches and neither returns replies (probably after he said it all very clearly previously)?

[steveL]

Why, oh why is that so hard to grasp? At least the last argument, how the Mutt Dude does not accept patches and neither returns replies (probably after he said it all very clearly previously)?

Eh? I was only talking about the point you made about portage and make install.

Also, you don't need to quote everything, just as an aside. It only clutters the thread, and doesn't add to the conversation.

[Slippery Jim]

Just want to say to MiroR, nice job on the HOWTO. I found it very clear and easy to follow. Whether or not there's a better way to do the same thing, it's that clarity which makes it easy for people to try it out and easily compare the results against other experiments. So thanks for contributing to the community, and being positive and constructive.

[miroR]

Just want to say to MiroR, nice job on the HOWTO. I found it very clear and easy to follow. Whether or not there's a better way to do the same thing, it's that clarity which makes it easy for people to try it out and easily compare the results against other experiments. So thanks for contributing to the community, and being positive and constructive.

Thanx!
But see also the post below on how it should be done; and why I can't yet do it the really right way... Even though the other right way is kindly pressure the maintainers to do it for us who don't possess the know-how yet...
Maintainers, pls. get us the option to use plain gnupg-1 with portage's mutt! (But I won't send them another message in that regard. Our devs are often short with time, and really busy...)

[miroR]

Eh? I was only talking about the point you made about portage and make install.

I see your point with the emphasis that you gave it now.
But... But I can't currently find time such as a weekend (or two or three; I'm not unintelligent, but neither top class for sure)... to study and figure out how to do it by modifying the ebuild in my own overlay. I wish soo much to be able to do that one day, even though I'm 60 in three years... I don't even understand what "...<expletive> you can just -r bump it and rework it if you want"...

[Navar]

@OP, some suggestions:

app-crypt/gnupg-1.4.18 is available in the tree. Why not simply mask >app-crypt/gnupg-2.0 if avoiding such is what you want? You could even keep your own localized out of tree copy if >=gpg 2.0 remains a concern.

Regarding available documentation (vague RTFM comments, etc.--it's one thing to state, it's another to leave it as an ambiguous egg hunt):

DESTDIR:

• info make "Makefile Conventions" DESTDIR
• man 5 ebuild
• https://devmanual.gentoo.org/quickstart/

/var/db/pkg/*/*/CONTENTS:

• man portage

"-r bump":

• http://devmanual.gentoo.org/general-con ... revisions/ , you could easily make your own copy of a particular ebuild, and suffix -r# appropriately to change for some personal need.

If you're reasoning to your reader to do something using someone (e.g., khayyam) else as a reference, a link to specifically what you're referring to would aid understanding.

Please consider editing your OP guide down (in particular the large, and seemingly irrelevant listings, political discourse, etc.). The whys (i.e. basis) for what you're trying to present are not so important as the hows.

Try to focus in brief on technical aspects only without any ill-conceived prejudice against portage (i.e. ignorance is not bliss). Afterall, this is Gentoo and portage is there to aid, in good confidence, with playing secretary and accountant. Let the tool work for you instead of being fearful from it; read up on matters you're confused with and attempt to ask concisely on issue(s) that remain unclear. If using portage remains an issue of distrust for you, perhaps you should instead look into LFS, where you're entirely on your own and all the caveats that entails, particularly over time.

@i92guboj, thank you, would have been good to see such a response in some past threads.

@others, Please consider editing down quoted lengthy responses as a proof of reason to rant if the OP adjusts their main post in appropriate manner in the interests of improving the quality of this thread.

[Navar]

EDIT: The issue seems to be with the gpg use dependency in the mutt ebuild, not gnupg. Have you tried simply editing gpgme out in those requirements?

Code: Select all

$ USE="gpg" emerge -pvt gnupg mutt

[ebuild  N     ] mail-client/mutt-1.5.22-r3  USE="berkdb crypt gdbm gpg nls sasl ssl -debug -doc -gnutls -idn -imap -kerberos -mbox -nntp -pop -qdbm (-selinux) -sidebar -slang -smime -smtp -tokyocabinet" 3,773 kB
[ebuild  N     ]  app-crypt/gpgme-1.3.2-r1:1  USE="-common-lisp -static-libs" 0 kB
[ebuild   R    ]   app-crypt/gnupg-2.0.26  USE="bzip2 nls readline -adns -doc -ldap -mta (-selinux) -smartcard -static -usb" 0 kB

$ grep -i gpgme /usr/portage/mail-client/mutt/mutt-1.5.2*
/usr/portage/mail-client/mutt/mutt-1.5.22-r3.ebuild:	gpg?     ( >=app-crypt/gpgme-0.9.0 )
/usr/portage/mail-client/mutt/mutt-1.5.22-r3.ebuild:		$(use_enable gpg gpgme) \
/usr/portage/mail-client/mutt/mutt-1.5.23-r2.ebuild:	gpg?     ( >=app-crypt/gpgme-0.9.0 )
/usr/portage/mail-client/mutt/mutt-1.5.23-r2.ebuild:		$(use_enable gpg gpgme) \

I don't want to be forced to install gnupg-2, because it introduces GUI based gpgme.

Unclear:

Code: Select all

$ grep -i gpgme /usr/portage/app-crypt/gnupg/*.ebuild
$ equery l -p gpgme
 * Searching for gpgme ...
[-P-] [  ] app-crypt/gpgme-1.3.2-r1:1
[-P-] [ ~] app-crypt/gpgme-1.4.4:1/11
[-P-] [ ~] app-crypt/gpgme-1.5.1:1/11
$ equery d gpgme
 * These packages depend on gpgme:
$ emerge -pv gnupg mutt

[ebuild   R    ] app-crypt/gnupg-2.0.26  USE="bzip2 nls readline -adns -doc -ldap -mta (-selinux) -smartcard -static -usb" 0 kB
[ebuild  N     ] mail-client/mutt-1.5.22-r3  USE="berkdb crypt gdbm nls sasl ssl -debug -doc -gnutls -gpg -idn -imap -kerberos -mbox -nntp -pop -qdbm (-selinux) -sidebar -slang -smime -smtp -tokyocabinet" 3,773 kB

[miroR]

See my immediately next post a little further below, pls.

[xaviermiller]

Moved to Unsupported Software, since installing a software not using portage is not supported.

[miroR]

I wrote the following five or so lines some twenty, and some ten hours ago.
EDIT: The text below I might need to change yet.
[[EDIT 2: 90% of my waking hours since the time of EDIT above spent only on man ebuild (and then man bash when otherwise incomprehensible the man ebuild). A few hours more I need, at least.]]

Two more senior Gentoo Forum members suggested (the two that suggested so nicely count only), so I am going to try now and read the documentation as Navar gave straight links to.

He suggested so above in this same topic the visitor is currently rreading.

I did try to read those a few months ago already, and a few months before then and yet before probably... Patience... If I am able to figure some out in not more than a few hours of dedicated work, I won't desist. If it will require more from me, then I will have to postpone it for who knows when.

==================================
So I spoke before these maybe ten or more and twenty or so dedicated hours of studying in the suggested direction. Time for a reply. You are reading the changed text.

No, I haven't done the work, I was right, it is a weekend or two or more worth of work, and not only ten or fifteen hours.

But I was right in a few aspects, and, also, I don't regret my studying, which I will not desist from now. Only post a reply. So the two senior Gentoo Forums members, were right in their suggestions too.

man ebuild I have never yet so relatively well understood. Problem is to memorize it, actually remember all the sleuth of symbols, so it's still a constant looking it up again whenever I check some ebuild.

The Quickstart Ebuild Guide is based on explaining things on old packages that none of it is anymore in portage, so I am going, now that I studied (it's not about reading, that's called studying, not reading, in my book ) the man ebuild... So I am going for Ebuild Writing and probably first, actually, General Concepts. I'll think about the "-r bump" when I reach it in that online manual (not soon).

This is what I wrote previously:

Just one thing. I wish I could, but I don't have one or two or three weekends' worth of time to go through the ebuild manual and othe documentation, notwithstanding that it appeals to me very much...

It appeals to me, so I have to stick for longer with studying it, now.

I'll postpone other things for a another while.

EDIT: The issue seems to be with the gpg use dependency in the mutt ebuild, not gnupg. Have you tried simply editing gpgme out in those requirements?

Code: Select all

$ USE="gpg" emerge -pvt gnupg mutt

[ebuild  N     ] mail-client/mutt-1.5.22-r3  USE="berkdb crypt gdbm gpg nls sasl ssl -debug -doc -gnutls -idn -imap -kerberos -mbox -nntp -pop -qdbm (-selinux) -sidebar -slang -smime -smtp -tokyocabinet" 3,773 kB
[ebuild  N     ]  app-crypt/gpgme-1.3.2-r1:1  USE="-common-lisp -static-libs" 0 kB
[ebuild   R    ]   app-crypt/gnupg-2.0.26  USE="bzip2 nls readline -adns -doc -ldap -mta (-selinux) -smartcard -static -usb" 0 kB

$ grep -i gpgme /usr/portage/mail-client/mutt/mutt-1.5.2*
/usr/portage/mail-client/mutt/mutt-1.5.22-r3.ebuild:	gpg?     ( >=app-crypt/gpgme-0.9.0 )
/usr/portage/mail-client/mutt/mutt-1.5.22-r3.ebuild:		$(use_enable gpg gpgme) \
/usr/portage/mail-client/mutt/mutt-1.5.23-r2.ebuild:	gpg?     ( >=app-crypt/gpgme-0.9.0 )
/usr/portage/mail-client/mutt/mutt-1.5.23-r2.ebuild:		$(use_enable gpg gpgme) \

I don't want to be forced to install gnupg-2, because it introduces GUI based gpgme.

Unclear:

Code: Select all

$ grep -i gpgme /usr/portage/app-crypt/gnupg/*.ebuild
$ equery l -p gpgme
 * Searching for gpgme ...
[-P-] [  ] app-crypt/gpgme-1.3.2-r1:1
[-P-] [ ~] app-crypt/gpgme-1.4.4:1/11
[-P-] [ ~] app-crypt/gpgme-1.5.1:1/11
$ equery d gpgme
 * These packages depend on gpgme:
$ emerge -pv gnupg mutt

[ebuild   R    ] app-crypt/gnupg-2.0.26  USE="bzip2 nls readline -adns -doc -ldap -mta (-selinux) -smartcard -static -usb" 0 kB
[ebuild  N     ] mail-client/mutt-1.5.22-r3  USE="berkdb crypt gdbm nls sasl ssl -debug -doc -gnutls -gpg -idn -imap -kerberos -mbox -nntp -pop -qdbm (-selinux) -sidebar -slang -smime -smtp -tokyocabinet" 3,773 kB

I'm familiar with that, mostly. But also there is this: the gnupg-1 can work without gpg-agent, while gnupg-2 can not. Look up (you seem to have, in 'man gpg' something like '--no-use-agent' and something to the effect that it does nothing, again can not be disabled.

Whlie I in my gnupg-1 have, in my Debian (online, so I'll copy from it), or Gentoo --I'm currently working--no! I have to postopne it now because of _this_ work, to enable a clone for online (

after problems somewhat peculiar and very unclear still:
Syslog-ng from Delay Logging to BrokenPipe/no Loggin [SOLVED
http://forums.gentoo.org/viewtopic-t-1001994.html

)...

I have in both:

--use-agent

--no-use-agent
Try to use the GnuPG-Agent. With this option, GnuPG first tries to connect to the
agent before it asks for a passphrase
--no-use-agent disables this option.

And my case (no gpg-agent) is by definition safer than more layers for a program just to read a password which is the case for gnupg-2, which you have.

And think about that I got rid of dbus and kits and all poetteringware:

Uninstalling dbus and *kits (to Unfacilitate Remote Seats)
http://forums.gentoo.org/viewtopic-t-99 ... ight-.html

as well as in Debian (thanks to Thorsten mirabilos Glaser's MirDebian WTF):

How to Remove Systemd and Related Packages from Your Debian
http://forums.debian.net/viewtopic.php?f=16&t=118197

I insist that the best solution would be:
Maintainers, pls. get us the option to use plain gnupg-1 with portage's mutt!

So, let's see the post that I didn't see promptly before.

@OP, some suggestions:

app-crypt/gnupg-1.4.18 is available in the tree. Why not simply mask >app-crypt/gnupg-2.0 if avoiding such is what you want? You could even keep your own localized out of tree copy if >=gpg 2.0 remains a concern.

I did that, and it is in the link that I gave, to Air-Gapped Gentoo Install Tentative. So, it's you who haven't considered all the information that I gave, in this case.

DESTDIR:

• info make "Makefile Conventions" DESTDIR
• man 5 ebuild
• https://devmanual.gentoo.org/quickstart/

/var/db/pkg/*/*/CONTENTS:

• man portage

"-r bump":

• http://devmanual.gentoo.org/general-con ... revisions/ , you could easily make your own copy of a particular ebuild, and suffix -r# appropriately to change for some personal need.

"Easy" depends on how nimble my mind is...

If you're reasoning to your reader to do something using someone (e.g., khayyam) else as a reference, a link to specifically what you're referring to would aid understanding.

Again, it is there, it's exactly based on the information that khayyam gave about gnupg-1 and that he made it available in portage, that that section of my "Air-Gapped Gentoo Install Tentative" is about. So, you're telling me off for things that are there, and not missing like you didn't investigate to find out. But it doesn't matter.

Please consider editing your OP guide down (in particular the large, and seemingly irrelevant listings, political discourse, etc.). The whys (i.e. basis) for what you're trying to present are not so important as the hows.

No, I don't agree. The title says why use gnupg-1 and not 2. Because we're air-gappers, we don't want intrusion, and we want to minimize it this way, and that is the reason we want to do it this way.
And the one sentence to explain why I used ukrainian instead of my regular user name, that must be bearable for everyone. It's little and it's about justice. It's a nation mine is related to. I'm saying this without ill feelings.

... perhaps you should instead look into LFS

I've been an LFSer. LFS is easy in comparison. BLFS, and getting the system to work in all aspects is not.

@others, Please consider editing down quoted lengthy responses as a proof of reason to rant if the OP adjusts their main post in appropriate manner in the interests of improving the quality of this thread.

No, the Slippery Jim obviously found it right. Look it up again, and see the logic of it.

For now, this is the solution! This is a good tutorial.

It I make it (it's not certain, it's hard, lots of work!), to set the right ebuild in my own overlay and get at least as good resuls and (look up what I said about the Mutt Dude's ways above) as it's suppose to (which certainly isn't as it can be had via Portage, i.e. with gnupg-2 only), than I'll post another tutorial and point to it from the very top of this one!

But in the meantime, it really would be best if the orthodoxy (as Mutt Dude and his devs preach it) is respected and not given away for the inferior and certainly much more intrusion prone gnupg-2, along with those other layer added, as I explained correctly already. I mean:

Maintainers, pls. get us the option to use plain gnupg-1 with portage's mutt!

And surely, navar, thanks a lot for your support! !

[miroR]

I just made it to install the portage way. To some extent.

I installed Mutt with emerge, and from local modified portage.

Here's the ebuild that I used:

mutt-1.5.23-r8.ebuild:

Code: Select all

# Copyright 1999-2014 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: /var/cvsroot/gentoo-x86/mail-client/mutt/mutt-1.5.23-r2.ebuild,v 1.1 2014/04/08 20:37:42 grobian Exp $

# Modified by miroR from portage repo to accept gnupg-1 and no gpg-agent and GUI

EAPI="5"

inherit eutils flag-o-matic autotools

# No patches, plain upstream
#PATCHSET_REV="-r2"

DESCRIPTION="A small but very powerful text-based mail client"
HOMEPAGE="http://www.mutt.org/"
SRC_URI="mirror://sourceforge/${PN}/${P}.tar.gz
	https://bitbucket.org/${PN}/${PN}/downloads/${P}.tar.gz
	ftp://ftp.mutt.org/mutt/devel/${P}.tar.gz"
#	mirror://gentoo/${P}-gentoo-patches${PATCHSET_REV}.tar.bz2
#	http://dev.gentoo.org/~grobian/distfiles/${P}-gentoo-patches${PATCHSET_REV}.tar.bz2"
IUSE="berkdb crypt debug doc gdbm gnutls gpg idn imap kerberos mbox nls nntp pop qdbm sasl selinux sidebar slang smime smtp ssl tokyocabinet"
SLOT="0"
LICENSE="GPL-2"
KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~mips ~ppc ~ppc64 ~sparc ~x86 ~x86-fbsd ~x64-freebsd ~x86-freebsd ~x86-interix ~amd64-linux ~x86-linux ~ppc-macos ~x64-macos ~x86-macos ~sparc-solaris ~sparc64-solaris ~x64-solaris ~x86-solaris"
RDEPEND="
	app-misc/mime-types
	nls? ( virtual/libintl )
	tokyocabinet?  ( dev-db/tokyocabinet )
	!tokyocabinet? (
		qdbm?  ( dev-db/qdbm )
		!qdbm? (
			gdbm?  ( sys-libs/gdbm )
			!gdbm? ( berkdb? ( >=sys-libs/db-4 ) )
		)
	)
	imap?    (
		gnutls?  ( >=net-libs/gnutls-1.0.17 )
		!gnutls? ( ssl? ( >=dev-libs/openssl-0.9.6 ) )
		sasl?    ( >=dev-libs/cyrus-sasl-2 )
	)
	kerberos? ( virtual/krb5 )
	pop?     (
		gnutls?  ( >=net-libs/gnutls-1.0.17 )
		!gnutls? ( ssl? ( >=dev-libs/openssl-0.9.6 ) )
		sasl?    ( >=dev-libs/cyrus-sasl-2 )
	)
	smtp?     (
		gnutls?  ( >=net-libs/gnutls-1.0.17 )
		!gnutls? ( ssl? ( >=dev-libs/openssl-0.9.6 ) )
		sasl?    ( >=dev-libs/cyrus-sasl-2 )
	)
	idn?     ( net-dns/libidn )
	gpg?     ( >=app-crypt/gnupg-1.4.18 )
	smime?   ( >=dev-libs/openssl-0.9.6 )
	selinux? ( sec-policy/selinux-mutt )
	slang? ( sys-libs/slang )
	!slang? ( >=sys-libs/ncurses-5.2 )
"
DEPEND="${RDEPEND}
	net-mail/mailbase
	doc? (
		dev-libs/libxml2
		dev-libs/libxslt
		app-text/docbook-xsl-stylesheets
		|| ( www-client/lynx www-client/w3m www-client/elinks )
	)"

# No patches, plain upstream
#PATCHDIR="${WORKDIR}"/${P}-gentoo-patches${PATCHSET_REV}

src_prepare() {
# 56 lines cut out here, all commented, all patching
	# don't just build documentation (lengthy process, with big dependencies)
	if use !doc ; then
#		sed -i -e '/SUBDIRS =/s/doc//' Makefile.in || die
	echo "I don't think src_prepare() can be empty."
	fi
}

src_configure() {
	local myconf="
		$(use_enable crypt pgp) \
		$(use_enable debug) \
		$(use_enable gpg) \
		$(use_enable imap) \
		$(use_enable nls) \
		$(use_enable nntp) \
		$(use_enable pop) \
		$(use_enable smime) \
		$(use_enable smtp) \
		$(use_with idn) \
		$(use_with kerberos gss) \
		$(use slang && echo --with-slang) \
		--enable-compressed \
		--enable-external-dotlock \
		--enable-mixmaster \
		--enable-nfs-fix \
		--sysconfdir=${EPREFIX}/etc/${PN} \
		--with-curses \
		--with-docdir=${EPREFIX}/usr/share/doc/${PN}-${PVR} \
		--with-regex \
		--with-exec-shell=${EPREFIX}/bin/sh"

	case $CHOST in
		*-solaris*)
			# Solaris has no flock in the standard headers
			myconf="${myconf} --enable-fcntl --disable-flock"
		;;
		*)
			myconf="${myconf} --disable-fcntl --enable-flock"
		;;
	esac

	# mutt prioritizes gdbm over bdb, so we will too.
	# hcache feature requires at least one database is in USE.
	if use tokyocabinet; then
		myconf="${myconf} --enable-hcache \
			--with-tokyocabinet --without-qdbm --without-gdbm --without-bdb"
	elif use qdbm; then
		myconf="${myconf} --enable-hcache \
			--without-tokyocabinet --with-qdbm --without-gdbm --without-bdb"
	elif use gdbm ; then
		myconf="${myconf} --enable-hcache \
			--without-tokyocabinet --without-qdbm --with-gdbm --without-bdb"
	elif use berkdb; then
		myconf="${myconf} --enable-hcache \
			--without-tokyocabinet --without-qdbm --without-gdbm --with-bdb"
	else
		myconf="${myconf} --disable-hcache \
			--without-tokyocabinet --without-qdbm --without-gdbm --without-bdb"
	fi

	# there's no need for gnutls, ssl or sasl without socket support
	if use pop || use imap || use smtp ; then
		if use gnutls; then
			myconf="${myconf} --with-gnutls"
		elif use ssl; then
			myconf="${myconf} --with-ssl"
		fi
		# not sure if this should be mutually exclusive with the other two
		myconf="${myconf} $(use_with sasl)"
	else
		myconf="${myconf} --without-gnutls --without-ssl --without-sasl"
	fi

	if use mbox; then
		myconf="${myconf} --with-mailpath=${EPREFIX}/var/spool/mail"
	else
		myconf="${myconf} --with-homespool=Maildir"
	fi

	econf ${myconf} || die "configure failed"
}

src_install() {
	emake DESTDIR="${D}" install || die "install failed"
	if use mbox; then
		insinto /etc/mutt
		newins "${FILESDIR}"/Muttrc.mbox Muttrc
	else
		insinto /etc/mutt
		doins "${FILESDIR}"/Muttrc
	fi

# really plain upstream first
#	# A newer file is provided by app-misc/mime-types. So we link it.
#	rm "${ED}"/etc/${PN}/mime.types
#	dosym /etc/mime.types /etc/${PN}/mime.types

	# A man-page is always handy, so fake one
	if use !doc; then
		emake -C doc DESTDIR="${D}" muttrc.man || die
		# make the fake slightly better, bug #413405
		sed -e 's#@docdir@/manual.txt#http://www.mutt.org/doc/devel/manual.html#' \
			-e 's#in @docdir@,#at http://www.mutt.org/,#' \
			-e "s#@sysconfdir@#${EPREFIX}/etc/${PN}#" \
			-e "s#@bindir@#${EPREFIX}/usr/bin#" \
			doc/mutt.man > mutt.1
		cp doc/muttbug.man flea.1
		cp doc/muttrc.man muttrc.5
		doman mutt.1 flea.1 muttrc.5
# really plain upstream first
#	else
#		# nuke manpages that should be provided by an MTA, bug #177605
#		rm "${ED}"/usr/share/man/man5/{mbox,mmdf}.5 \
#			|| ewarn "failed to remove files, please file a bug"
	fi

	if use !prefix ; then
		fowners root:mail /usr/bin/mutt_dotlock
		fperms g+s /usr/bin/mutt_dotlock
	fi

	dodoc BEWARE COPYRIGHT ChangeLog NEWS OPS* PATCHES README* TODO VERSION
}

pkg_postinst() {
	echo
	elog "If you are new to mutt you may want to take a look at"
	elog "the Gentoo QuickStart Guide to Mutt E-Mail:"
	elog "   http://www.gentoo.org/doc/en/guide-to-mutt.xml"
	echo
}

It's been two and a half days of which most of my waking hours I have dedicated to studying this issue.

For the newbies, briefly and incompletely: You need that ebuild, you need to set it in your local overlay (and that is some studying to get the know-how!), and you only need in your local overlay:

Code: Select all

mkdir files/

and copy just Muttrc and Muttrc.mbox from the $PORTDIR.

And, just to say that the error is still there, it looks the same, ths one:
After doing a regular "emerge mutt", and trying to use it, I got:

Code: Select all

ukrainian@mybox ~ $ mutt
Error in /home/ukrainian/.muttrc, line 11: ssl_usesystemcerts: unknown variable
source: errors in /home/ukrainian/.muttrc
Press any key to continue...
ukrainian@mybox ~ $

(where ukrainian I am not, but I didn't feel like putting my real user name), and that error is probably caused by something in the Gentoo-customization over the vanilla Mutt.

But: I can use gnupg-1 just fine, and also can login to my dovecot.

Wee hours here, so I'd just like to thank esp. Navar again, who may have not read the link I gave, and repeated the same advice that was already in the link given, but thanks to whom I've made it, because his gentle advice was crucial to my going this way. (because that made two senior Gentoo Forums members with the then likely to-be-regarded-as right advice).

I don't know if I can explain more in detail, but then it needs to be done in a separate topic...

But it's wee hours in Europe, so: Good Night!

[miroR]

@OP, some suggestions:
...[snip]...
@others, Please consider editing down quoted lengthy responses as a proof of reason to rant if the OP adjusts their main post in appropriate manner in the interests of improving the quality of this thread.

I'll leave it for another day or so, so that it can be clear, that Navar was right in part.
In the first post, I let the:

Code: Select all

/usr/src/mutt-1.5.23/...
...[snip]...

lots of those unfiltered out. Only those I need to remove. Sorry about that.
EDIT: 2014-10-29 later in the afternoon. Fixed.

[miroR]

This wiki page:

Overlay/Local overlay
http://wiki.gentoo.org/wiki/Overlay/Local_overlay

made me lose a lot of time.

Because it's, in some details, wrong for the versions of layman:

(for layman 1.3 and later)
# echo "source /var/lib/layman/make.conf" >> /etc/portage/make.conf

as you can find in:
Gentoo Overlays: Users' Guide
http://www.gentoo.org/proj/en/overlays/userguide.xml

I'd like to, if I find more endurance, correct that with my ebuild. Namely I used strictly only local overlay, just that single category/package: mail-client/mutt, have all in history and in logs, I think I can explain.

But: I need to sort that Mutt ebuild above to be a little nicer. I don't understand yet a few things...

Thanks again for the support, Navar and steveL!

[steveL]

Good work mir0r; file a bug with the ebuild attached.

If you want to get in touch with the maintainers, that's the only approach. They don't read the forums looking for things to correct; that's what bugzilla is for.

They can help you with the ebuild, it doesn't have to be perfect. Just keep it terse (ie: brief) on the bugzilla; it's not a discussion forum, so only talk about the specific issues.

[miroR]

Good work mir0r; file a bug with the ebuild attached.
....[snip]...
...get in touch with the maintainers, that's the only approach
....[snip]...

Thanks! I second your proposition . I'll be acted upon.

[miroR]

I filed:

Bug 527338 - mail-client/mutt - add support for app-crypt/gnupg as alternative to app-crypt/gpgme
https://bugs.gentoo.org/show_bug.cgi?id=527338

It seems to have been noticed fine and that it is being worked on.

Admins, can this topic be put back to:
Documentation, Tips & Tricks

now that even a bug is based on it, with huge amount of my work, and rather obedient (and expensive) following of good senior advice ?

[steveL]

I filed [bug=527338]Bug 527338[/bug]

Good. Note the link above ("Bug" button)

It seems to have been noticed fine and that it is being worked on.

*sigh* you are new ;) It's being processed, not worked on, since all that's happened is that a bug-wrangler has assigned it to
the correct team. So yes, they'll see it soon enough. When it will get corrected is another matter.

now that even a bug is based on it, with huge amount of my work, and rather obedient (and expensive) following of advice?

Trust me, you've done very little so far, compared to what Gentoo users do on a daily or weekly basis.

If it was expensive, I presume you mean your time. You could save us all a lot of that with shorter posts, yourself most of all.

And in passing, our time ain't free either.