qstat-devx — Discussion between QStat developers

[QStat-devx] pariah...

[Ludwig N. <l-...@us...>]

Hi,

To not introduce yet another variable that just makes the mess even
more messy I'd suggest the following patch.

Index: qserver.h
===================================================================
RCS file: /cvsroot/qstat/qstat2/qserver.h,v
retrieving revision 1.3
diff -u -p -r1.3 qserver.h
--- qserver.h	16 Apr 2005 01:19:44 -0000	1.3
+++ qserver.h	24 Apr 2005 10:25:04 -0000
@@ -101,7 +101,6 @@ struct qserver {
     int max_players;
     int num_players;
     int protocol_version;
-	int state;
 
     SavedData saved_data;
 
Index: qstat.c
===================================================================
RCS file: /cvsroot/qstat/qstat2/qstat.c,v
retrieving revision 1.76
diff -u -p -r1.76 qstat.c
--- qstat.c	16 Apr 2005 01:19:44 -0000	1.76
+++ qstat.c	24 Apr 2005 10:25:09 -0000
@@ -7560,6 +7560,7 @@ pariah_rule_packet( struct qserver *serv
 	if ( no_rules == seen )
 	{
 		// all done
+		server->next_rule = NULL;
 		return 1;
 	}
 
@@ -7572,6 +7573,9 @@ ut2003_rule_packet( struct qserver *serv
     char *key, *value;
     int result= 0;
 
+    // Packet Type
+    rawpkt++;
+
 	// we get size encoded key = value pairs
     while ( rawpkt < end )
     {
@@ -7678,7 +7682,10 @@ STATIC int
 pariah_player_packet( struct qserver *server, char *rawpkt, char *end)
 {
 	unsigned char no_players = rawpkt[1];
-	unsigned char seen = 0;
+	unsigned char seen = 0; /* XXX: cannot work this way, it takes only
+				   this packet into consideration. What if
+				   player info is spread across multiple
+				   packets? */
 
 	// type + no_players + some unknown preamble
 	rawpkt += 3;
@@ -7715,6 +7722,7 @@ pariah_player_packet( struct qserver *se
 	if ( no_players == seen )
 	{
 		// all done
+		server->num_players = server->n_player_info;
 		return 1;
 	}
 
@@ -7946,15 +7954,14 @@ deal_with_ut2003_packet( struct qserver 
 			int requests = server->n_requests;
 			server->next_rule = "";
 			server->retry1 = n_retries;
-			server->retry2 = n_retries;
+			server->retry2 = 0;
 			server->n_requests = requests; // would produce wrong ping
 		}
 		break;
 
 	case 0x11:
 		// Game info
-		server->state += pariah_rule_packet( server, rawpkt, end );
-		server->next_rule = "";
+		pariah_rule_packet( server, rawpkt, end );
 		server->retry1 = 0; /* we received at least one rule packet so
 				       no need to retry. We'd get double
 				       entries otherwise. */
@@ -7963,8 +7970,7 @@ deal_with_ut2003_packet( struct qserver 
 	case 0x12:
 		// Player info
 		before = server->n_player_info;
-		server->state += pariah_player_packet( server, rawpkt, end );
-		server->retry2 = 0;
+		pariah_player_packet( server, rawpkt, end );
 		if ( before == server->n_player_info )
 		{
 			error = 1;
@@ -7981,15 +7987,11 @@ deal_with_ut2003_packet( struct qserver 
 	 * rule packets as we don't know how many we get
 	 * We do clean up if we don't fetch server rules so we don't
 	 * need to wait for timeout.
-	 * 2 == server->state indicates finished for pariah as both
-	 * rule and player packets state how many to expect
 	 */
-	if (
-		error ||
-		(!get_server_rules && !get_player_info) ||
-		(!get_server_rules && server->num_players == server->n_player_info) ||
-		2 == server->state 
-	)
+	if (error
+	|| (!get_server_rules && !get_player_info)
+	|| (!get_server_rules && server->num_players == server->n_player_info)
+	|| (server->next_rule == NULL && server->num_players == server->n_player_info))
 	{
 		cleanup_qserver( server, 1 );
 	}

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Re: [qstat-users] qstat 2.8 released

[Steven H. <ki...@mu...>]

That patch works. I've binary win32 to sourceforge.
Steve I've notice there are support requests there but I don't seem
to have sufficient rights to update it could you delegate me
the required rights please :)

    Steve / K
----- Original Message ----- 
From: "Ludwig Nussel" <l-...@us...>
To: <qst...@li...>
Sent: Monday, April 04, 2005 7:59 PM
Subject: Re: [QStat-devx] Re: [qstat-users] qstat 2.8 released


> Steven Hartland wrote:
>> This fails to compile on windows as Makefile it totally incompatible :(
>> Where did the Makefile.w32 go?
> 
> Use Makefile.noauto as described in COMPILE.txt. Now that you
> mentioned it you likely need this patch:
> 
> Index: Makefile.noauto
> ===================================================================
> RCS file: /cvsroot/qstat/qstat2/Makefile.noauto,v
> retrieving revision 1.2
> diff -u -p -r1.2 Makefile.noauto
> --- Makefile.noauto 12 Dec 2004 12:04:46 -0000 1.2
> +++ Makefile.noauto 4 Apr 2005 18:54:32 -0000
> @@ -9,7 +9,7 @@
> ## NOTE: if you get errors when linking qstat (missing symbols or
> ## libraries), then modify LDFLAGS or LDLIBS
> 
> -SRC = config.c debug.c hcache.c md5.c qserver.c qstat.c template.c ut2004.c
> +SRC = config.c debug.c hcache.c md5.c qserver.c qstat.c template.c ut2004.c a2s.c
> OBJ = $(SRC:.c=.obj)
> O   = $(SRC:.c=.o)
> 
> Due to lack of a windows development environment I can't test the
> windows build all the time.


================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Re: [qstat-users] qstat 2.8 released

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> This fails to compile on windows as Makefile it totally incompatible :(
> Where did the Makefile.w32 go?

Use Makefile.noauto as described in COMPILE.txt. Now that you
mentioned it you likely need this patch:

Index: Makefile.noauto
===================================================================
RCS file: /cvsroot/qstat/qstat2/Makefile.noauto,v
retrieving revision 1.2
diff -u -p -r1.2 Makefile.noauto
--- Makefile.noauto	12 Dec 2004 12:04:46 -0000	1.2
+++ Makefile.noauto	4 Apr 2005 18:54:32 -0000
@@ -9,7 +9,7 @@
 ## NOTE: if you get errors when linking qstat (missing symbols or
 ## libraries), then modify LDFLAGS or LDLIBS
 
-SRC = config.c debug.c hcache.c md5.c qserver.c qstat.c template.c ut2004.c
+SRC = config.c debug.c hcache.c md5.c qserver.c qstat.c template.c ut2004.c a2s.c
 OBJ = $(SRC:.c=.obj)
 O   = $(SRC:.c=.o)
 
Due to lack of a windows development environment I can't test the
windows build all the time.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

[QStat-devx] Re: [qstat-users] qstat 2.8 released

[Steven H. <ki...@mu...>]

This fails to compile on windows as Makefile it totally incompatible :(
Where did the Makefile.w32 go?
  ----- Original Message -----=20
  From: Ludwig Nussel=20
  To: qst...@ya...=20
  Sent: Sunday, April 03, 2005 9:36 PM
  Subject: [qstat-users] qstat 2.8 released


  Hi,

  Since there doesn't seem to be a lot of people testing the cvs
  version and I don't know how much time I have the next weeks I
  decided to do the 2.8 release now rather than to defer it
  indefinitely:

  http://sourceforge.net/project/showfiles.php?group_id=3D56603&package_id=3D51888&release_id=3D317684

  Binary packages for Linux/x86 are available from the xqf site as
  ususal.

  cu
  Ludwig

  --=20
  (o_  Lud...@gm...
  //\
  V_/_ PGP Key ID: FF8135CE


------------------------------------------------------------------------------
  Yahoo! Groups Links

    a.. To visit your group on the web, go to:
    http://groups.yahoo.com/group/qstat-users/
     =20
    b.. To unsubscribe from this group, send an email to:
    qst...@ya...
     =20
    c.. Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.=20



=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it.=20

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Re: [qstat-users] test build for 2.8

[Steven H. <ki...@mu...>]

----- Original Message ----- 
From: "Ludwig Nussel" <l-...@us...>

> Do you know when it's going to change? Does the old protocol stop
> working? If so it doesn't make much sense to release without it
> indeed. Especially sinc hl2 support is new in xqf 1.0.3...

Two weeks iirc

    Steve / K



================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Re: [qstat-users] test build for 2.8

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> Few ./autogen.sh warnings here:
> configure.ac:3: warning: do not use m4_patsubst: use patsubst or m4_bpatsubst
> configure.ac:39: warning: do not use m4_regexp: use regexp or m4_bregexp

Looks like your autoconf doesn't like itself.

> I would like to get the new Source protocol in before 2.8 is released.
> Is anyone looking at this? I'm very stretched for time atm due to a big event
> coming up in two weeks and me being away on hols for the past 2.
> New docs for source:
> http://www.valve-erc.com/srcsdk/Code/Networking/serverqueries.html

Do you know when it's going to change? Does the old protocol stop
working? If so it doesn't make much sense to release without it
indeed. Especially sinc hl2 support is new in xqf 1.0.3...

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

[QStat-devx] Re: [qstat-users] test build for 2.8

[Steven H. <ki...@mu...>]

Few ./autogen.sh warnings here:
configure.ac:3: warning: do not use m4_patsubst: use patsubst or m4_bpatsubst
configure.ac:39: warning: do not use m4_regexp: use regexp or m4_bregexp

I would like to get the new Source protocol in before 2.8 is released.
Is anyone looking at this? I'm very stretched for time atm due to a big event
coming up in two weeks and me being away on hols for the past 2.
New docs for source:
http://www.valve-erc.com/srcsdk/Code/Networking/serverqueries.html

    Steve / K
  ----- Original Message -----=20
  From: Ludwig Nussel=20
  To: qst...@ya...=20
  Sent: Monday, March 07, 2005 8:39 PM
  Subject: [qstat-users] test build for 2.8


  Hi,

  I'd like to release XQF 1.0.3 and qstat 2.8 soon. A test build of
  the current cvs of both is available at
  http://xqf.sourceforge.net/xqf-1.0.2.1/ if anyone wants to
  beta-test. Feedback welcome.

  cu
  Ludwig

  --=20
  (o_  Lud...@gm...
  //\
  V_/_ PGP Key ID: FF8135CE

        Yahoo! Groups Sponsor=20
              ADVERTISEMENT
            =20
      =20
      =20


------------------------------------------------------------------------------
  Yahoo! Groups Links

    a.. To visit your group on the web, go to:
    http://groups.yahoo.com/group/qstat-users/
     =20
    b.. To unsubscribe from this group, send an email to:
    qst...@ya...
     =20
    c.. Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.=20



=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it.=20

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

[QStat-devx] Known Bugs of 2.7

[Ludwig N. <l-...@us...>]

Hi,

*sigh*

- version string in qstat.c is wrong therefore compilation on
  Windows shows 2.6.CVS instead of 2.7
- ut2004 master support is broken because a timeout is triggered too
  early.

The latter one is quite annoying as I consider it the main feature
of 2.7. I have no idea why it worked in in the first place but I
swear it did work. We should probably release a bugfix release soon.

cu
Ludwig

PS: is Pierre (webmaster of qstat.org) on this list?

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] release?

[Ludwig N. <l-...@us...>]

Ludwig Nussel wrote:
> No feedback at all from the beta list. Are we ready to release 2.7?

No objections? So I'll release qstat 2.7 tomorrow then.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

[QStat-devx] release?

[Ludwig N. <l-...@us...>]

Hi,

No feedback at all from the beta list. Are we ready to release 2.7?

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> ----- Original Message ----- 
> From: "Ludwig Nussel" <l-...@us...>
> >>But we're not doing releases.  And we're not doing beta releases.  So use 
> >>of
> >>the current CVS is now serving as a beta test.  But we can't get any beta
> >>feedback if people can't compile it.
> >
> >Why not? 2.6 is from August, time for a 2.7 release.
> 
> I think what Steve's trying to say is that unless we do interim releases for
> each patch applied then we are going to lack important feedback from
> people who would usually compile from CVS quite easily. Given it took me
> the best part of a day with your help to get it compiling it doesn't bode 
> well.

That was an initial hurdle, it's fixed now.

> I suggest what we do is reinstate the old Makefile into the CVS
> tree but when we do a release it gets excluded and your new

We can just as well leave it in then. We need a Makefile for W32
anyways. I'll add it again.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Steven H. <ki...@mu...>]

----- Original Message ----- 
From: "Ludwig Nussel" <l-...@us...>
>> But we're not doing releases.  And we're not doing beta releases.  So use of
>> the current CVS is now serving as a beta test.  But we can't get any beta
>> feedback if people can't compile it.
> 
> Why not? 2.6 is from August, time for a 2.7 release.

I think what Steve's trying to say is that unless we do interim releases for
each patch applied then we are going to lack important feedback from
people who would usually compile from CVS quite easily. Given it took me
the best part of a day with your help to get it compiling it doesn't bode well.

I suggest what we do is reinstate the old Makefile into the CVS tree but when
we do a release it gets excluded and your new configure based system shipped.
This will then have all benefits that you stated with none of the pitfalls. The
only slight downside is that the old Makefile will need to be maintained but
given its simplicity that shouldn't be a problem.

    Steve / K


================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

Steve Jankowski wrote:
> 
> --- Ludwig Nussel <l-...@us...> wrote:
> 
> > [there is a massive delay of several hours from this list whereas
> > the mails directly addressed to me arrived instantly]
> > 
> > Steve Jankowski wrote:
> > > Yeah, I thought the 10 line makefile was just fine.  But the autoconf
> > benefits
> > > you list are some what compelling.  They aren't enough for me, given the
> > head
> > > aches you've encountered; I've got zero patience for unreliable tools.
> > > 
> > > I suggest restoring the simple works-on-most-unixes makefile.  The fewer
> > > barriers to entry, the more successful QStat will be.  Autoconf will be
> > nice
> > > for the qstat developers (assuming all the developers can get it working),
> > but
> > > we shouldn't expose it to users.
> > 
> > Users of release tarballs don't see it. They just run the usual
> > ./configure && make install. The problems with older automake are
> > hopefully fixed now.
> 
> But we're not doing releases.  And we're not doing beta releases.  So use of
> the current CVS is now serving as a beta test.  But we can't get any beta
> feedback if people can't compile it.

Why not? 2.6 is from August, time for a 2.7 release.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Steve J. <ste...@ya...>]

--- Ludwig Nussel <l-...@us...> wrote:

> [there is a massive delay of several hours from this list whereas
> the mails directly addressed to me arrived instantly]
> 
> Steve Jankowski wrote:
> > Yeah, I thought the 10 line makefile was just fine.  But the autoconf
> benefits
> > you list are some what compelling.  They aren't enough for me, given the
> head
> > aches you've encountered; I've got zero patience for unreliable tools.
> > 
> > I suggest restoring the simple works-on-most-unixes makefile.  The fewer
> > barriers to entry, the more successful QStat will be.  Autoconf will be
> nice
> > for the qstat developers (assuming all the developers can get it working),
> but
> > we shouldn't expose it to users.
> 
> Users of release tarballs don't see it. They just run the usual
> ./configure && make install. The problems with older automake are
> hopefully fixed now.
> 
> cu
> Ludwig
> 

But we're not doing releases.  And we're not doing beta releases.  So use of
the current CVS is now serving as a beta test.  But we can't get any beta
feedback if people can't compile it.

Steve

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

[there is a massive delay of several hours from this list whereas
the mails directly addressed to me arrived instantly]

Steve Jankowski wrote:
> Yeah, I thought the 10 line makefile was just fine.  But the autoconf benefits
> you list are some what compelling.  They aren't enough for me, given the head
> aches you've encountered; I've got zero patience for unreliable tools.
> 
> I suggest restoring the simple works-on-most-unixes makefile.  The fewer
> barriers to entry, the more successful QStat will be.  Autoconf will be nice
> for the qstat developers (assuming all the developers can get it working), but
> we shouldn't expose it to users.

Users of release tarballs don't see it. They just run the usual
./configure && make install. The problems with older automake are
hopefully fixed now.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Steve J. <ste...@ya...>]

Yeah, I thought the 10 line makefile was just fine.  But the autoconf benefits
you list are some what compelling.  They aren't enough for me, given the head
aches you've encountered; I've got zero patience for unreliable tools.

I suggest restoring the simple works-on-most-unixes makefile.  The fewer
barriers to entry, the more successful QStat will be.  Autoconf will be nice
for the qstat developers (assuming all the developers can get it working), but
we shouldn't expose it to users.

Steve


--- Ludwig Nussel <l-...@us...> wrote:

> Steven Hartland wrote:
> > Nope autoconf doesnt seem to be a standard tool so I'd say we best
> > avoid it:
> > 
> > steven@besson [steven] autoconf
> > bash: autoconf: command not found
> > 
> > root@dev1:installs> autoconf
> > autoconf: Command not found.
> > 
> > root@yoda:qstat2> ./autogen.sh 
> > + which autoreconf
> > + aclocal
> > aclocal: not found
> > + autoconf
> > autoconf: not found
> > + autoheader
> > autoheader: not found
> > + automake
> 
> automake/autoconf are standard GNU development tools and they are
> probably just as annoying as useful sometimes. The reasons I want it
> for are at least, make dist, make distcheck, the standard install
> target and the automatic dependency rules it generates. When I asked
> last month the only concern was that it may not work on Windows,
> therefore I added a simple Makefile.w32 for that. Note that you do
> not need automake or autoconf if you compile the released version.
> Only the CVS doesn't contain configure etc since those files are
> automatically generated.
> 
> cu
> Ludwig
> 
> -- 
> (o_     Ludwig Nussel
> //\
> V_/_  PGP Key: FF8135CE
> 
> 
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now. 
> http://productguide.itmanagersjournal.com/
> _______________________________________________
> QStat-devx mailing list
> QSt...@li...
> https://lists.sourceforge.net/lists/listinfo/qstat-devx
> 

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <lud...@we...>]

Steven Hartland wrote:
> Still no joy:
> ./autogen.sh
> + which autoreconf
> + autoreconf -i
> autoreconf: `aclocal.m4' is updated
> exec: autom4te --language=autoconf --output=-: not found
> automake: configure.ac: `AM_INIT_AUTOMAKE' must be used
> automake: Makefile.am: required file `./NEWS' not found
> automake: Makefile.am: required file `./README' not found
> automake: Makefile.am: required file `./AUTHORS' not found
> configure.ac: 3: required file `./gnuconfig.h.in' not found
> autoreconf: automake failed with exit status: 1
> at /usr/local/bin/autoreconf line 368

gnuconfig.h.in is supposed to be created automatically. Maybe it got
confused by previous errors. Try autoreconf -f -i or start with a
fresh checkout.

cu
Ludwig

-- 
(o_  Lud...@we...
//\
V_/_ PGP Key ID: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Steven H. <ki...@mu...>]

Still no joy:
./autogen.sh
+ which autoreconf
+ autoreconf -i
autoreconf: `aclocal.m4' is updated
exec: autom4te --language=autoconf --output=-: not found
automake: configure.ac: `AM_INIT_AUTOMAKE' must be used
automake: Makefile.am: required file `./NEWS' not found
automake: Makefile.am: required file `./README' not found
automake: Makefile.am: required file `./AUTHORS' not found
configure.ac: 3: required file `./gnuconfig.h.in' not found
autoreconf: automake failed with exit status: 1
 at /usr/local/bin/autoreconf line 368


    Steve / K
----- Original Message ----- 
From: "Ludwig Nussel" <l-...@us...>
 
> That's the annoying side of automake. I don't have a system with
> such an old version so I can't check compatability myself. I've
> changed AC_CONFIG_HEADER into AM_CONFIG_HEADER, please try again. If
> automake still complains about required files not found, remove the
> autoreconf call and add --foreign to automake.


================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> Also installing automake still gives errors:
> 
> root@yoda:qstat2> ./autogen.sh
> + which autoreconf
> + autoreconf -i
> autoreconf: `aclocal.m4' is updated
> exec: autom4te --language=autoconf --output=-: not found
> configure.ac: 3: `automake requires `AM_CONFIG_HEADER', not 
> `AC_CONFIG_HEADER'
> automake: configure.ac: `AM_INIT_AUTOMAKE' must be used
> automake: configure.ac: installing `./install-sh'
> automake: configure.ac: installing `./mkinstalldirs'
> automake: configure.ac: installing `./missing'
> automake: Makefile.am: installing `./INSTALL'
> automake: Makefile.am: required file `./NEWS' not found
> automake: Makefile.am: required file `./README' not found
> automake: Makefile.am: installing `./COPYING'
> automake: Makefile.am: required file `./AUTHORS' not found
> configure.ac: 3: required file `./gnuconfig.h.in' not found
> automake: configure.ac: installing `./depcomp'
> autoreconf: automake failed with exit status: 1
> at /usr/local/bin/autoreconf line 368
> 
> pkg_info |grep auto
> autoconf-2.53_1     Automatically configure source code on many Un*x 
> platforms
> automake-1.5,1      GNU Standards-compliant Makefile generator
> 
> This is not convincing me this is a worth while exercise :(

That's the annoying side of automake. I don't have a system with
such an old version so I can't check compatability myself. I've
changed AC_CONFIG_HEADER into AM_CONFIG_HEADER, please try again. If
automake still complains about required files not found, remove the
autoreconf call and add --foreign to automake.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> Hmm not had to download said tools for building anything else and I
> build almost all things from source which makes me question how
> standard it is?

If you ever had to enter ./configure before make you have used it's
service. You do not need to autoconf/automake installed as long as
you don't modify Makefile.am&friends.

> Not sure how helpful it is not being able to compile from CVS
> without installing additional tools especially as there is no mention
> of it in the docs

Well, that's easy to fix.

> Also just installed autoconf and it still fails:
> ./autogen.sh 
> + which autoreconf
> + autoreconf -i
> Can't exec "aclocal": No such file or directory at 
> /usr/local/share/autoconf/Autom4te/General.pm line 459.
> autoreconf: aclocal failed with exit status: 16777215
> 
> pkg_info |grep autoc
> autoconf-2.53_1   Automatically configure source code on many Un*x platforms

You also need automake.

> Seems like a bit to much pain to ask people to go through just to compile
> the CVS version when a 10 line Makefile did the job nicely.

It lacked the useful features I mentioned. Ideally people don't
compile from CVS. We need to do releases more frequently instead.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Steven H. <ki...@mu...>]

Also installing automake still gives errors:

root@yoda:qstat2> ./autogen.sh
+ which autoreconf
+ autoreconf -i
autoreconf: `aclocal.m4' is updated
exec: autom4te --language=autoconf --output=-: not found
configure.ac: 3: `automake requires `AM_CONFIG_HEADER', not `AC_CONFIG_HEADER'
automake: configure.ac: `AM_INIT_AUTOMAKE' must be used
automake: configure.ac: installing `./install-sh'
automake: configure.ac: installing `./mkinstalldirs'
automake: configure.ac: installing `./missing'
automake: Makefile.am: installing `./INSTALL'
automake: Makefile.am: required file `./NEWS' not found
automake: Makefile.am: required file `./README' not found
automake: Makefile.am: installing `./COPYING'
automake: Makefile.am: required file `./AUTHORS' not found
configure.ac: 3: required file `./gnuconfig.h.in' not found
automake: configure.ac: installing `./depcomp'
autoreconf: automake failed with exit status: 1
 at /usr/local/bin/autoreconf line 368

pkg_info |grep auto
autoconf-2.53_1     Automatically configure source code on many Un*x platforms
automake-1.5,1      GNU Standards-compliant Makefile generator

This is not convincing me this is a worth while exercise :(

    Steve / K

----- Original Message ----- 
From: "Steven Hartland" <ki...@mu...>
To: "Ludwig Nussel" <l-...@us...>; <qst...@li...>
Sent: Tuesday, November 23, 2004 6:36 PM
Subject: Re: [QStat-devx] Current CVS how to build?


> Hmm not had to download said tools for building anything else and I
> build almost all things from source which makes me question how
> standard it is?
> Not sure how helpful it is not being able to compile from CVS
> without installing additional tools especially as there is no mention
> of it in the docs
> 
> Also just installed autoconf and it still fails:
> ./autogen.sh 
> + which autoreconf
> + autoreconf -i
> Can't exec "aclocal": No such file or directory at /usr/local/share/autoconf/Autom4te/General.pm line 459.
> autoreconf: aclocal failed with exit status: 16777215
> 
> pkg_info |grep autoc
> autoconf-2.53_1   Automatically configure source code on many Un*x platforms
> 
> Seems like a bit to much pain to ask people to go through just to compile
> the CVS version when a 10 line Makefile did the job nicely.



================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Current CVS how to build?

[Steven H. <ki...@mu...>]

Hmm not had to download said tools for building anything else and I
build almost all things from source which makes me question how
standard it is?
Not sure how helpful it is not being able to compile from CVS
without installing additional tools especially as there is no mention
of it in the docs

Also just installed autoconf and it still fails:
./autogen.sh 
+ which autoreconf
+ autoreconf -i
Can't exec "aclocal": No such file or directory at /usr/local/share/autoconf/Autom4te/General.pm line 459.
autoreconf: aclocal failed with exit status: 16777215

pkg_info |grep autoc
autoconf-2.53_1   Automatically configure source code on many Un*x platforms

Seems like a bit to much pain to ask people to go through just to compile
the CVS version when a 10 line Makefile did the job nicely.

    Steve / K
----- Original Message ----- 
From: "Ludwig Nussel" <l-...@us...>
> 
> automake/autoconf are standard GNU development tools and they are
> probably just as annoying as useful sometimes. The reasons I want it
> for are at least, make dist, make distcheck, the standard install
> target and the automatic dependency rules it generates. When I asked
> last month the only concern was that it may not work on Windows,
> therefore I added a simple Makefile.w32 for that. Note that you do
> not need automake or autoconf if you compile the released version.
> Only the CVS doesn't contain configure etc since those files are
> automatically generated.


================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> Nope autoconf doesnt seem to be a standard tool so I'd say we best
> avoid it:
> 
> steven@besson [steven] autoconf
> bash: autoconf: command not found
> 
> root@dev1:installs> autoconf
> autoconf: Command not found.
> 
> root@yoda:qstat2> ./autogen.sh 
> + which autoreconf
> + aclocal
> aclocal: not found
> + autoconf
> autoconf: not found
> + autoheader
> autoheader: not found
> + automake

automake/autoconf are standard GNU development tools and they are
probably just as annoying as useful sometimes. The reasons I want it
for are at least, make dist, make distcheck, the standard install
target and the automatic dependency rules it generates. When I asked
last month the only concern was that it may not work on Windows,
therefore I added a simple Makefile.w32 for that. Note that you do
not need automake or autoconf if you compile the released version.
Only the CVS doesn't contain configure etc since those files are
automatically generated.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE

Re: [QStat-devx] Current CVS how to build?

[Steven H. <ki...@mu...>]

Nope autoconf doesnt seem to be a standard tool so I'd say we best
avoid it:

steven@besson [steven] autoconf
bash: autoconf: command not found

root@dev1:installs> autoconf
autoconf: Command not found.

root@yoda:qstat2> ./autogen.sh 
+ which autoreconf
+ aclocal
aclocal: not found
+ autoconf
autoconf: not found
+ autoheader
autoheader: not found
+ automake

    Steve / K

----- Original Message ----- 
From: "Ludwig Nussel" <l-...@us...>


> Steven Hartland wrote:
>> ./autogen.sh 
>> + autoreconf -i
>> autoreconf: not found
>> 
>> Seems like your relying on something that's pretty none standard.
> 
> autoreconf is part of autoconf. Maybe you have an older version.
> I've changed autogen.sh to call the tools individually if autoreconf
> is not available.


================================================
This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. 

In the event of misdirection, illegible or incomplete transmission please telephone (023) 8024 3137
or return the E.mail to pos...@mu....

Re: [QStat-devx] Current CVS how to build?

[Ludwig N. <l-...@us...>]

Steven Hartland wrote:
> ./autogen.sh 
> + autoreconf -i
> autoreconf: not found
> 
> Seems like your relying on something that's pretty none standard.

autoreconf is part of autoconf. Maybe you have an older version.
I've changed autogen.sh to call the tools individually if autoreconf
is not available.

cu
Ludwig

-- 
(o_     Ludwig Nussel
//\
V_/_  PGP Key: FF8135CE