From: Kevin L. Mitchell Date: Wed, 20 Jun 2001 20:56:32 +0000 (+0000) Subject: Author: Kev X-Git-Url: http://git.pk910.de/?p=ircu2.10.12-pk.git;a=commitdiff_plain;h=a767b7fcb66ccfcff3e7fd56ec1fd7d39222e13f Author: Kev Log message: Beginning to update some of the important out-of-date documentation. git-svn-id: file:///home/klmitch/undernet-ircu/undernet-ircu-svn/ircu2/trunk@508 c9e4aea6-c8fd-4c43-8297-357d70d61c8c --- diff --git a/ChangeLog b/ChangeLog index f830303..71d4497 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,7 @@ +2001-06-20 Kevin L. Mitchell + + * INSTALL: partial update of INSTALL for u2.10.11 release... + 2001-06-14 Kevin L. Mitchell * ircd/table_gen.c (makeTables): finally got tired of the diff --git a/INSTALL b/INSTALL index 01ebc04..628a9d3 100644 --- a/INSTALL +++ b/INSTALL @@ -1,5 +1,6 @@ INSTALL file by Run Updated by Isomer + Updated for u2.10.11 by Kev This is the UnderNet IRC daemon. @@ -7,18 +8,17 @@ The installation of the IRC daemon (ircd) exists of the following steps: 1) Retrieve the package. 2) cd into the base directory. -3) `./configure' -4) `make config' -5) `make' -6) `make install' +3) "./configure" +4) "make" +5) "make install" 1) Retrieve the package. -==================== +======================== The recommended way to get the ircu package now is to use CVS. CVS makes upgrades a lot less painful and lets you get the latest package. -1.1) The first thing you need to do is 'authenticate' yourself against the +1.1) The first thing you need to do is "authenticate" yourself against the server. This is done with: @@ -26,38 +26,45 @@ This is done with: cvs -d :pserver:anoncvs@coder.com.undernet.org:/home/coder-com/cvs login (we recommend that you cut and paste the above line to use it :) -When it prompts you for a password enter 'anoncvs'. +When it prompts you for a password enter "anoncvs". -1.2) Then you have to decide which of the trees you want to use: +1.2) The module you will check out is ircu2.10. There are three branches +you can check out. -irc2.10 - This is the *STABLE* tree. If in doubt use this one! +stable - This branch is always the recommended version for use on the + Undernet. To check out the stable tree, add "-r u2_10_11" to + the CVS command line. -beta - This tree is undergoing testing before being promoted to ircu2.10 - It may be buggy. Use on the undernet production network is prohibited, - except for certain authorised servers. +beta - This branch is undergoing testing before being promoted to the + stable branch. It may be buggy. Use on the Undernet's production + network is prohibited, except for certain authorized servers. The + "-r" flag you need to check this branch out is documented on the + Coder Committee's web site, http://coder-com.undernet.org. + +alpha - This is the development branch. It is not guaranteed to even + compile, and should be considered HIGHLY unstable. It is NOT + intended for production use. If you wish to play with the server, + you may wish to investigate the test network. To check this branch + out, use no "-r" flags. -alpha - This is the development tree, if you are planning on making patches - to be submitted to the main source tree we recommend you use this - tree. However this tree is *not* guaranteed to compile, and should - be concidered HIGHLY unstable. It is not intended for production - use. to check out the tree, type: cvs -d :pserver:anoncvs@coder-com.undernet.org:/home/coder-com/cvs checkout - -P irc2.10 + -P ircu2.10 -The above two lines shouldn't have an enter between them. if you want to -use another tree, replace 'irc2.10' with the tree you want to use. This -will create a directory irc2.10, and put all the files in there. +The above two lines shouldn't have an enter between them. If you want to +use another branch, place the appropriate "-r" flag after the "checkout" +command. This will create a directory ircu2.10, and put all the files in +there. -to get the latest version, from within the tree type "cvs update -dP". +To get the latest version, from within the tree type "cvs update -dP". For more information see the coder-com website at http://coder-com.undernet.org/ The old (tried and true) method that works even when the website isn't being -DoS'd (sigh) is included below. Using the method below means you /cant/ +DoS'd (sigh) is included below. Using the method below means you can't just type 'cvs update -dP' to get the latest version. The name of the package is something like `ircu2.x.y.z.tgz', where @@ -90,25 +97,18 @@ or ircu2.10 if you used cvs. where "ircu2.x.y.z" is the name of the unpacked directory. -3) `./configure' -============== +3) "./configure" +================ -This will generate 'config/setup.h', your Operating System dependend +This will generate "config.h", your operating system-dependent configuration. -If this produces a 'Permission Denied' error message, then try typing -`chmod a+x ./configure` first to give yourself permission to run the file. - -4) `make config' -================ +If this produces a "Permission Denied" error message, then try typing +"chmod a+x ./configure" first to give yourself permission to run the file. -This will (re)generate the include/config.h file. You can run this -as often as you like and it will use your last values as defaults. -At every question you can type a '?' (followed by a return) to get -extensive help, or a 'c' to continue using your old values by default -(quickly finishing the script). +For information on configure command line options, type "./configure --help". -5) `make' +4) "make" ========= Type: @@ -117,25 +117,16 @@ make in the base directory. It should compile without errors or warnings. Please mail any problem to the maintainer, but only AFTER you made sure -you did everything the right way. If you want your Operating System +you did everything the right way. If you want your operating system to be supported in future releases, you best make a patch that actually fixes the problem. -If you have problems here with it complaining about unresolved symbols in -res.o, try "make config" again and add -lresolv to LDFLAGS. Note, there is -no 'e' on the end of resolv. It's not a typo, it's supposed to be like -that. - -If you have problems here with it complaining about unresolved symbols -possibly in s_user.c for 'crypt', add -lcrypt to LDFLAGS. - -6) `make install' +5) "make install" ================= This should install the ircd and the man page. Please recheck the permissions of the binary. -You need to create some of the logfiles that you have chosen by hand -(for instance with 'touch') before the ircd starts writing to them. + Of course, you need a syntactically correct ircd.conf in DPATH. See the docs for some info on this. Also create an ircd.motd with the text of your MOTD. And finally create a remote.motd with three lines of text @@ -148,17 +139,20 @@ In case of problems If you have problems configuring the server you might consider installing GNU make in your PATH. In some cases a brain-dead /bin/sh is causing the -problem, in which case I suggest to install 'bash' and use that (as sh -> bash). -Finally, any other compile problem should be solved when you install gcc. +problem, in which case I suggest to install "bash" and use that (as sh -> +bash). Finally, any other compile problem should be solved when you install +gcc. -If you have problems with starting the ircd, run 'make config' again -and define DEBUGMODE. Recompile the ircd, and run it by hand as: +If you have problems with starting the ircd, run "./configure" again +and give it the "--enable-debug" command line option. Recompile the ircd, +and run it by hand as: ircd -t -x9 This will write debug output to your screen, probably showing why it doesn't start. -Don't use a server with DEBUGMODE defined on a production net. +DO NOT USE A SERVER WITH DEBUGGING ENABLED ON A PRODUCTION NETWORK. Doing +so is a severe privacy risk. If things still don't work, try emailing coder-com@undernet.org