re-alpine-devel — Development discussion

[Re-alpine-devel] Dependencies for re-alpine

[D.J.J. R. Jr. <n1...@ar...>]

Hello,

I'm running Slackware, actually a wonderful version called "Slint"
which is Slackware International and also for accessible for vision
impaired and blind people.

We have some people in our group that have brain damage and cannot
manage the "long password" that now protects alpine, and we have no
packaged re-alpine available.  The Distro leader (one man operation)
would probably package it if I could tell him the dependencies.

I tried to configure the latest re-alpine 2.03.

I got this error when I tried to ./configure re-alpinie:
/usr/bin/rm: cannot remove 'libtoolT': No such file or directory

When I run make as user, it runs but I start getting mentions of errors here:

libtool: link: gcc -shared  .libs/wpcomm.o   -L/usr/lib -ltcl
/usr/lib64/libldap.so -L/usr/lib64 /usr/lib64/libsasl2.so -ldl
-lcrypto -lncurses /usr/lib64/liblber.so -lresolv -lssl  -pthread
-pthread -Wl,-soname -Wl,libwpcomm.so.1 -o .libs/libwpcomm.so.1.0.0
libtool: link: (cd ".libs" && rm -f "libwpcomm.so.1" && ln -s
"libwpcomm.so.1.0.0" "libwpcomm.so.1")
libtool: link: (cd ".libs" && rm -f "libwpcomm.so" && ln -s
"libwpcomm.so.1.0.0" "libwpcomm.so")
libtool: link: /usr/bin/ar cru .libs/libwpcomm.a  wpcomm.o
libtool: link: ranlib .libs/libwpcomm.a
libtool: link: ( cd ".libs" && rm -f "libwpcomm.la" && ln -s
"../libwpcomm.la" "libwpcomm.la" )
gcc -DHAVE_CONFIG_H   -I../../../include -I../../../include
-I/usr/include  -g -pthread -g -O2 -MT alpined.o -MD -MP -MF
.deps/alpined.Tpo -c -o alpined.o alpined.c
alpined.c: In function ‘main’:
alpined.c:763:48: error: ‘Tcl_Interp {aka struct Tcl_Interp}’ has no
member named ‘result’
       case TCL_OK   : peReturn(cs, "OK", interp->result); break;
                                                ^
alpined.c:764:53: error: ‘Tcl_Interp {aka struct Tcl_Interp}’ has no
member named ‘result’
       case TCL_ERROR  : peReturn(cs, "ERROR", interp->result); break;
                                                     ^
alpined.c:765:53: error: ‘Tcl_Interp {aka struct Tcl_Interp}’ has no
member named ‘result’
       case TCL_BREAK  : peReturn(cs, "BREAK", interp->result); break;
                                                     ^
alpined.c:766:54: error: ‘Tcl_Interp {aka struct Tcl_Interp}’ has no
member named ‘result’
       case TCL_RETURN : peReturn(cs, "RETURN", interp->result); break;
                                                      ^
alpined.c: In function ‘peLoadConfig’:
alpined.c:10847:45: warning: cast to pointer from integer of different
size [-Wint-to-pointer-cast]
       mail_parameters(NULL, SET_RSHTIMEOUT, (void *) rv);
                                             ^
alpined.c:10851:45: warning: cast to pointer from integer of different
size [-Wint-to-pointer-cast]
       mail_parameters(NULL, SET_SSHTIMEOUT, (void *) rv);
                                             ^
alpined.c: In function ‘peRssFetch’:
alpined.c:16275:21: warning: implicit declaration of function
‘date_to_local_time_t’ [-Wimplicit-function-declaration]
         theirdate = date_to_local_time_t(q);
                     ^
Makefile:457: recipe for target 'alpined.o' failed
make[2]: *** [alpined.o] Error 1
make[2]: Leaving directory
'/home/djringjr/Downloads/re-alpine-2.03/web/src/alpined.d'
Makefile:421: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/home/djringjr/Downloads/re-alpine-2.03'
Makefile:312: recipe for target 'all' failed
make: *** [all] Error 2

If anyone can help me with this error, I'd be very happy.

If anyone is using slackware and who would volunteer making a version
of re-alpine with --with-passfile=.pine-passfile it would help us
because it seems alpine is the most friendly for people using
espeak-ng and speech-dispatcher screen reader in console.

Blind people like console because there's no mouse to find.

Regards,
David

Re: [Re-alpine-devel] [Alpine-info] install Alpine on MacBook Pro

[Matt A. <ma...@ap...>]

On Wed, 24 Jan 2018, Jacob Wegelin wrote:
> The latest Apple Macintosh upgrade, to macOS Sierra 10.12.6, has made alpine
> much slower to launch. This has nothing to do with access to the internet. Ten
> seconds are required simply before the software launches.

what do you mean before it launches? Did you run it under lldb? Has it reached main() yet?

I'm not denying it's slower, but I *highly* doubt it hasn't actually launched the
binary yet.

I had problems long ago where for some reason 1.x would start up way faster than
newer versions for me. It seemed to be an alpine config issue for me, I never
was able to narrow it down.  At some point I went back to a newer version
with the same pine config file (and reading a remote config for my acct info)
and it got fast again.

> I have used fink before; is that still the way to install alpine?

You could recompile it from source, but there are various things you have
to tweak to even get that to work. (I think Eduardo had some patches that
fixes at least one of the issues.)

[Re-alpine-devel] install Alpine on MacBook Pro

[Jacob W. <jac...@fa...>]

Is there a best-practices way to install Alpine on a MacBook Pro?

I run alpine at the command line (no GUI) in the so-called Terminal application, which provides a bash shell. I also occasionally use the iTerm application, which also gives a bash shell.

The latest Apple Macintosh upgrade, to macOS Sierra 10.12.6, has made alpine much slower to launch. This has nothing to do with access to the internet. Ten seconds are required simply before the software launches.

By contrast: before I "upgraded" from the previous macOS version, launch was instantaneous.

Thus, perhaps I need to re-install alpine in a way that is compatible with the new version.

How does one best do this?

I have used fink before; is that still the way to install alpine?

Is there any reason to believe that re-installing will make it launch more quickly?

Thank you

Jacob Wegelin

[Re-alpine-devel] Message-ID when posting news

[Vitus J. <vj...@gm...>]

Hello guys,

Every once in a while I'm trying to access newsgroups with alpine; I do 
have an account with individual.net and feel I should show up so they 
don't cancel the server alltogether.

However, since several years I can't post using alpine.  The issue is the 
message-id: my articels are rejected with "441 435 Bad Message-ID".  I'm 
not that common with NNTP but from what I read it seems the creation of 
the message-id may be relayed to the server.  If the client provides a 
Message-ID individual.net's FAQ says:

   I ... would like to generate the Message-IDs by myself. Can I get a FQDN
   (Fully Qualified Domain Name) from your name space?

   Sure. The FQDN is a combination of the string "ID-", your User ID, and
   the subdomain "user.individual.net" ...

   (See http://news.individual.net/faq.php#4.4 )


Alpine uses my machine's FQDN in the message-id.  How do I either change 
that (for news only) to the above string or remove the message-id 
alltogether?

I couldn't find a solution any of those options.

Best regards,
    Vitus

Re: [Re-alpine-devel] [PATCH] Add config option for an M$ Outlook friendly encryption algo

[David N. <dav...@gm...>]

Hi,

This was affecting far later versions than that.
It was a hang/crash, not an error message.
Changing cipher seems to solve the issue.

Best Regards,
David

On Tue, Feb 16, 2016 at 5:44 AM, <ja...@us...> wrote:

> Indeed this is a problem with unsupported versions of Outlook.
>
> https://support.microsoft.com/en-us/kb/2710636
>

Re: [Re-alpine-devel] [PATCH] Add config option for an M$ Outlook friendly encryption algo

[<ja...@us...>]

Indeed this is a problem with unsupported versions of Outlook.

https://support.microsoft.com/en-us/kb/2710636

[Re-alpine-devel] [PATCH] Add config option for an M$ Outlook friendly encryption algo

[David N. <dav...@gm...>]

In some versions of M$ Outlook, it seems the default encryption
algorithm will cause the receiving M$ Outlook client to hang/crash.
Adds a config option for people unfortunate enough to work in
such a situation.

Signed-off-by: David Nyström <dav...@gm...>
---
 alpine/smime.c                | 11 +++++++++++
 doc/tech-notes/config.html    | 16 ++++++++++++++++
 doc/tech-notes/tech-notes.txt | 11 +++++++++++
 pith/conf.c                   |  2 ++
 pith/conftype.h               |  1 +
 pith/pine.hlp                 | 30 ++++++++++++++++++++++++++++++
 pith/smime.c                  | 11 +++++++++--
 7 files changed, 80 insertions(+), 2 deletions(-)

diff --git a/alpine/smime.c b/alpine/smime.c
index 4cfdeae..24127d8 100644
--- a/alpine/smime.c
+++ b/alpine/smime.c
@@ -873,6 +873,17 @@ smime_config_init_display(struct pine *ps, CONF_S **ctmp, CONF_S **first_line)
     (*ctmp)->varmem		= ind;
     (*ctmp)->value		= pretty_value(ps, (*ctmp));
 
+    ind = feature_list_index(F_CIPHER_DES_EDE3_CBC);
+    feature = feature_list(ind);
+    new_confline(ctmp)->var 	= vtmp;
+    (*ctmp)->varnamep		= ctmpb;
+    (*ctmp)->keymenu		= &config_checkbox_keymenu;
+    (*ctmp)->help		= config_help(vtmp-ps->vars, feature->id);
+    (*ctmp)->tool		= checkbox_tool;
+    (*ctmp)->valoffset		= feature_indent();
+    (*ctmp)->varmem		= ind;
+    (*ctmp)->value		= pretty_value(ps, (*ctmp));
+
     ind = feature_list_index(F_USE_CERT_STORE_ONLY);
     feature = feature_list(ind);
     new_confline(ctmp)->var 	= vtmp;
diff --git a/doc/tech-notes/config.html b/doc/tech-notes/config.html
index b935483..80e1078 100644
--- a/doc/tech-notes/config.html
+++ b/doc/tech-notes/config.html
@@ -7611,6 +7611,22 @@ certificate).
 This feature is displayed as &quot;S/MIME -- Sign by Default&quot;.
 <P>
 
+<DT> <A NAME="smime-encrypt-des-ede3-cbc"><EM>smime-encrypt-des-ede3-cbc</EM></A>
+
+<DD> UNIX <EM>Alpine</EM> only.
+<P>
+This feature only has an effect if your version of <EM>Alpine</EM> includes
+support for S/MIME.
+It affects <EM>Alpine</EM>'s behavior when you send a message.
+It changes the default encryption algorithm to DES-EDE3-CBC since it seems to
+be able to crash at least some versions of M$ Outlook.
+<P>
+<UL>
+<LI><A HREF="config-notes.html#smime-general">General S/MIME Overview</A>
+</UL><P>
+This feature is displayed as &quot;S/MIME -- Encrypt with DES-EDE3-CBC&quot;.
+<P>
+
 <DT> <A NAME="sort-default-fcc-alpha"><EM>sort-default-fcc-alpha</EM></A>
 
 <DD> This feature controls an aspect of <EM>Alpine</EM>'s FOLDER LIST screen.
diff --git a/doc/tech-notes/tech-notes.txt b/doc/tech-notes/tech-notes.txt
index e1eb4b9..bca7dcd 100644
--- a/doc/tech-notes/tech-notes.txt
+++ b/doc/tech-notes/tech-notes.txt
@@ -6570,6 +6570,17 @@ John: on a job well done!
           certificate).
           + General S/MIME Overview
           This feature is displayed as "S/MIME -- Sign by Default".
+   _smime-encrypt-des-ede3-cbc_
+          UNIX _Alpine_ only.
+          This feature only has an effect if your version of _Alpine_
+          includes support for S/MIME. It affects _Alpine_'s behavior when
+          you send a message.
+	  If this option is set, the "Sign" option
+          will change the default encryption algorithm to DES-EDE3-CBC since
+	  some version out M$ Outlook seems to crash when receiving signed and
+	  encrypted mail using the default algorithm.
+          + General S/MIME Overview
+          This feature is displayed as "S/MIME -- Encrypt with DES-EDE3-CBC".
    _sort-default-fcc-alpha_
           This feature controls an aspect of _Alpine_'s FOLDER LIST
           screen. If set, the default FCC folder will be sorted
diff --git a/pith/conf.c b/pith/conf.c
index 6a80393..3a003a6 100644
--- a/pith/conf.c
+++ b/pith/conf.c
@@ -3290,6 +3290,8 @@ feature_list(int index)
 	 F_REMEMBER_SMIME_PASSPHRASE, h_config_smime_remember_passphrase, PREF_HIDDEN, 0},
 	{"smime-sign-by-default", "S/MIME -- Sign by Default",
 	 F_SIGN_DEFAULT_ON, h_config_smime_sign_by_default, PREF_HIDDEN, 0},
+	{"smime-encrypt-des-ede3-cbc", "S/MIME -- Encrypt with DES-EDE3-CBC",
+	 F_CIPHER_DES_EDE3_CBC, h_config_smime_encrypt_des_ede3_cbc, PREF_HIDDEN, 0},
 	{"smime-use-store-only", "S/MIME -- Validate Using Certificate Store Only",
 	 F_USE_CERT_STORE_ONLY, h_config_smime_use_cert_store, PREF_HIDDEN, 1},
 #ifdef APPLEKEYCHAIN
diff --git a/pith/conftype.h b/pith/conftype.h
index abae2b0..df3595a 100644
--- a/pith/conftype.h
+++ b/pith/conftype.h
@@ -549,6 +549,7 @@ typedef enum {
 	F_ENCRYPT_DEFAULT_ON,
 	F_REMEMBER_SMIME_PASSPHRASE,
 	F_USE_CERT_STORE_ONLY,
+	F_CIPHER_DES_EDE3_CBC,
 #ifdef	APPLEKEYCHAIN
 	F_PUBLICCERTS_IN_KEYCHAIN,
 #endif
diff --git a/pith/pine.hlp b/pith/pine.hlp
index de174d2..e65adca 100644
--- a/pith/pine.hlp
+++ b/pith/pine.hlp
@@ -4205,6 +4205,7 @@ There are also additional details on
 <li><a href="h_config_smime_encrypt_by_default">S/MIME FEATURE: <!--#echo var="FEAT_smime-encrypt-by-default"--></a>
 <li><a href="h_config_smime_remember_passphrase">S/MIME FEATURE: <!--#echo var="FEAT_smime-remember-passphrase"--></a>
 <li><a href="h_config_smime_sign_by_default">S/MIME FEATURE: <!--#echo var="FEAT_smime-sign-by-default"--></a>
+<li><a href="h_config_smime_encrypt_des_ede3_cbc">S/MIME FEATURE: <!--#echo var="FEAT_smime-encrypt-des-ede3-cbc"--></a>
 <li><a href="h_config_smime_use_cert_store">S/MIME FEATURE: <!--#echo var="FEAT_smime-use-store-only"--></a>
 <li><a href="h_config_smime_pubcerts_in_keychain">S/MIME FEATURE: <!--#echo var="FEAT_publiccerts-in-keychain"--></a>
 <li><a href="h_config_smime_cacertcon">S/MIME OPTION: <!--#echo var="VAR_smime-cacert-container"--></a>
@@ -34940,6 +34941,35 @@ certificate).
 &lt;End of help on this topic&gt;
 </BODY>
 </HTML>
+========== h_config_smime_encrypt_des_ede3_cbc ==========
+<HTML>
+<HEAD>
+<TITLE>S/MIME FEATURE: <!--#echo var="FEAT_smime-encrypt-des-ede3-cbc"--></TITLE>
+</HEAD>
+<BODY>
+<H1>S/MIME FEATURE: <!--#echo var="FEAT_smime-encrypt-des-ede3-cbc"--></H1>
+
+UNIX Alpine only.
+<P>
+This feature only has an effect if your version of Alpine includes
+support for S/MIME.
+It affects Alpine's behavior when you send a message.
+If this option is set, the &quot;Sign&quot; option will default to ON when sending messages.
+<P>
+Encrypt using DES-EDE3-CBC instead of the default algorith, since it in some deployments
+seem to crash M$ Outlook when opening encrypted and signed emails.
+<P>
+<UL>
+<LI><A HREF="h_mainhelp_smime">General S/MIME help</A>
+</UL><P>
+<P>
+
+<UL>   
+<LI><A HREF="h_finding_help">Finding more information and requesting help</A>
+</UL><P>
+&lt;End of help on this topic&gt;
+</BODY>
+</HTML>
 ========== h_config_smime_use_cert_store ==========
 <HTML>
 <HEAD>
diff --git a/pith/smime.c b/pith/smime.c
index 6eacf74..c9b7281 100644
--- a/pith/smime.c
+++ b/pith/smime.c
@@ -2250,7 +2250,11 @@ encrypt_file(char *fp, char *text, PERSONAL_CERT *pc)
   if(pc == NULL)
     return 0;
 
-  cipher = EVP_aes_256_cbc();
+  if(F_ON(F_CIPHER_DES_EDE3_CBC, ps_global))
+    cipher = EVP_des_ede3_cbc();
+  else
+    cipher = EVP_aes_256_cbc();
+
   encerts = sk_X509_new_null();
 
   sk_X509_push(encerts, X509_dup(pc->cert));
@@ -2307,7 +2311,10 @@ encrypt_outgoing_message(METAENV *header, BODY **bodyP)
     dprint((9, "encrypt_outgoing_message()"));
     smime_init();
 
-    cipher = EVP_aes_256_cbc();
+    if(F_ON(F_CIPHER_DES_EDE3_CBC, ps_global))
+      cipher = EVP_des_ede3_cbc();
+    else
+      cipher = EVP_aes_256_cbc();
 
     encerts = sk_X509_new_null();
 
-- 
2.5.0

[Re-alpine-devel] configure: crypto error

[Jacob W. <jac...@fa...>]

I am working in the bash shell under Mac OS 10.9.5.

I have ALPINE 2.01(1266) installed and am trying to install the
latest version, 2.20, downloaded from http://patches.freeiz.com/alpine/.

Based on emails on this list from 7 November 2014 (Subject: "new
SSL error", I understand that I should have an updated openssl
working. My first question is whether the way I updated openssl was
correct.

I downloaded openssl-1.0.1j, untarred it, cd-ed into the directory
created by the "untar", and did

./config

make

make test

sudo make install

Subsequently, though, things were not updated:

/Volumes/spin/SoftwareDownload/OpenSSL/openssl-1.0.1j> openssl version
OpenSSL 0.9.8za 5 Jun 2014
/Volumes/spin/SoftwareDownload/OpenSSL/openssl-1.0.1j> which openssl
/usr/bin/openssl

My public path is

echo $PATH
/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/X11/bin:/usr/texbin:/Users/qqt/bin

My solution was to create a symlink to a place in my public path:

/usr/bin> sudo ln -s /usr/local/ssl/bin/openssl openssl
/usr/bin> ls -latT openssl
lrwxr-xr-x  1 root  wheel  26 Jan 28 14:58:21 2015 openssl@ -> /usr/local/ssl/bin/openssl
/usr/bin> openssl version
OpenSSL 1.0.1j 15 Oct 2014

Is this a kluge? Is there a better way to do it?  Or at least is this sufficient to get alpine to work properly?

INSTALL mentions code such as

./config --prefix=/usr/local --openssldir=/usr/local/openssl

but a colleague said that specifying openssldir would not solve the problem.

Subsequently, at any rate, my attempt to install the newest alpine failed on the first command,

./configure

Here are the last few lines of output. Note that openssl appears to have
been properly updated.

checking Openssl library version >= 1.0.1c... yes
checking for library containing TLSv1_2_client_method... no
checking for library containing SSLeay... no
configure: error: crypto library NOT found

How do I solve the crypto or SSLeay error?

Thanks

Jacob Wegelin

Re: [Re-alpine-devel] Possible problem with recent "Reliability fixes" commit

[Stéphane G. <ste...@wa...>]

Hello,

On Sun, 25 Jan 2015, ja...@us... wrote:
> Thanks for the input. You are correct, [470972] introduced a bug as you 
> described and it has been reverted in [51a66f].

   All right, then I can continue my nitpicking :-) :  your new patch drew 
my attention to these lines:

printf("%5d: Average messages in inbox (%ld/%d)\n",
 	inboxes>0?inbox_mess/inboxes:0, inbox_mess, inboxes); 
printf("%5d: Largest inbox in messages\n", inbox_mess_max);

   It is not very important, but all fields are /long/, so all 4 of them 
should have "%ld" format instead of "%d", not just one of them.

   (One may also wonder why /inboxes/, which is by construction less than 
or equal to /sofar/, has type /long/, when /sofar/ has type /int/...)


   And, even less important: in the same main function, it would be cleaner 
for the variable /now/ to have the type /time_t/ (instead of /long/).



Goodbye,
   Stéphane.

Re: [Re-alpine-devel] Possible problem with recent "Reliability fixes" commit

[<ja...@us...>]

Stéphane,

Thanks for the input. You are correct, [470972] introduced a bug as you 
described and it has been reverted in [51a66f].

jamesaj

[Re-alpine-devel] Possible problem with recent "Reliability fixes" commit

[Stéphane G. <ste...@wa...>]

Hello,


   I think a bug was possibly introduced in the recent commit 470972, named 
"Reliability fixes", by James Jerkins.

   In the file "pine-use.c", a /strcmp/ is replaced with a /strncmp/.
The problem is that it does not test the same thing.

   The older version
> if(strcmp(pw->pw_dir, "/") == 0)
>   continue;

   is equivalent to:  if(pw->pw_dir[0]=='/'&&pw->pw_dir[1]=='\0') continue;
which means only the string "/" will match.

   The new version
> if(strncmp(pw->pw_dir, "/", 1) == 0)
>   continue;

   is equivalent to:  if(pw->pw_dir[0]=='/') continue;
which means any string starting with '/' will match.


    I have not tested anything, I do not know if the real input can trigger 
a bug (but I guess it is likely because pw_dir is very likely to be a 
string starting with '/' and not ending there, such as "/home/joe"), I did 
not try to understand the goal of the function; I just browsed this random 
commit and it hit me that the new version was not the same as the original 
one.


Goodbye,
   Stéphane Goujet.

Re: [Re-alpine-devel] [Alpine-info] new SSL error

[Jacob W. <jac...@fa...>]

On 2014-11-07 Fri 14:02, Andrew Morgan wrote:
> On Fri, 7 Nov 2014, Jacob Wegelin wrote:
>
>> 
>> A couple days ago, my institution (vcu.edu), which uses Lotus Notes as its 
>> mail server, apparently disabled the current SSL they were using (v.3) and 
>> switched to some kind of patch which they say is version 2.
>> 
>> Now I am unable to get into my vcu inbox.
>> 
>> Alpine works fine with my fastmail.fm inbox.
>> 
>> The IT people here point the finger at my client (Alpine), saying that, 
>> with proper upgrades, other secure IMAP clients work with the patched SSL. 
>> They named "Mac Mail" as an example.
>> 
>> I am running ALPINE 2.01(1266).
>> 
>> Is there a solution for this?
>> 
>> And if I need to reinstall Alpine on my Mac (OS 10.7.5), can you remind me 
>> how to do install?
>
> Alpine is going to use your system's encryption library, which is usually 
> OpenSSL.  Maybe you have a very old version of OpenSSL installed?

~/Documents> which openSSL
/usr/bin/openSSL
~/Documents> man openSSL
No manual entry for openSSL
~/Documents> openSSL
OpenSSL> 
OpenSSL> version
OpenSSL 0.9.8za 5 Jun 2014
OpenSSL> exit

Since the Mac Mini computer was purchased in December 2011, it appears
that OpenSSL has somehow been updated since then.

Is this the kind of thing that Mac automatically updates? Would an
OpenSSL update fix the IMAP connection problem?

>
> I thought I would see what protocols and ciphers your IMAP server supports. 
> I googled around for the VCU instructions.  However, they want me to login to 
> https://beech.vcu.edu/serverfinder to fetch the IMAP server name.

The imap servers that I know of are aspen4.vcu.edu and oak4.vcu.edu.
They have been listed in that way in my alpine .pinrc since around
January 2008.  Until a few days ago, I regularly checked vcu email
through those servers, with the lines

{aspen4.vcu.edu/ssl/novalidate-cert/user=jwegelin}inbox
{oak4.vcu.edu/ssl/novalidate-cert/user=jwegelin}inbox

> Hilariously, I cannot connect to that website because I have disabled SSLv3 
> in my browser!  Scanning that website reveals that it only supports SSLv3 and 
> the list of ciphers are all weak ciphers that should no longer be used.  :)

I would not be surprised if the website's information was out of
date.  In the past circa 48 hours, Collaboration Services, which
is VCU's office for dealing with the Lotus Notes client, shut down
SSLv3 because of a published vulnerability.  This morning, someone
from Collaboration Services told me that they now have a patched
SSL, which he said was version 2.

Jake

[Re-alpine-devel] new SSL error

[Jacob W. <jac...@fa...>]

A couple days ago, my institution (vcu.edu), which uses Lotus Notes as its mail server, apparently disabled the current SSL they were using (v.3) and switched to some kind of patch which they say is version 2.

Now I am unable to get into my vcu inbox.

Alpine works fine with my fastmail.fm inbox.

The IT people here point the finger at my client (Alpine), saying that, with proper upgrades, other secure IMAP clients work with the patched SSL. They named "Mac Mail" as an example.

I am running ALPINE 2.01(1266).

Is there a solution for this?

And if I need to reinstall Alpine on my Mac (OS 10.7.5), can you remind me how to do install?

Thanks

Jacob A. Wegelin
Assistant Professor
Department of Biostatistics
Virginia Commonwealth University
830 E. Main St., Seventh Floor
P. O. Box 980032
Richmond VA 23298-0032
U.S.A. 
CTSA grant: UL1TR000058
E-mail: jac...@fa... 
URL: http://www.people.vcu.edu/~jwegelin

Re: [Re-alpine-devel] re-alpine 2.03 patches

[Eduardo C. <ec...@gm...>]

On Sun, 20 Apr 2014, James Jerkins wrote:


> --- ./alpine/pine-use.c.orig	Sat Apr 19 14:36:00 2014
> +++ ./alpine/pine-use.c	Sat Apr 19 14:45:36 2014

ALl of these hunks are already in version 2.19.12.

> --- ./alpine/help.c.orig	Thu Dec 20 22:06:57 2012
> +++ ./alpine/help.c	Sat Apr 19 14:55:43 2014
> @@ -613,7 +613,7 @@ url_local_config(char *url)
>  {
>      if(!struncmp(url, "x-alpine-config:", 16)){
>  	char **config;
> -	int    rv;
> +	int    rv = MC_NONE;

I agree with this change.

> --- ./alpine/mailcmd.c.orig	Sat Apr 19 15:07:56 2014
> +++ ./alpine/mailcmd.c	Sat Apr 19 15:09:05 2014
> @@ -400,7 +400,7 @@ view_text:
>  			     : (in_index == View)
>  			       ? MH_ANYTHD : MH_NONE);
>  		if(i == mn_get_cur(msgmap)){
> -		    PINETHRD_S *thrd, *topthrd;
> +		    PINETHRD_S *thrd = NULL, *topthrd = NULL;

One only needs to initialize topthrd, not thrd, but it does not hurt to 
initialize them anyway.

> --- ./alpine/mailindx.c.orig	Sat Apr 19 18:54:47 2014
> +++ ./alpine/mailindx.c	Sat Apr 19 18:55:50 2014
> @@ -3633,7 +3633,7 @@ view_in_new_window(void)
>  	  mswin_tw = mswin_displaytext(title, NULL, 0, text,
>  				       NULL, MSWIN_DT_USEALTWINDOW);
>
> -	if(mswin_tw)
> +	if(mswin_tw != NULL)

I think you forgot to initialize mswin_tw to NULL too, so I added that.

> --- ./alpine/mailview.c.orig	Thu Dec 20 22:06:57 2012
> +++ ./alpine/mailview.c	Sat Apr 19 18:47:04 2014
> @@ -1399,6 +1399,8 @@ url_launch(HANDLE_S *handle)
>  	PIPE_S  *syspipe;
>
>  	toolp = handle->h.url.tool;
> +	left_double_quote = NULL;
> +	right_double_quote = NULL;

this code is not needed in version 2.19.12. In fact, you cannot quote 
successfully without knowing which shell you are using, and this code 
disregards the shell, so it is unsafe. All shells interpret special 
characters differently. Alpine 2.19.12 does not have this problem.

> --- ./pith/adrbklib.c.orig	Sat Apr 19 22:22:10 2014
> +++ ./pith/adrbklib.c	Sat Apr 19 22:23:40 2014
> @@ -3468,7 +3468,7 @@ io_error:
>
>      writing = 0;
>
> -    if(ab_stream){
> +    if(ab_stream != NULL){

No harm in adding that.

> --- ./pith/conf.c.orig	Sat Apr 19 22:25:47 2014
> +++ ./pith/conf.c	Sat Apr 19 22:28:12 2014
> @@ -5854,6 +5854,7 @@ write_pinerc(struct pine *ps, EditWhich which, int fla
>  	char datebuf[200];
>
>  	datebuf[0] = '\0';
> +	we_cancel = 0;

I agree with this change. Thank you.

> --- ./pith/help.c.orig	Sat Apr 19 22:48:58 2014
> +++ ./pith/help.c	Sat Apr 19 22:58:19 2014
> @@ -180,6 +180,8 @@ debugjournal_to_file(FILE *dfile)
>  	    else
>  	      phi = NULL;
>
> +	    p = NULL;
> +

This is not needed, the code that follows checks if rmcat has any of the 
possible values {No, Jo, Lo, Hi}, and so since it checks for all 
possibilities, it determines the value of p after that code, so it is not 
really neccesary. I suspect that anyone who modifies that code should take 
care of what happens to p after the next if statement, so I do not see the 
need for this.

> @@ -215,7 +217,7 @@ debugjournal_to_file(FILE *dfile)
>  		p = NULL;
>  	    }
>
> -	    if(p){
> +	    if(p != NULL){

that is an easy change, if you want to do it.

> --- ./pith/imap.c.orig	Sat Apr 19 22:50:54 2014
> +++ ./pith/imap.c	Sat Apr 19 22:52:23 2014
> @@ -210,6 +210,8 @@ mm_expunged(MAILSTREAM *stream, long unsigned int rawn
>      int           is_current = 0;
>      MSGNO_S      *msgmap;
>
> +    mc = NULL;
> +

Added.

> --- ./pith/remote.c.orig	Sat Apr 19 23:19:20 2014
> +++ ./pith/remote.c	Sat Apr 19 23:19:37 2014
> @@ -613,7 +613,7 @@ rd_read_metadata(REMDATA_S *rd)
>  void
>  rd_write_metadata(REMDATA_S *rd, int delete_it)
>  {
> -    char *tempfile;
> +    char *tempfile = NULL;

This is not necessary, the code does not use an unassigned tempfile 
variable, and before it uses it, it assigns it, and checks that it is not 
null. If it can not assign it it executes error code and bails out.

> --- ./pith/string.c.orig	Sat Apr 19 23:27:22 2014
> +++ ./pith/string.c	Sat Apr 19 23:32:52 2014
> @@ -382,6 +382,8 @@ short_str(char *src, char *buf, size_t buflen, int wid
>      char *ans;
>      unsigned alen, first, second;
>
> +    alen = first = second = 0;

This is not necessary. These variables get assigned before they are used. 
This is like the change to help.c before. More to the point. The variables 
first and second get assigned according to the value of variable where. 
This variable can have only three values: FrontDots, MidDots, EndDots. All 
these cases are accounted for. The variable alen is initialized by the 
call

else if((alen = utf8_width(src)) <= wid)

> @@ -419,7 +421,7 @@ short_str(char *src, char *buf, size_t buflen, int wid
>  	    }
>
>  	    q = buf;
> -	    if(first){
> +	    if(first > 0){

the variable first is unsigned, but it does not hurt to add this. I assume 
you also want to add "if(second > 0)" below also.

> @@ -2710,6 +2712,8 @@ copy_quoted_string_asis(char *src)
>  {
>      char *q, *p;
>      int   done = 0, quotes = 0;
> +
> +    q = NULL;

added.

> --- ./pith/thread.c.orig	Sat Apr 19 23:34:37 2014
> +++ ./pith/thread.c	Sat Apr 19 23:34:51 2014
> @@ -883,7 +883,7 @@ collapse_or_expand(struct pine *state, MAILSTREAM *str
>  void
>  select_thread_stmp(struct pine *state, MAILSTREAM *stream, MSGNO_S *msgmap)
>  {
> -    PINETHRD_S   *thrd;
> +    PINETHRD_S   *thrd = NULL;

Added.

Thank you.

-- 
Eduardo
http://patches.freeiz.com/alpine/

Re: [Re-alpine-devel] re-alpine 2.03 patches

[Eduardo C. <ec...@gm...>]

On Sun, 20 Apr 2014, James Jerkins wrote:

> In January of this year I had the opportunity to clean up some buffer 
> overflow bugs in the Alpine 2.11 release. I sent these to the maintainer 
> and received an email reply indicating they would be included. They have 
> not been applied as of 2.19.11.

Dear James,

   I am sorry to see that you think I did not include your contribution. I 
did, so let me go through your patches in this message and let me show you 
how the source code looks like, and we will see if there is a security 
problem. I appreciate all your feedback:

file alpine/help.c: hunk 1 is a cosmetic change, no need to argue about it 
in any way.

hunk 2 was applied.

@@ -414,7 +413,8 @@ help_processor(int cmd, MSGNO_S *msgmap, SCROLL_S *spa
          help_name[0] = '\0';
          for(t = h_texts; t->help_text != NO_HELP; t++)
            if(t->help_text == ((HELP_SCROLL_S *)sparms->proc.data.p)->help_source){
-            strcpy(help_name,t->tag);
+            strncpy(help_name, t->tag, sizeof(help_name)-1);
+            help_name[sizeof(help_name)-1] = '\0';
              break;
            }

file alpine/imap.c: Hunk 1:

-  char *s, c;
+  char *s;

applied.

Hunk 2:

-     if(strlen(*textp) < len - 1)
+     if(strlen(*textp) < len - 1){
         strcpy(tmp, *textp);
+       strcat(tmp, "\n");
+       *textp = s+1;
+     }
       else
         return 0;
-     strcat(tmp, "\n");
-     *textp = s+1;

this was rewritten in a different way, and your comment on it was:

<Quote>
>  I also simplified the code of the line_get function in alpine/imap.c.
>

Whew, thanks. That bit of code made my head hurt a little and seriously 
*needed* simplification. Your rewrite is much nicer than my quick hack.
</Quote>

File alpine/mailcmd.c:

-       strcpy(tmp, "Name of HEADER to match : ");
+       strlcpy(tmp, "Name of HEADER to match : ", sizeof(tmp));
         flags = OE_APPEND_CURRENT;

your substitution of strcpy by strlcpy is by a non standard function. The 
new code is

        strncpy(tmp, "Name of HEADER to match : ", sizeof(tmp)-1);
        tmp[sizeof(tmp)-1] = '\0';

our conversation on this matter was:

<Quote>
>  The original patch to alpine/mailcmd.c uses strlcpy, which is not 
> available in my distribution of gcc, so I switched it to strncpy.

Oops, my system has strlcpy and strlcat and I originally fixed numerous 
things with those functions and then (re)discovered in testing on one of 
my Linux boxes they don't exist there. strncpy is correct.
</Quote>


file alpine/pine-use.c

all hunks applied. This "security bug" was not introduced after version 
2.03, this is code that comes from the time Alpine was Pine.

> Further inspection of the 2.19.11 Alpine code has led me to revert to 
> the re-alpine 2.03.code base because, in my opinion, it is superior code 
> with fewer problems. Interestingly, the buffer overflows in Alpine I 
> submitted patches for had all been introduced since 2.03.

Could you please let me know what are the issues you see with it? I 
thought we had solved the problems you mentioned before, and that they had 
been solved to your satisfaction. As you know, version 2.19.11 is an alpha 
version, the point of it is to find bugs, and fix them before a full 
release is made.

> I undertook a review of the alpine and pith folders sources and found a 
> few reliability issues to fix. I am now running re-alpine 2.03, with my 
> patches and a patched libssl, without any problems.

I am running version 2.19.12 without any problems too. Please test it and 
let me know any problems you have with it. The patch that you sent to this 
list is the first time that I see this code. I will review it too and add 
it or tell you why I did not add it again on a separate message.

Thank you.

-- 
Eduardo
http://patches.freeiz.com/alpine/

[Re-alpine-devel] re-alpine 2.03 patches

[James J. <txt...@ja...>]

Hello,

In January of this year I had the opportunity to clean up some buffer 
overflow bugs in the Alpine 2.11 release. I sent these to the maintainer 
and received an email reply indicating they would be included. They have 
not been applied as of 2.19.11.

Further inspection of the 2.19.11 Alpine code has led me to revert to the 
re-alpine 2.03.code base because, in my opinion, it is superior code with 
fewer problems. Interestingly, the buffer overflows in Alpine I submitted 
patches for had all been introduced since 2.03.

I undertook a review of the alpine and pith folders sources and found a 
few reliability issues to fix. I am now running re-alpine 2.03, with my 
patches and a patched libssl, without any problems.

I have pasted my patches to re-alpine below for your consideration.



--- ./alpine/pine-use.c.orig	Sat Apr 19 14:36:00 2014
+++ ./alpine/pine-use.c	Sat Apr 19 14:45:36 2014
@@ -71,10 +71,10 @@ main(argc, argv)
              printf("%5d users processed so far\n", so_far);
          }

-        if(strcmp(pw->pw_dir, "/") == 0)
+        if(strncmp(pw->pw_dir, "/", 1) == 0)
            continue;

-        sprintf(filename, "%s/.pinerc", pw->pw_dir);
+        snprintf(filename, sizeof(filename), "%s/.pinerc", pw->pw_dir);
          if(stat(filename, &statb) < 0)
            continue;
          if(statb.st_mtime + 7 * DAYSEC > now) 
@@ -98,11 +98,11 @@ main(argc, argv)
              }
          }

-        sprintf(filename, "%s/.signature", pw->pw_dir);
+        snprintf(filename, sizeof(filename), "%s/.signature", pw->pw_dir);
          if(access(filename, 0) == 0)
            sig_files++;

-        sprintf(filename, "%s/core", pw->pw_dir);
+        snprintf(filename, sizeof(filename), "%s/core", pw->pw_dir);
          if((f = fopen(filename, "r")) != NULL) {
              fflush(stdout);
              while((c = getc(f)) != EOF) {
@@ -121,7 +121,7 @@ main(argc, argv)
                      *p = '\0';
                      if(c == EOF)
                        break;
-                    if(strcmp(&buf[strlen(buf) - 13], "(olivebranch)") == 0) {
+                    if(strncmp(&buf[strlen(buf) - 13], "(olivebranch)", 13) == 0) {
                          printf("%s\t%s\n", filename, buf + 14);
                          core_files++;
                          if(core != NULL) {
@@ -161,7 +161,7 @@ mail_file_size(user)
      FILE *f;
      char buf[20480];

-    sprintf(buf, MAILSPOOLPCTS, user);
+    snprintf(buf, sizeof(buf), MAILSPOOLPCTS, user);

      f = fopen(buf, "r");
      if(f  == NULL)


--- ./alpine/help.c.orig	Thu Dec 20 22:06:57 2012
+++ ./alpine/help.c	Sat Apr 19 14:55:43 2014
@@ -613,7 +613,7 @@ url_local_config(char *url)
  {
      if(!struncmp(url, "x-alpine-config:", 16)){
  	char **config;
-	int    rv;
+	int    rv = MC_NONE;

  	config = get_supported_options();
  	if(config){


--- ./alpine/mailcmd.c.orig	Sat Apr 19 15:07:56 2014
+++ ./alpine/mailcmd.c	Sat Apr 19 15:09:05 2014
@@ -400,7 +400,7 @@ view_text:
  			     : (in_index == View)
  			       ? MH_ANYTHD : MH_NONE);
  		if(i == mn_get_cur(msgmap)){
-		    PINETHRD_S *thrd, *topthrd;
+		    PINETHRD_S *thrd = NULL, *topthrd = NULL;

  		    if(THRD_INDX_ENABLED()){
  			mn_dec_cur(stream, msgmap, MH_ANYTHD);


--- ./alpine/mailindx.c.orig	Sat Apr 19 18:54:47 2014
+++ ./alpine/mailindx.c	Sat Apr 19 18:55:50 2014
@@ -3633,7 +3633,7 @@ view_in_new_window(void)
  	  mswin_tw = mswin_displaytext(title, NULL, 0, text,
  				       NULL, MSWIN_DT_USEALTWINDOW);

-	if(mswin_tw)
+	if(mswin_tw != NULL)
  	  mswin_set_readonly(mswin_tw, FALSE);
      }
  }


--- ./alpine/mailview.c.orig	Thu Dec 20 22:06:57 2012
+++ ./alpine/mailview.c	Sat Apr 19 18:47:04 2014
@@ -1399,6 +1399,8 @@ url_launch(HANDLE_S *handle)
  	PIPE_S  *syspipe;

  	toolp = handle->h.url.tool;
+	left_double_quote = NULL;
+	right_double_quote = NULL;

  	/*
  	 * Figure out if we need to quote the URL. If there are shell


--- ./pith/adrbklib.c.orig	Sat Apr 19 22:22:10 2014
+++ ./pith/adrbklib.c	Sat Apr 19 22:23:40 2014
@@ -3468,7 +3468,7 @@ io_error:

      writing = 0;

-    if(ab_stream){
+    if(ab_stream != NULL){
  	fclose(ab_stream);
  	ab_stream = (FILE *)NULL;
      }


--- ./pith/conf.c.orig	Sat Apr 19 22:25:47 2014
+++ ./pith/conf.c	Sat Apr 19 22:28:12 2014
@@ -5854,6 +5854,7 @@ write_pinerc(struct pine *ps, EditWhich which, int fla
  	char datebuf[200];

  	datebuf[0] = '\0';
+	we_cancel = 0;

  	if(!(flags & WRP_NOUSER))
  	  we_cancel = busy_cue(_("Copying to remote config"), NULL, 1);


--- ./pith/help.c.orig	Sat Apr 19 22:48:58 2014
+++ ./pith/help.c	Sat Apr 19 22:58:19 2014
@@ -180,6 +180,8 @@ debugjournal_to_file(FILE *dfile)
  	    else
  	      phi = NULL;

+	    p = NULL;
+
  	    if(pjo && (!plo || pjo->seq <= plo->seq)
  	       && (!phi || pjo->seq <= phi->seq))
  	      rmcat = Jo;
@@ -215,7 +217,7 @@ debugjournal_to_file(FILE *dfile)
  		p = NULL;
  	    }

-	    if(p){
+	    if(p != NULL){
  		if(p->timestamp && p->timestamp[0]
  		   && (fputs(p->timestamp, dfile) == EOF
  		       || fputs(": ", dfile) == EOF))


--- ./pith/imap.c.orig	Sat Apr 19 22:50:54 2014
+++ ./pith/imap.c	Sat Apr 19 22:52:23 2014
@@ -210,6 +210,8 @@ mm_expunged(MAILSTREAM *stream, long unsigned int rawn
      int           is_current = 0;
      MSGNO_S      *msgmap;

+    mc = NULL;
+
  #ifdef DEBUG
      if(ps_global->debug_imap > 1 || ps_global->debugmem)
        dprint((3, "mm_expunged(%s,%lu)\n",


--- ./pith/remote.c.orig	Sat Apr 19 23:19:20 2014
+++ ./pith/remote.c	Sat Apr 19 23:19:37 2014
@@ -613,7 +613,7 @@ rd_read_metadata(REMDATA_S *rd)
  void
  rd_write_metadata(REMDATA_S *rd, int delete_it)
  {
-    char *tempfile;
+    char *tempfile = NULL;
      FILE *fp_old = NULL, *fp_new = NULL;
      char *p = NULL, *pinerc_dir = NULL, *metafile = NULL;
      char *rel_filename, *key;


--- ./pith/string.c.orig	Sat Apr 19 23:27:22 2014
+++ ./pith/string.c	Sat Apr 19 23:32:52 2014
@@ -382,6 +382,8 @@ short_str(char *src, char *buf, size_t buflen, int wid
      char *ans;
      unsigned alen, first, second;

+    alen = first = second = 0;
+
      if(wid <= 0){
  	ans = buf;
  	if(buflen > 0)
@@ -419,7 +421,7 @@ short_str(char *src, char *buf, size_t buflen, int wid
  	    }

  	    q = buf;
-	    if(first){
+	    if(first > 0){
  		q += utf8_to_width(q, src, buflen, first, &got_width);
  		if(got_width != first){
  		  if(second)
@@ -2710,6 +2712,8 @@ copy_quoted_string_asis(char *src)
  {
      char *q, *p;
      int   done = 0, quotes = 0;
+
+    q = NULL;

      if(src){
  	p = q = (char *)fs_get(strlen(src) + 1);


--- ./pith/thread.c.orig	Sat Apr 19 23:34:37 2014
+++ ./pith/thread.c	Sat Apr 19 23:34:51 2014
@@ -883,7 +883,7 @@ collapse_or_expand(struct pine *state, MAILSTREAM *str
  void
  select_thread_stmp(struct pine *state, MAILSTREAM *stream, MSGNO_S *msgmap)
  {
-    PINETHRD_S   *thrd;
+    PINETHRD_S   *thrd = NULL;
      unsigned long rawno, in_thread, set_in_thread, save_branch;

      /* ugly bit means the same thing as return of 1 from individual_select */

Re: [Re-alpine-devel] re-alpine instructions/tutorial/manual?

[Andreas S. <sch...@fa...>]

Hi David,

On Fri, 11 Apr 2014, at 18:24, David Melik wrote:

> I used to be able to find instructions how to set it up for 
> Unix/mbox mail... now it seems the projects originators no longer 
> have any documentation on it. How about writing some instructions, 
> ...

Most of what is available and linked to
at https://www.washington.edu/alpine/ is still correct.
Don't miss the "Technical Notes for Alpine".

Besides, the alpine-info mailing list
http://mailman13.u.washington.edu/mailman/listinfo/alpine-info
is still active, and a friendly and helpful place.

HTH,
-- 
-- Andreas

[Re-alpine-devel] re-alpine instructions/tutorial/manual?

[David M. <dn...@hi...>]

Hello,

I have used ((Re-)Al)Pine on and off since the late 1990s... I used to be able to find instructions how to set it up for Unix/mbox mail... now it seems the projects originators no longer have any documentation on it. How about writing some instructions, like a manual or tutorial so users can use the program? When I go to setup and configuration, I no longer recall what to put in for inbox path, and it assumes I am using IMAP, when I am only using Unix/mbox mail... I would also like to keep the mailbox compatible with a GUI mail client that uses mbox... I, and probably many others, still use these command-line tools to keep updated on system mail, or occasionally send mail from our own to somewhere else on the 'net... some rudimentary instructions would be appreciated.

David
http://www.cwu.edu/~melikd/


_____________________________________________________________
Find out what's HIP!
Visit Hip Planet for news, shopping, forums, chatrooms, free personal and classified ads and much more!
Get FREE E-MAIL! at HipPlanet now!

It's all waiting for you, at http://www.hipplanet.com

[Re-alpine-devel] encryption of outgoing email

[Jacob W. <jac...@fa...>]

Recently, following a software update on my Macintosh 10.7.5, Alpine has begun to prompt me for a password when I send email. After ctrl X, if I merely hit Y at the confirmation, it tells me it doesn't have "a certificate" for the email address. I have to hit E for Encrypt. Then it sometimes prompts me for the password of the smtp server. Apparently, if I've recently typed in a password for some other recipient, it doesn't demand the password again.

I am running ALPINE 2.01(1266). I have not recently updated Alpine nor had I changed my config file. Thus, this appears to be some interaction between the Mac OS and Alpine? Or with the network or firewall of the organization that provides the internet access?

Does anyone have any insights?

Thanks

Jacob A. Wegelin
Assistant Professor
Department of Biostatistics
Virginia Commonwealth University
830 E. Main St., Seventh Floor
P. O. Box 980032
Richmond VA 23298-0032
U.S.A. 
CTSA grant: UL1TR000058
E-mail: jac...@fa... 
URL: http://www.people.vcu.edu/~jwegelin

[Re-alpine-devel] Build "configure" Reported Missing "factory.pem"

[M.Ross <mro...@vf...>]

Build "configure" Reported Missing "factory.pem"

HW: Lenovo R60e 06573AU Laptop
OS: Ubuntu v12.04.3 LTS (newly installed).
SW: recent download of Re-Alpine v2.03 from:
<http://sourceforge.net/projects/re-alpine/files/re-alpine-2.03.tar.bz2/download>

I don't know where "factory.pem" is supposed to be or come from, it is 
called for by "configure", but I know it was not on this drive until I 
put it there! - I searched! - not part of Ubuntu issue! - not 
obviously a ReAlpine issue! Only an archived UW mailing list thread 
refs & explains part of it, as I can find.

See:
Jean-Noel BOUVIER Jean-Noel.Bouvier at imag.fr
Wed Mar 26 07:19:30 PDT 2008
<http://mailman2.u.washington.edu/pipermail/alpine-info/2008-March/000498.html>

 From a plain "configure" run I got:

-snip-
checking for sigprocmask... yes
checking for library containing syslog... none required
configure: * * * SSL file "/etc/ssl/certs/factory.pem" is missing.
configure: * * *   This might indicate that CA certs did not get properly
configure: * * *   installed.  If you get certificate validation failures
configure: * * *   in Alpine, this might be the reason for them.
configure: * * * TCL libraries could not be found.
configure: * * * WEB ALPINE COMPONENT WILL NOT BE BUILT.
configure: creating ./config.status
config.status: creating m4/Makefile
-snip-

I always suffered some cert not found type warnings in Pine/Alpine 
installations for a few servers I visit, requiring ignore, 
no-val-cert, etc...

Is there a whitepaper explaining the requirements, design, source, or 
intent of this "factory.pem" file? other than the obvious, & do 
Pine/Alpine programmers notes of external requirements exist, or any ref?

Anyway, I made a link to /etc/ssl/certs/ca-certificates.crt in that 
directory, & renamed the link "factory.pem" - it is done!

But I suffer make problems now with x509... - too many pkgs needed to 
build not loaded by the OS it seems.

-- 
The Best To You & Yours,
M.RossARR
~~~
The ends of life can be viewed in grave yards.
---end-of-message---
email via Seamonkey v2.23

-------------------------------------------------

VFEmail.net - http://www.vfemail.net
ONLY AT VFEmail! - Use our Metadata Mitigator to keep your email out of the NSA's hands!
$24.95 ONETIME Lifetime accounts with Privacy Features!  
15GB disk! No bandwidth quotas!
Commercial and Bulk Mail Options!  

[Re-alpine-devel] Installing re-alpine

[uroonyc <ur...@gm...>]

Hello,

It's been a while but now I have a solution that works for me.

This thread got Eduardo Chappa's attention and he helped me clarify which
version to choose (alpine vs. re-alpine). Alpine (2.11) is the more recent
version and it's still taken care of by Eduardo - my source of information
(wikipedia, especially the German site) is a little confusing regarding
this.

With further help by Eduardo I was able to compile and run alpine on my
machine. Still I wanted to have a packaged version, just out of curiosity.
I don't really know how to get the advantages of a package (deinstalling
alpine via synaptic leaves the config files although I have selected
complete removal) but that's another issue.

In case anyone is interested I'm posting the steps I had to do in order to
build a package from the source files.
My machine runs on Ubuntu 13.10.

- Get the following programs/dependency-files: build-essential, debhelper,
dh-make, autotools-dev, libncurses5-dev, libpam0g-dev, libssl-dev (I'm not
really sure about libncursesw5-dev; I've installed them, too)

- Download alpine from Eduardo's site http://patches.freeiz.com/alpine/

- Extract the archive
$ tar -xf alpine-2.11.tar.lzma

- cd in the created folder and run dh-make
$ dh_make -f ../alpine-2.11.tar.lzma

- cd in /debian and delete the optional files
$ rm *.ex *.EX README.Debian README.source

- Adjust the rules file to add a configure option so that alpine can store
passwords
  Add these lines at the end of the file:
override_dh_auto_configure:
    dh_auto_configure -- --with-passfile=.pine-passfile

- Run these commands in the source directory to build the .deb package
$ dpkg-buildpackage -S -us -uc
$ dpkg-buildpackage -us -uc


With kind regards,
uroonyc
 <http://patches.freeiz.com/alpine/release/src/alpine-2.11.tar.lzma>

Re: [Re-alpine-devel] Installing re-alpine

[uroonyc <ur...@gm...>]

2013/10/20 Asheesh Laroia <li...@as...>

> Hi there uroonyc,
>
> I'm the maintainer of (re-)alpine in Debian. But it's a bit of a
> complicated situation -- lately we have been packaging whichever of
> re-alpine or Eduardo Chappa's alpine seems best, and it's been pretty
> arbitrary. I want to fix that shortly.
>
> But until then, have you tried simply 'sudo apt-get install alpine'?
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60135031&iu=/4140/ostg.clktrk
> _______________________________________________
> Re-alpine-devel mailing list
> Re-...@li...
> https://lists.sourceforge.net/lists/listinfo/re-alpine-devel
>

Thank you for your reply.
It seems that I don't know much about the difference between alpine and
re-alpine. After reading the wikipedia page about alpine I thought that the
original alpine has a greater version number (2.10) but development stopped
early this year whereas re-alpine is at 2.03 but the lastest version was
published in August and there will be further patches. That's why I assumed
it would be best to use re-alpine.
There is alpine in version 2.10 in the official Ubuntu repository and I'm
afraid that I'm not experienced enough to be able to switch to re-alpine
(configuration and saved mails) if I'd like to do that in the future.

Otherwise if there isn't that much of a difference between both versions
and re-alpine won't get developed any further I could just stick with the
"original".

Furthermore I thought that if package building wouldn't be that hard I
could learn by trying it for re-alpine first and use that for other
programs I'd like to have in a package. It's way simpler to deinstall a
package than a compiled binary which saves files who knows where on the
disk.

[Re-alpine-devel] Fwd: Installing re-alpine

[uroonyc <ur...@gm...>]

2013/10/20 Nicolas Pitre <ni...@fl...>

> On Sun, 20 Oct 2013, uroonyc wrote:
>
> > I have already looked for an apt package because that would be the
> easiest
> > way to try out re-alpine, in my opinion. If I wouldn't like it I could
> > simple deinstall it with apt. I have no experience with installing
> programs
> > manually. I wouldn't know where all those files would be in the end and
> how
> > to delete them thoroughly without messing up something.
> >
> > After having no luck in finding a package I thought I could try to make
> one
> > myself :P That didn't work.
>
> Ubuntu does ship a package for it.
>
> Try: apt-get install alpine
>
>
> Nicolas
>

Thank you for your quick reply.
I have found that package, too. And as far as I can see it's for alpine and
not for the program I was looking for, i.e. re-alpine. I know about alpine
because I used it some years ago for a short period on Windows.
That's why I got to re-alpine in the first place. I need it for Linux and I
thought it would be better to use a more recent version of the original
program.

Re: [Re-alpine-devel] Installing re-alpine

[Nicolas P. <ni...@fl...>]

On Sun, 20 Oct 2013, uroonyc wrote:

> I have already looked for an apt package because that would be the easiest
> way to try out re-alpine, in my opinion. If I wouldn't like it I could
> simple deinstall it with apt. I have no experience with installing programs
> manually. I wouldn't know where all those files would be in the end and how
> to delete them thoroughly without messing up something.
> 
> After having no luck in finding a package I thought I could try to make one
> myself :P That didn't work.

Ubuntu does ship a package for it.

Try: apt-get install alpine


Nicolas

Re: [Re-alpine-devel] Installing re-alpine

[Asheesh L. <li...@as...>]

Hi there uroonyc,

I'm the maintainer of (re-)alpine in Debian. But it's a bit of a
complicated situation -- lately we have been packaging whichever of
re-alpine or Eduardo Chappa's alpine seems best, and it's been pretty
arbitrary. I want to fix that shortly.

But until then, have you tried simply 'sudo apt-get install alpine'?