Alan DeKok wrote:
Alan DeKok wrote:
It's about time we have a formal testing process. I have some hosted
machines with spare cycles.
I'll install CruiseControl...
Nope.
After a quick review of continuous integration systems:
.
I think it's easier just to write a few
To: FreeRadius users mailing list freeradius-users@lists.freeradius.org
Cc: a.l.m.bu...@lboro.ac.uk
Sent: Thursday, March 12, 2009 2:58:28 PM
Subject: Re: Version 2.1.4 has been released
Alan DeKok wrote:
Alan DeKok wrote:
It's about time we have a formal testing process. I have some hosted
machines
piston wrote:
You might consider to take a look on the mysql module on freeradius 2.1.4.
I have tried with Debian Lenny + mysql , error: mysql module not found. Same
machine with 2.1.3, no such issue.
Is it so difficult to provide more information?
Like... configure logs, build logs,
I saw that there is a new tar file for the new release so I downloaded
it and tried to build it, alas it won't build because the new tar file
is schizophrenic with multiple identities.
The tar file is named 2.1.4. The tar root directory is named 2.1.4. But
the file
John Dennis wrote:
I saw that there is a new tar file for the new release so I downloaded
it and tried to build it, alas it won't build because the new tar file
is schizophrenic with multiple identities.
I was trying to fix things...
The tar file is named 2.1.4. The tar root directory is
This version comes 3 months after 2.1.3, which is a bit more of a
delay than we would like. However, it includes a number of minor bug
fixes, and some interesting new features.
The best new feature is one that has been needed for a long time. The
(easy) ability to see debugging output from
Am 10.03.2009 um 13:17 schrieb Alan DeKok:
This version comes 3 months after 2.1.3, which is a bit more of a
delay than we would like. However, it includes a number of minor bug
fixes, and some interesting new features.
The best new feature is one that has been needed for a long
time.
Nicolas Goutte wrote:
FreeRADIUS 2.1.4 Thu Dec 25 17:40:00 CEST 2008; , urgency=medium
Just a nitpick: the date above is probaly the one of 2.1.3 (around three
months ago) and not the date of today.
Whoops.. missed that. Oh well. It's not critical.
Alan DeKok.
-
List
Hi,
I couldn't reach bug.freeradius.org, so I'm reporting this here. There
is a typo in src/freeradius.devel/rad_assert.h:
#elsif !defined(FR_SCAN_BUILD)
should be
#elif !defined(FR_SCAN_BUILD)
HTH
-John
Alan DeKok wrote:
This version comes 3 months after 2.1.3, which is a
Alan DeKok wrote:
This version comes 3 months after 2.1.3, which is a bit more of a
delay than we would like. However, it includes a number of minor bug
fixes, and some interesting new features.
Thank you Alan and everyone else for the new release, your community
efforts are very much
Am Dienstag, 10. März 2009 16:32:32 schrieb John Center:
Hi,
I couldn't reach bug.freeradius.org, so I'm reporting this here. There
is a typo in src/freeradius.devel/rad_assert.h:
#elsif !defined(FR_SCAN_BUILD)
should be
#elif !defined(FR_SCAN_BUILD)
HTH
-John
Alan DeKok
Michael Schwartzkopff wrote:
I reported a bug in the create-users.pl create users script. Since
bugs.freeradius.org was not reachable I reported the bug here in the list.
But the solution is not included in the new version.
That fix is pending based on other changes to create-users.pl.
John Dennis wrote:
I noticed a couple of small problems in building the new release. There
was an incorrect #elif in radassert.h which causes the compile to fail
(it was a typo #elsif). The Makefile in etc/raddb contained a dangling
reference otp.conf which is no longer present. I've included
hi,
thanks for the rad_assert pointers etc. still coming a cropper
on another part of the build process:
gmake[6]: Entering directory
`/usr/src/freeradius-server-2.1.4/src/modules/rlm_smsotp'
gmake[7]: Entering directory
`/usr/src/freeradius-server-2.1.4/src/modules/rlm_smsotp'
gmake[7]: ***
a.l.m.bu...@lboro.ac.uk wrote:
thanks for the rad_assert pointers etc. still coming a cropper
on another part of the build process:
Arg. I don't usually build with experimental modules, so I didn't
catch that.
suggest that some of us are called in (like cattle? ;-) ) when the release
is
Hi,
I'll re-spin 2.1.4, unless there are objections.
for reference, i did the old classic 'rm -rf src/modules/rlm_smsotp'
and 'make install' then worked (it was the install part failing with that
message, not the main make process). it built. it runs fine (after
blowing away the old
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
a.l.m.bu...@lboro.ac.uk wrote:
hi,
thanks for the rad_assert pointers etc. still coming a cropper
on another part of the build process:
gmake[6]: Entering directory
`/usr/src/freeradius-server-2.1.4/src/modules/rlm_smsotp'
gmake[7]:
Alan DeKok wrote:
It's about time we have a formal testing process. I have some hosted
machines with spare cycles.
I'll install CruiseControl...
Nope.
After a quick review of continuous integration systems:
- few integrate with git
- most integrate with other build systems (Ant,
Hi,
Is CPPFLAGS used? I see it defined in Make.inc, but I don't see it
actually used. I've been adding it to CFLAGS to make sure it gets included.
Thanks.
-John
Alan DeKok wrote:
This version comes 3 months after 2.1.3, which is a bit more of a
delay than we would like.
John Center wrote:
Is CPPFLAGS used? I see it defined in Make.inc, but I don't see it
actually used. I've been adding it to CFLAGS to make sure it gets
included.
It's not used anywhere.
Alan DeKok.
-
List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users.html
Hi Alan,
Another thing I noticed, if you set WITH_VMPS=no, it isn't consistent:
listen.c, line 1795: undefined symbol: RAD_LISTEN_VQP
cc: acomp failed for listen.c
gmake[4]: *** [listen.lo] Error 1
This code is not surrounded with #defines:
if (this-type ==
Hi,
2009/3/10 Alan DeKok al...@deployingradius.com
The best new feature is one that has been needed for a long time. The
(easy) ability to see debugging output from a live server. You can
now do this via the raddebug command.
Brilliant feature - however I needed to mod the shell
Hi Alan,
Compiling for 64-bit Solaris 10 (SPARC) using Sun Studio 12, see the
following warnings (with appropriate lines):
ttls.c, line 78: warning: integer overflow detected: op
if ((length (1 31)) != 0) {
ttls.c, line 217: warning: integer overflow detected: op
Rupert Finnigan wrote:
Brilliant feature - however I needed to mod the shell script. Path to
radmin was set to
/Users/alandekok/git/2_1_x.git.freeradius.org/src/main/
http://2_1_x.git.freeradius.org/src/main/. Hardly the end of the
world, but thought I'd mention it.
Damn it... that makes
John Center wrote:
Hi Alan,
Another thing I noticed, if you set WITH_VMPS=no, it isn't consistent:
listen.c, line 1795: undefined symbol: RAD_LISTEN_VQP
cc: acomp failed for listen.c
gmake[4]: *** [listen.lo] Error 1
OK. I'll fix that. It shouldn't affect anything else, though.
John Center wrote:
Hi Alan,
Compiling for 64-bit Solaris 10 (SPARC) using Sun Studio 12, see the
following warnings (with appropriate lines):
ttls.c, line 78: warning: integer overflow detected: op
if ((length (1 31)) != 0) {
Hmm... un-typed numbers are int, but that
26 matches
Mail list logo
