Re: akpop3d questions
On 24 Nov 2005, at 4:28 PM, J Moore wrote:
On Wed, Nov 23, 2005 at 10:08:13PM -0600, the unit calling itself J
Moore wrote:
I need to set up a POP3 server for a while, and after a quick survey,
akpop3d seemed like a good choice - partly because it supports
POP3 via
SSL. So I built it from the ports tree (3.8 -stable), and
installed it.
I am currently starting from the command line as follows:
# akpop3d -d -s
Attempts to connect result in immediate complaints from the client
(I've
tried two of them: Evolution (Linux) and Pegasus (Windoze).
I'm assuming this is due to the fact that I have no cert or key file
installed or generated?
Before I invest any more time in this, I thought I'd ask if anyone
else
is using akpop3d, what the consensus of opinion is on it, and if
there
is any documentation on how to generate the .pem (Base64-encoded?)
cert
and key files.
Perhaps some fwd progress... got cert key files installed, but I am
bombing during the authentication process. Following is part of the
debug output from my client. I double-checked the password value, and
it's correct (changed here, but my client's log shows it correctly).
The culprit seems to be the group not found error... WTF, O??
23:17:13.312 0009 USER jm\0D\0A
23:17:13.359 0005 +OK\0D\0A
23:17:13.359 0017 PASS abcdefghij\0D\0A
23:17:13.500 0033 -ERR [SYS/TEMP] group not found\0D\0A
23:17:20.718 0006 QUIT\0D\0A
23:17:20.718 18: SSL read error -41 (locus 0, type 0, code 0, 'No data
was read because the remote system closed the connection (recv() ==
0)')
--- Connection closed normally at Wed, 23 Nov 2005 23:17:20. ---
Thanks,
Jay
OK, It looks like the port needs some work as it doesn't handle the
default group name.
main.c:# define DEFAULT_GROUP_NAME mail
It appears that this can be changed with a command line arg.
main.c: case 'g': group_name = optarg; break;
It looks like the groupname is used as an argument to lock the users
mail box.
pop3_session.c.orig: g_inf = getgrnam(mail);
pop3_session.c.orig: if (g_inf==NULL) {
pop3_session.c.orig: if (setegid(g_inf-gr_gid)!=0 real_username
[0] == 0) {
pop3_session.c.orig:syslog(LOG_ERR,%s: %u: %s,setegid()
failed,g_inf-gr_gid,strerror(errno));
pop3_session.c.orig: if (setgid(g_inf-gr_gid)!=0 real_username
[0] == 0) {
pop3_session.c.orig:syslog(LOG_ERR,%s: %u: %s,setgid()
failed,g_inf-gr_gid,strerror(errno));
pop3_session.c.orig: if ((rc=lock_maildrop(maildrop,u_inf-
pw_uid,g_inf-gr_gid))=0)
yup, it then fchowns the lock file
lock_maildrop.c: fchown(fd,uid,gid);
So I would assume on other unix systems /var/mail is group mail by
default, maybe??.
if you want to add mail to the /etc/group file
man -k groupadd
groupadd (8) - add a group to the system
Ian McWilliam
Re: UPDATE: graphics/cairo and graphics/py-cairo
On 11/24/05, Aleksander Piotrowski [EMAIL PROTECTED] wrote: Eric Faurot [EMAIL PROTECTED] wrote: I understand. I am using this patch locally and I have xchat, firefox 1.5rc3 working well. I can make a full package build, run make newlib-depends-check and see what happens. Would that help? Anyway, we can leavecairo like this for now. It simply adds a useless lib dependency, but nothing harmful as far as I can tell. However py-gtk2 really requires the updated py-cairo (for pycairo.pc) Yeah, but 150+ ports are more important that one py-gtk2 port. Sorry, it was not clear: the py-cairo update does not depend on the cairo update. I was suggesting to update py-cairo and py-gtk2 but leave cairo untouched for now. And yes, as you are updating port that has lots of dependant ports, you should, erm, you have to run full pkg build to be sure that everything is OK after that change. I'll do that. Eric.
Re: New port: highmoon
On Thu, Nov 24, 2005 at 11:33:39AM +0100, M. Schatzl wrote:
Hi,
this is a little gravity shooter game with cool graphics:
http://highmoon.gerdsmeier.net/
Since it is the first port I prepared, I'd be grateful for any feedback
concerning the Makefile and patches. As far as I can say, it runs fine.
Please test it nevertheless, for the above reason.
some quick feedback:
- you need a # in 1st line of makefile
- instead of patching the bundled makefile, try using ${INSTALL_DATA} and
friends in do-install in the port's makefile.
- you need to honor CXX, CXXFLAGS etc. which are passed by make
e.g. we do not use -O3 and there's no need for -g
- no need for your header locations patches, just make sure you get
-I${LOCALBASE}/include in CXXFLAGS
Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
sparc64 breakage
Hi, besides regular breakage, these ports could need an eye: net/gaimcannot find silc net/mldonkey/stable ??? These will be fixed rsn: graphics/digikam-docmissing dependency on kdelibs3 lang/gcc/4.0needs fixed bison www/clearsilver 64bit patches got removed x11/kde/bindings3 python problem Nikolay
Re: UPDATE from maintainer: cssed-0.4
On Thu, Nov 24, 2005 at 10:14:05AM +, Antoine Jacoutot wrote: Hi. New version of cssed is out. Diff included. Tested under i386/current. I tested on sparc64. looks okay, and seems they fixed the validation since the previous version. I think your PLIST contains a few locale directories too many. Try make update-plist ? -- steven Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
Re: UPDATE from maintainer: cssed-0.4
Selon steven mestdagh [EMAIL PROTECTED]: then you are probably not -current or don't have the gettext package in your local packages dir. this is the only change needed in PLIST Well, I'm using the most current snapshot available for i386 and available via ftp (I updated everything this morning before updating the port). Anyway, besides that, everything looks fine, right ? Antoine
Re: centericq
so, i downloaded the newest, compiled with --without-msn (no curl here) and so far it's good. what's the issue i should be looking for? You precisely avoided the problem as it's in the msn code, I've the diff --without-msn in my tree but I was hoping it would get fixed. Ignoring it's broken won't do that, spending some time or submitting a bug report to the author will.
LuaSocket update from version 2.0beta3 to 2.0
Tested on i386, macppc and sparc64.
-p.
Index: Makefile
===
RCS file: /cvs/ports/net/luasocket/Makefile,v
retrieving revision 1.3
diff -u -r1.3 Makefile
--- Makefile10 Sep 2005 17:32:15 - 1.3
+++ Makefile24 Nov 2005 17:35:38 -
@@ -3,11 +3,10 @@
NOT_FOR_ARCHS= ${NO_SHARED_ARCHS}
COMMENT= network support for the lua language
-DISTNAME= luasocket-2.0-beta3
-PKGNAME= ${DISTNAME:S/0-b/0b/}p0
+DISTNAME= luasocket-2.0
CATEGORIES=net
-MASTER_SITES= ${HOMEPAGE}
+MASTER_SITES= http://luaforge.net/frs/download.php/1095/
HOMEPAGE= http://www.cs.princeton.edu/~diego/professional/luasocket/
MAINTAINER=Pedro Martelletto [EMAIL PROTECTED]
@@ -24,30 +23,33 @@
RUN_DEPENDS= ${BUILD_DEPENDS}
LIB_DEPENDS= luacompat::devel/luacompat
-MAKE_FILE= makefile.Linux # heh, yeah, linux, sure...
+MAKE_FILE= makefile
post-extract:
@sed -e s:@socketpath@:${PREFIX}/lua/luasocket/?.lua:g \
--e s:@socketcpath@:${PREFIX}/lib/?.so:g \
+-e s:@socketcpath@:${PREFIX}/lua/luasocket/?.so:g \
${FILESDIR}/init.lua ${WRKSRC}/init.lua
+LUASOCKETDIR= ${PREFIX}/lua/luasocket
+
do-install:
- ${INSTALL_DATA_DIR} ${PREFIX}/lua/luasocket
- ${INSTALL_DATA_DIR} ${PREFIX}/lua/luasocket/socket
+ ${INSTALL_DATA_DIR} ${LUASOCKETDIR}
+ ${INSTALL_DATA_DIR} ${LUASOCKETDIR}/socket
+ ${INSTALL_DATA_DIR} ${LUASOCKETDIR}/mime
${INSTALL_DATA_DIR} ${PREFIX}/share/doc/luasocket
${INSTALL_DATA_DIR} ${PREFIX}/share/examples/luasocket
- ${INSTALL_DATA} ${WRKSRC}/luasocket*.so.*.* ${PREFIX}/lib/luasocket.so
- ${INSTALL_DATA} ${WRKSRC}/luamime*.so* ${PREFIX}/lib/luamime.so
- ${INSTALL_DATA} ${WRKSRC}/init.lua ${PREFIX}/lua/luasocket
- ${INSTALL_DATA} ${WRKSRC}/lua/socket.lua ${PREFIX}/lua/luasocket
- ${INSTALL_DATA} ${WRKSRC}/lua/ltn12.lua ${PREFIX}/lua/luasocket
- ${INSTALL_DATA} ${WRKSRC}/lua/mime.lua ${PREFIX}/lua/luasocket
- ${INSTALL_DATA} ${WRKSRC}/lua/ftp.lua ${PREFIX}/lua/luasocket/socket
- ${INSTALL_DATA} ${WRKSRC}/lua/http.lua ${PREFIX}/lua/luasocket/socket
- ${INSTALL_DATA} ${WRKSRC}/lua/smtp.lua ${PREFIX}/lua/luasocket/socket
- ${INSTALL_DATA} ${WRKSRC}/lua/tp.lua ${PREFIX}/lua/luasocket/socket
- ${INSTALL_DATA} ${WRKSRC}/lua/url.lua ${PREFIX}/lua/luasocket/socket
- ${INSTALL_DATA} ${WRKSRC}/manual/* ${PREFIX}/share/doc/luasocket
- ${INSTALL_DATA} ${WRKSRC}/examples/* ${PREFIX}/share/examples/luasocket
+ ${INSTALL_DATA} ${WRKSRC}/src/socket.so ${LUASOCKETDIR}/socket/core.so
+ ${INSTALL_DATA} ${WRKSRC}/src/mime.so ${LUASOCKETDIR}/mime/core.so
+ ${INSTALL_DATA} ${WRKSRC}/init.lua ${LUASOCKETDIR}
+ ${INSTALL_DATA} ${WRKSRC}/src/socket.lua ${LUASOCKETDIR}
+ ${INSTALL_DATA} ${WRKSRC}/src/ltn12.lua ${LUASOCKETDIR}
+ ${INSTALL_DATA} ${WRKSRC}/src/mime.lua ${LUASOCKETDIR}
+ ${INSTALL_DATA} ${WRKSRC}/src/ftp.lua ${LUASOCKETDIR}/socket
+ ${INSTALL_DATA} ${WRKSRC}/src/http.lua ${LUASOCKETDIR}/socket
+ ${INSTALL_DATA} ${WRKSRC}/src/smtp.lua ${LUASOCKETDIR}/socket
+ ${INSTALL_DATA} ${WRKSRC}/src/tp.lua ${LUASOCKETDIR}/socket
+ ${INSTALL_DATA} ${WRKSRC}/src/url.lua ${LUASOCKETDIR}/socket
+ ${INSTALL_DATA} ${WRKSRC}/doc/* ${PREFIX}/share/doc/luasocket
+ ${INSTALL_DATA} ${WRKSRC}/samples/* ${PREFIX}/share/examples/luasocket
.include bsd.port.mk
Index: distinfo
===
RCS file: /cvs/ports/net/luasocket/distinfo,v
retrieving revision 1.3
diff -u -r1.3 distinfo
--- distinfo19 Apr 2005 15:14:38 - 1.3
+++ distinfo24 Nov 2005 17:35:38 -
@@ -1,4 +1,4 @@
-MD5 (luasocket-2.0-beta3.tar.gz) = ba32118746205cb088ce7fed06d8d1b7
-RMD160 (luasocket-2.0-beta3.tar.gz) = 49deb7b7050f831717b7e34852cf044745f0e62a
-SHA1 (luasocket-2.0-beta3.tar.gz) = e49638ea4dee174dba26af58d9bb37b93a12bd14
-SIZE (luasocket-2.0-beta3.tar.gz) = 115472
+MD5 (luasocket-2.0.tar.gz) = 8416fed2eb2fc223439d7f8ffd90eb85
+RMD160 (luasocket-2.0.tar.gz) = abede8f3618130781567157bcd6b42f69d180dde
+SHA1 (luasocket-2.0.tar.gz) = ffef5b9faf9f802101d3121a60c77f912deb38aa
+SIZE (luasocket-2.0.tar.gz) = 118540
Index: patches/patch-buffer_c
===
RCS file: patches/patch-buffer_c
diff -N patches/patch-buffer_c
--- patches/patch-buffer_c 16 Dec 2004 13:38:12 - 1.1.1.1
+++ /dev/null 1 Jan 1970 00:00:00 -
@@ -1,23 +0,0 @@
-$OpenBSD: patch-buffer_c,v 1.1.1.1 2004/12/16 13:38:12 pedro Exp $
buffer.c.orig Mon Jul 26 17:17:35 2004
-+++ buffer.c Tue Aug 10 21:11:45 2004
-@@ -75,7 +75,9 @@ int buf_meth_setstats(lua_State *L, p_bu
- \*-*/
- int buf_meth_send(lua_State *L, p_buf buf) {
- int top =
update: sysutils/tpb
0.6.3 - 0.6.4
Index: Makefile
===
RCS file: /cvs/ports/sysutils/tpb/Makefile,v
retrieving revision 1.4
diff -u -r1.4 Makefile
--- Makefile9 Nov 2005 01:25:38 - 1.4
+++ Makefile24 Nov 2005 18:16:08 -
@@ -5,8 +5,7 @@
COMMENT= utility to use ibm thinkpad(tm) special keys
-DISTNAME= tpb-0.6.3
-PKGNAME= ${DISTNAME}p1
+DISTNAME= tpb-0.6.4
CATEGORIES=sysutils
HOMEPAGE= http://www.nongnu.org/tpb/
Index: distinfo
===
RCS file: /cvs/ports/sysutils/tpb/distinfo,v
retrieving revision 1.2
diff -u -r1.2 distinfo
--- distinfo5 Jan 2005 17:29:39 - 1.2
+++ distinfo24 Nov 2005 18:16:08 -
@@ -1,4 +1,4 @@
-MD5 (tpb-0.6.3.tar.gz) = fc11a0af992178e4013fd9c041dbaa7e
-RMD160 (tpb-0.6.3.tar.gz) = f5f4348aaad676b9a1f86b0006728c715b5ca291
-SHA1 (tpb-0.6.3.tar.gz) = e2787d53f5e4caa30c2b3b0ef0f960bcad152f1c
-SIZE (tpb-0.6.3.tar.gz) = 277058
+MD5 (tpb-0.6.4.tar.gz) = 563b544fbedcc1763130a0bca1ad63b6
+RMD160 (tpb-0.6.4.tar.gz) = 09cf46ed1933d568ced7756f85fef35b23c1ded2
+SHA1 (tpb-0.6.4.tar.gz) = 481894d8289c75d821f50568f9c2d0005486e0e8
+SIZE (tpb-0.6.4.tar.gz) = 280316
Index: patches/patch-man_tpb_1
===
RCS file: /cvs/ports/sysutils/tpb/patches/patch-man_tpb_1,v
retrieving revision 1.1
diff -u -r1.1 patch-man_tpb_1
--- patches/patch-man_tpb_1 13 May 2005 20:12:24 - 1.1
+++ patches/patch-man_tpb_1 24 Nov 2005 18:16:08 -
@@ -1,17 +1,16 @@
-$OpenBSD: patch-man_tpb_1,v 1.1 2005/05/13 20:12:24 msf Exp $
man/tpb.1.orig Fri May 13 12:41:46 2005
-+++ man/tpb.1 Fri May 13 12:42:14 2005
-@@ -265,13 +265,6 @@ value middle is only supported by xosd 2
- .TP 8
- .B OSDALIGN
- Defines the alignment of the osd. Possible values are left, center and right.
Default is left.
+--- man/tpb.1.orig Thu Nov 24 15:36:31 2005
man/tpb.1 Thu Nov 24 15:36:41 2005
+@@ -331,13 +331,6 @@
+ Defines the alignment of the osd.
+ Possible values are left, center and right.
+ Default is left.
-.SH NOTES
--Requires NVRAM device /dev/nvram. This must be enabled in your kernel.
--Although the device file must exist. The device file could be created with
--\'mknod /dev/nvram c 10 144\'
+-Requires NVRAM device /dev/nvram.
+-This must be enabled in your kernel.
+-Although the device file must exist.
+-The device file could be created with \'mknod /dev/nvram c 10 144\'
-
--This program was developed on a Thinkpad Model T21. I don't know if it works
on
--other models.
+-This program was developed on a Thinkpad Model T21.
.SH EXAMPLE
tpb \-\-osd=off \-\-verbose \-\-thinkpad=xterm \-T ntpctl \-e ntpctl
.SH COPYRIGHT
Re: centericq
hmm, on Thu, Nov 24, 2005 at 09:13:30AM -0700, Peter Valchev said that so, i downloaded the newest, compiled with --without-msn (no curl here) and so far it's good. what's the issue i should be looking for? You precisely avoided the problem as it's in the msn code, I've the diff --without-msn in my tree but I was hoping it would get fixed. Ignoring it's broken won't do that, spending some time or submitting a bug report to the author will. i feel your pain. but i have never used msn and don't plan to. is there some more info about this flaw somewhere? i would gladly write the author if i knew what's the deal. i presume it's openbsd specific if the code is released and noone else complains? or just telling the author that having msn compiled in the program core dumps on openbsd? -f -- you're old when you forget how to start your rocking chair.
Re: akpop3d questions
On Thu, Nov 24, 2005 at 07:17:54PM +1100, the unit calling itself Ian McWilliam
wrote:
Perhaps some fwd progress... got cert key files installed, but I am
bombing during the authentication process. Following is part of the
debug output from my client. I double-checked the password value, and
it's correct (changed here, but my client's log shows it correctly).
The culprit seems to be the group not found error... WTF, O??
23:17:13.312 0009 USER jm\0D\0A
23:17:13.359 0005 +OK\0D\0A
23:17:13.359 0017 PASS abcdefghij\0D\0A
23:17:13.500 0033 -ERR [SYS/TEMP] group not found\0D\0A
23:17:20.718 0006 QUIT\0D\0A
23:17:20.718 18: SSL read error -41 (locus 0, type 0, code 0, 'No data
was read because the remote system closed the connection (recv() ==
0)')
--- Connection closed normally at Wed, 23 Nov 2005 23:17:20. ---
OK, It looks like the port needs some work as it doesn't handle the
default group name.
main.c:# define DEFAULT_GROUP_NAME mail
It appears that this can be changed with a command line arg.
It can - that's how I finally got it to work. According to man akpop3d,
-g groupID does it. (and apparently I'm confused - I thought group ID
was the number, but akpop3d wants the group name, ... whatever)
main.c: case 'g': group_name = optarg; break;
It looks like the groupname is used as an argument to lock the users
mail box.
pop3_session.c.orig: g_inf = getgrnam(mail);
pop3_session.c.orig: if (g_inf==NULL) {
pop3_session.c.orig: if (setegid(g_inf-gr_gid)!=0 real_username
[0] == 0) {
pop3_session.c.orig:syslog(LOG_ERR,%s: %u: %s,setegid()
failed,g_inf-gr_gid,strerror(errno));
pop3_session.c.orig: if (setgid(g_inf-gr_gid)!=0 real_username
[0] == 0) {
pop3_session.c.orig:syslog(LOG_ERR,%s: %u: %s,setgid()
failed,g_inf-gr_gid,strerror(errno));
pop3_session.c.orig: if ((rc=lock_maildrop(maildrop,u_inf-
pw_uid,g_inf-gr_gid))=0)
yup, it then fchowns the lock file
lock_maildrop.c: fchown(fd,uid,gid);
So I would assume on other unix systems /var/mail is group mail by
default, maybe??.
That may be... I checked a FreeBSD and a Linux (Fedora) box - both
listed mail as the group for /var/mail. So OpenBSD would appear to be
in a minority position.
if you want to add mail to the /etc/group file
man -k groupadd
groupadd (8) - add a group to the system
I thought about this, but wouldn't you actually have to change group
ownership of /var/mail to group mail for this to make any difference?
And if you did this, wouldn't you risk breaking something else?
Thanks for the insight,
Jay
fontconfig and freetype (was Re: New and some updated ports)
Antti Harri wrote: Hello, Here are some ports I've made and a couple of updated ones: http://openbsd.fi/~iku/files/ports/ fontconfig-2.3.2-obsd_port.tar.gz freetype2-2.1.10-obsd_port.tar.gz [and other ports, that are not the issue here] The fontconfig in XF4 will be update to 2.3.2 with the next X.Org import. Are there any ports that need this update quickly ? freetype2 is part of XF4 too. The next X.Org import will use 2.1.9, since 2.1.10 is causing problems in X, and those have not been solved yet. Is there any port that's not happy with 2.1.9 ? More generally, it seems to me that the current consensus among OpenBSD developpers is that ports that need parts of X (typically fontconfig freetype, not to mention expat again) should install xbase. There's no security risk associated with it as there are no setuid programs there. So there is no need to duplicate code that is part of XF4 in ports. If parts of them get updated they can be updated in XF4, just talk to the X maintainers. -- Matthieu Herrb
