HipToday

From what I've seen and read these steps:

autoconf

automake --add-missing

automake src/bindings/ScriptDev2/Makefile

are unnecessary and will break things (though the threads discussing that had different issues than what you're describing).

it's easy for windows users.... I just managed to compile fine with ACE5.7.6. Just did

1.) download the zip file from http://download.dre.vanderbilt.edu/previous_versions/ACE-5.7.6.zip

2.) unzip the ace_wrappers into the %DRIVE%%MANGOS FOLDER%/dep folder overwritting the one that is there

3.) make a file called config.h with just #include "config-win32.h" in /ace_wrappers/ace

4.) compile

Except then you end up with over 100MB of stuff in your source code that you don't need.

I've tried both NetSky's 5.7.6 patch and ApoC's 5.7.5 patch for FreeBSD on OpenBSD 4.7-beta and neither compiled. I didn't have the time to dig into it very far, but I do know that with NetSky's patch I was having similar issues to the ones reported by ApoC (missing .inl files).

Are there linux distros that don't have pkg-config available ?

I don't know about Linux distros, but pkg-config doesn't work on OpenBSD (and I think FreeBSD too).

And can't we just check for SSL headers in a few predefined locations ?

That's exactly what this stuff is trying to do if pkg-config is either unavailable or doesn't give us what we need.

Vladimir, do you agree that the OpenSSL check should be setting OPENSSL_CFLAGS? Or should it indeed be OPENSSL_CPPFLAGS as in your patch?

@Snake_Plissken:

It looks like maybe the patch didn't apply cleanly.

@Samuell.Sk:

There's no indication at all that OpenSSL was even looked for in that log, I'd expect to at least see an attempt and a failure with the OpenSSL check. Also make sure the patch is applying cleanly for you.

@ALL:

When testing this stuff it is best to start with a clean clone of the repository and do everything from scratch, the idea being to eliminate as many variables as possible. I've tested this stuff on Ubuntu Desktop 9.10, Ubunter Server 8.04.3, Debian 5.03, and OpenBSD-current (all i386, 32-bit) and the only time I've been able to reproduce the compile errors is when I remove the libssl-dev package. I'm having a hell of a time figuring out what is different on the systems people are having issues on. This is frustrating for me too, my whole goal with this was to get MaNGOS compiling on more systems.

I appreciate all of your feedback .

not working, same result,

Dang, could you post your config.log (in objdir)?

Is this patch help?

Can someone with problems build mangos after SSL detection chnages check this?

I was looking at this more over the weekend and came up with something similar. But why did you change OPENSSL_INCLUDES to OPENSSL_CPPFLAGS? I was going to change it to OPENSSL_CFLAGS as that is what it looked like was being set previously. My alternate patch also changes the pkg-config commands a little bit, and adds an error message if the OpenSSL check fails (which should have been there all along, but I was hurrying to try to fix people's issues).

diff --git a/configure.ac b/configure.ac
index 57cf30e..cdc3ce1 100644
--- a/configure.ac
+++ b/configure.ac
@@ -97,7 +97,7 @@ AC_CHECK_LIB( pthread, pthread_create, [],
    ])
AC_CHECK_LIB( z, compress, [ZLIB=-lz],[AC_MSG_ERROR([Missing zlib])] )
AC_CHECK_LIB( compat, ftime, [COMPATLIB=-lcompat] )
-AX_CHECK_OPENSSL()
+AX_CHECK_OPENSSL([], [AC_MSG_ERROR([Missing OpenSSL])])

AC_ARG_WITH(postgresql,
[  --with-postgresql       Use PostgreSQL as a backend (default: no)],
diff --git a/m4/ax_check_openssl.m4 b/m4/ax_check_openssl.m4
index 7f5c319..3b5ed9c 100644
--- a/m4/ax_check_openssl.m4
+++ b/m4/ax_check_openssl.m4
@@ -31,6 +31,9 @@
#   permitted in any medium without royalty provided the copyright notice
#   and this notice are preserved. This file is offered as-is, without any
#   warranty.
+#
+# MaNGOS local changes: OPENSSL_INCLUDES replaced by OPENSSL_CFLAGS for
+# better compatibility with the way PKG_CONFIG previously worked.

AU_ALIAS([CHECK_SSL], [AX_CHECK_OPENSSL])
AC_DEFUN([AX_CHECK_OPENSSL], [
@@ -53,8 +56,8 @@ AC_DEFUN([AX_CHECK_OPENSSL], [
            if test x"$PKG_CONFIG" != x""; then
                OPENSSL_LDFLAGS=`$PKG_CONFIG openssl --libs-only-L 2>/dev/null`
                if test $? = 0; then
-                    OPENSSL_LIBS=`$PKG_CONFIG openssl --libs-only-l 2>/dev/null`
-                    OPENSSL_INCLUDES=`$PKG_CONFIG openssl --cflags-only-I 2>/dev/null`
+                    OPENSSL_LIBS=`$PKG_CONFIG openssl --libs 2>/dev/null`
+                    OPENSSL_CFLAGS=`$PKG_CONFIG openssl --cflags 2>/dev/null`
                    found=true
                fi
            fi
@@ -71,11 +74,11 @@ AC_DEFUN([AX_CHECK_OPENSSL], [
    # an 'openssl' subdirectory

    if ! $found; then
-        OPENSSL_INCLUDES=
+        OPENSSL_CFLAGS=
        for ssldir in $ssldirs; do
            AC_MSG_CHECKING([for openssl/ssl.h in $ssldir])
            if test -f "$ssldir/include/openssl/ssl.h"; then
-                OPENSSL_INCLUDES="-I$ssldir/include"
+                OPENSSL_CFLAGS="-I$ssldir/include"
                OPENSSL_LDFLAGS="-L $ssldir/lib"
                OPENSSL_LIBS="-lssl -lcrypto"
                found=true
@@ -95,14 +98,14 @@ AC_DEFUN([AX_CHECK_OPENSSL], [

    AC_MSG_CHECKING([whether compiling and linking against OpenSSL works])
    echo "Trying link with OPENSSL_LDFLAGS=$OPENSSL_LDFLAGS;" \\
-        "OPENSSL_LIBS=$OPENSSL_LIBS; OPENSSL_INCLUDES=$OPENSSL_INCLUDES" >&AS_MESSAGE_LOG_FD
+        "OPENSSL_LIBS=$OPENSSL_LIBS; OPENSSL_CFLAGS=$OPENSSL_CFLAGS" >&AS_MESSAGE_LOG_FD

    save_LIBS="$LIBS"
    save_LDFLAGS="$LDFLAGS"
    save_CPPFLAGS="$CPPFLAGS"
    LDFLAGS="$LDFLAGS $OPENSSL_LDFLAGS"
    LIBS="$OPENSSL_LIBS $LIBS"
-    CPPFLAGS="$OPENSSL_INCLUDES $CPPFLAGS"
+    CPPFLAGS="$OPENSSL_CFLAGS $CPPFLAGS"
    AC_LINK_IFELSE(
        AC_LANG_PROGRAM([#include <openssl/ssl.h>], [sSL_new(NULL)]),
        [
@@ -116,7 +119,7 @@ AC_DEFUN([AX_CHECK_OPENSSL], [
    LDFLAGS="$save_LDFLAGS"
    LIBS="$save_LIBS"

-    AC_SUBST([OPENSSL_INCLUDES])
+    AC_SUBST([OPENSSL_CFLAGS])
    AC_SUBST([OPENSSL_LIBS])
    AC_SUBST([OPENSSL_LDFLAGS])
])

I'm testing this as I post this on Ubuntu Server 8.04.3, but since I didn't have problems before, it won't really mean much unless some other folks out there test it.

I use linux ubuntu and for OpenSSL i use the same and before this comit all where ok ...

Now not

I understand that, but I still need the information I asked for if I'm going to try to help. What version of Ubuntu? What version of OpenSSL? How did you install OpenSSL (compiled from source, using apt-get, etc.)? Do you have libssl-dev installed?

same ...

What distro are you using? What version of OpenSSL? How did you install OpenSSL?

http://wow.evo.bg/config.log

I delete all of the repo in my server and download again.

I think now is ok. I compile in this moment.

So was that log you posted from before you downloaded a clean repository? Because it doesn't look like it is using the stuff from my patches.

Doesn't work to me.

I test it on another ubuntu server and I get the same problems.

If I use official revision (this include the first patch) I get configure error.

If I use official rev. + second patch I get compile error.

Could you post your configure.log (on pastebin.com or something similar, configure.log is quite large)?

Alright, with my fix posted above applied to 9225, I'm able to compile fine on Ubuntu 9.10 Desktop and OpenBSD-current (both i386). So now I don't know if my original patch to configure.ac should be reverted or if my fix, submitted afterwords, should just be committed. Have other people tried the fix I posted on freshly checked out code? alein?

Bah! This should probably be taken out until I can get myself a Linux distro setup for testing.

It seems the AX_CHECK_OPENSSL() macro already tries to use pkg-config to find OpenSSL before searching in some standard default locations, so nesting it in the PKG_CHECK_MODULES() macro was redundant. That caused some weird nesting issues in the configure script that I didn't see since pkg-config doesn't find OpenSSL on OpenBSD. Hopefully this new patch will fix these issues in environments where pkg-config and OpenSSL work as expected.

Thanks for the commit anyway!

Here's a patch that fixes what was committed from my previous patch (I hope):

diff --git a/configure.ac b/configure.ac
index 3406e1c..926da9b 100644
--- a/configure.ac
+++ b/configure.ac
@@ -97,7 +97,7 @@ AC_CHECK_LIB( pthread, pthread_create, [],
    ])
AC_CHECK_LIB( z, compress, [ZLIB=-lz],[AC_MSG_ERROR([Missing zlib])] )
AC_CHECK_LIB( compat, ftime, [COMPATLIB=-lcompat] )
-PKG_CHECK_MODULES(OPENSSL, [openssl], [], [CHECK_SSL()])
+AX_CHECK_OPENSSL()

AC_ARG_WITH(postgresql,
[  --with-postgresql       Use PostgreSQL as a backend (default: no)],

I figured it out, I had to change the 'minitems' and 'maxitems' of each auction house in the 'auctionhousebot' table (which defaulted to 0). Perhaps that's mentioned somewhere in this thread, but I didn't see it.

Hmmmm... not sure what I'm doing wrong, but I don't see any items appearing in my Auction House (I've checked in Stormwind and the auctionhouse table in the characters database is empty too).

This is my first attempt at using the AHBot. I created a new account and character to use specifically for the Auction House with GUID 4 and Account 7, and after patching then building MaNGOS I applied the characters_auctionhousebot.sql to my characters database.

MaNGOS version 9045, no other third-party patches, UDB 0.11.6 (386).

Database Settings:

INSERT INTO `auctionhousebot` VALUES
(2,'Alliance',0,0,8,24,0,27,12,10,1,0,0,0,10,30,8,2,0,0,100,150,150,250,800,1400,1250,1750,2250,4550,3250,5550,5250,6550,70,100,70,100,80,100,75,100,80,100,80,100,80,100,0,0,3,2,1,1,1,1,1,5,12,15,20,22,1,1),
(6,'Horde',0,0,8,24,0,27,12,10,1,0,0,0,10,30,8,2,0,0,100,150,150,250,800,1400,1250,1750,2250,4550,3250,5550,5250,6550,70,100,70,100,80,100,75,100,80,100,80,100,80,100,0,0,3,2,1,1,1,1,1,5,12,15,20,22,1,1),
(7,'Neutral',0,0,8,24,0,27,12,10,1,0,0,0,10,30,8,2,0,0,100,150,150,250,800,1400,1250,1750,2250,4550,3250,5550,5250,6550,70,100,70,100,80,100,75,100,80,100,80,100,80,100,0,0,3,2,1,1,1,1,1,5,12,15,20,22,1,1)
;

Config Settings:

AuctionHouseBot.EnableSeller = 1
AuctionHouseBot.EnableBuyer = 1
AuctionHouseBot.Account = 7
AuctionHouseBot.GUID = 4
AuctionHouseBot.VendorItems = 1
AuctionHouseBot.LootItems = 1
AuctionHouseBot.OtherItems = 1
AuctionHouseBot.No_Bind = 1
AuctionHouseBot.Bind_When_Picked_Up = 0
AuctionHouseBot.Bind_When_Equipped = 1
AuctionHouseBot.Bind_When_Use = 1
AuctionHouseBot.Bind_Quest_Item = 0
AuctionHouseBot.ItemsPerCycle = 200
AuctionHouseBot.UseBuyPriceForSeller = 0
AuctionHouseBot.UseBuyPriceForBuyer = 0

Server Log:

2009-12-24 11:58:00 Game Event system initialized.
2009-12-24 11:58:00 Initialize AuctionHouseBot...
2009-12-24 11:58:03 AuctionHouseBot:
2009-12-24 11:58:03 loaded 9 grey trade goods
2009-12-24 11:58:03 loaded 555 white trade goods
2009-12-24 11:58:03 loaded 68 green trade goods
2009-12-24 11:58:03 loaded 29 blue trade goods
2009-12-24 11:58:03 loaded 3 purple trade goods
2009-12-24 11:58:03 loaded 1 orange trade goods
2009-12-24 11:58:03 loaded 0 yellow trade goods
2009-12-24 11:58:03 loaded 1286 grey items
2009-12-24 11:58:03 loaded 2751 white items
2009-12-24 11:58:03 loaded 5481 green items
2009-12-24 11:58:03 loaded 1733 blue items
2009-12-24 11:58:03 loaded 732 purple items
2009-12-24 11:58:03 loaded 0 orange items
2009-12-24 11:58:03 loaded 0 yellow items
2009-12-24 11:58:03 AuctionHouseBot [AHBot-004-HotFix-06] is now loaded
2009-12-24 11:58:03 AuctionHouseBot updated by Naicisum (original by ChrisK and Paradox)
2009-12-24 11:58:03 AuctionHouseBot now includes AHBuyer by Kerbe and Paradox
2009-12-24 11:58:03 WORLD: World initialized
2009-12-24 11:58:03 SERVER STARTUP TIME: 1 minutes 20 seconds

Resolved with Vlad on IRC - telnet specification (and most implementations) really doesn't specify and header or "garbage", thus most telnet clients (including MS telnet in winXP) should work with mangos RA.

The problem is purely in Putty client. Putty v0.60 doesn't have configurable telnet port (well, it has, it just uses hardcoded 23 anyway, so I was unable to test anything with it), Puttytel allowed me to specify a custom port, but it added a 21-byte garbage before the first user-specified data (in a separate packet), thus effectively breaking any strict server-side data parsing mechanisms (based on raw TCP payload parsing).

0030                    ff fb  1f ff fb 20 ff fb 18 ff         .. ... ....
0040  fb 27 ff fd 01 ff fb 03  ff fd 03                  .'...... ...

This bunch of data was sent only once per connection.

The garbage above contain telnet options. Putty's "telnet" mode assumes a running telnetd server and sends terminal info data and some other things to the telnet server. Nearly all other telnet clients allow to send those options "on live" via a control sequence, but none of them sends them automatically (unless specified in a config file), thus being able to send/receive raw TCP payload data.

Putty's "raw" mode doesn't send anything automatically, so that one should be used, BSD/linux telnet implementations should work fine by default, MS winXP telnet works as well. Any other TCP client can be in fact used (netcat, ...).

Interesting. Thanks for the explanation. I never actually tried connecting via PuTTY (I only use OpenBSD's telnet client to connect to RA on localhost), but this is good to know.

Try setting Ra.Secure to 0 to see if you get some helpful output from the Remote Admin Console (instead of it just closing the connection on you, which is why your window is disappearing).

+1. I wanted this exact feature yesterday. Perhaps if I clean some other stuff off my plate and get ambitious I'll see if I can tackle it.

FragFrog is correct.

It was my patch that "broke" this. My original patch was designed to keep the old behavior unless it was specifically set to operate the new way, but it was decided that it wasn't worth it to mess around with a configuration option to maintain that backwards compatiblity. Sorry.

There was a pasto in my original diff here's the real one (I've also updated the original post):

http://paste2.org/p/531006

I've submitted a patch:

[PATCH] Add username/password prompts to Remote Admin

First one ever, so be gentle.

What features does the patch add?

Adds username and password prompts to Remote Admin Console if Ra.Interactive, a new configuration option, is set to 1. If Ra.Interactive is missing from mangos.conf or is set to 0, RA functions exactly as it did before.

For which repository revision was the patch created?

8870

Is there a thread in the bug report section or at lighthouse?

No.

Who has been writing this patch?

- Me (HipToday <nick [at] nicktempleton.com>)

Diff: http://paste2.org/p/531006