log4cpp-devel — technical discussions about the development of log4cpp

[Log4cpp-devel] log4cpp timestamp

[Monica L. <mll...@ya...>]

Hi all,
 
I'm very new to the log4cpp library, and I was just wondering if the library provides a function to translate the timestamp into the readable format that is provided by the "date" command?  For example, log4cpp logs the following:
 
995871335 ERROR main_cat: This is an error.
 
Instead of this, can log4cpp do this:
Thu Apr 8 18:17:38 PDT 2004 ERROR main_cat: This is an error.
 
Thanks for your time and help.
 
Regards,
Monica


---------------------------------
Do you Yahoo!?
Yahoo! Small Business $15K Web Design Giveaway - Enter today

[Log4cpp-devel] Can't make log4cpp-0.2.8 on a Windows machine

[Dimitri P. <dim...@gm...>]

Hello!

I'm trying to make log4cpp-0.2.8 on a Windows machine with cygwin.

configure works, but make results in an error:
<make-output-fragment>
Making all in tests
make[1]: Entering directory
`/cygdrive/c/work/daoes/sw-thirdparty/log4cpp-0.2.8/
tests'
source=3D'testmain.cpp' object=3D'testmain.o' libtool=3Dno \
depfile=3D'.deps/testmain.Po' tmpdepfile=3D'.deps/testmain.TPo' \
depmode=3Dgcc3 /bin/bash ../config/depcomp \
g++ -DHAVE_CONFIG_H -I. -I. -I../include -I../include    -g -O2 -Wall
-pedantic
-c -o testmain.o `test -f 'testmain.cpp' || echo './'`testmain.cpp
/bin/bash ../libtool --mode=3Dlink g++  -g -O2 -Wall -pedantic   -o
testmain.exe
testmain.o ../src/liblog4cpp.la
mkdir .libs
g++ -g -O2 -Wall -pedantic -o testmain.exe testmain.o
../src/.libs/liblog4cpp.a

source=3D'Clock.cpp' object=3D'Clock.o' libtool=3Dno \
depfile=3D'.deps/Clock.Po' tmpdepfile=3D'.deps/Clock.TPo' \
depmode=3Dgcc3 /bin/bash ../config/depcomp \
g++ -DHAVE_CONFIG_H -I. -I. -I../include -I../include    -g -O2 -Wall
-pedantic
-c -o Clock.o `test -f 'Clock.cpp' || echo './'`Clock.cpp
In file included from Clock.cpp:17:
Clock.hh:13: error: ISO C++ does not support `long long'
Clock.cpp:21:53: warning: use of C99 long long integer constant
make[1]: *** [Clock.o] Error 1
make[1]: Leaving directory
`/cygdrive/c/work/daoes/sw-thirdparty/log4cpp-0.2.8/t
ests'
make: *** [all-recursive] Error 1
</make-output-fragment>

How can I solve this problem?

TIA

dap

RE: [Log4cpp-devel] Compiling log4cpp on Mingw32.

[<Jak...@no...>]

Hi,

Try 0.3.4 beta.

It is quite stable.

cheers,
--jakub


-----Original Message-----
From: log...@li...
[mailto:log...@li...]On Behalf Of ext
Joshua Little
Sent: Friday, February 06, 2004 2:39 AM
To: log...@li...
Subject: [Log4cpp-devel] Compiling log4cpp on Mingw32.


Hi all,
I noticed that there hasn't been a whole lot of activity on the list in=20
a while but I hope there are still some people sticking around. I=20
downloaded log4cpp and tried to compile it using Mingw32 with gcc3.2 and =

MSYS. I ran configure and when I was running make and got the error :
In file included from ../include/log4cpp/Appender.hh:21,
                 from Appender.cpp:11:
../include/log4cpp/Priority.hh:32:2: #error Naming collision for 'ERROR' =

detected. Please read the FAQ for a workaround.

So I looked it up in the faq and I'm not quite sure how to apply the fix =

for this. If I'm reading this right I need to go through all of log4cpp=20
and put in that #define LOG4CPP_FIX_ERROR_COLLISION 1? I'm trying to=20
compile 0.3.3, do I need to try the 0.3.4 beta version maybe?
Thanks,
Joshua.



-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Log4cpp-devel mailing list
Log...@li...
https://lists.sourceforge.net/lists/listinfo/log4cpp-devel

[Log4cpp-devel] Compiling log4cpp on Mingw32.

[Joshua L. <lj...@sw...>]

Hi all,
I noticed that there hasn't been a whole lot of activity on the list in 
a while but I hope there are still some people sticking around. I 
downloaded log4cpp and tried to compile it using Mingw32 with gcc3.2 and 
MSYS. I ran configure and when I was running make and got the error :
In file included from ../include/log4cpp/Appender.hh:21,
                 from Appender.cpp:11:
../include/log4cpp/Priority.hh:32:2: #error Naming collision for 'ERROR' 
detected. Please read the FAQ for a workaround.

So I looked it up in the faq and I'm not quite sure how to apply the fix 
for this. If I'm reading this right I need to go through all of log4cpp 
and put in that #define LOG4CPP_FIX_ERROR_COLLISION 1? I'm trying to 
compile 0.3.3, do I need to try the 0.3.4 beta version maybe?
Thanks,
Joshua.

[Log4cpp-devel] Home page - Bad link to "Recommended reading"

[Ahmed <ah...@ba...>]

Hi,

On the home page,
the link to the kind of log4cpp-HOWTO available at Solaris Developer  
Connection leads nowhere.

"http://soldc.sun.com/articles/logging.html/"
is to be replaced by
"http://developers.sun.com/solaris/articles/logging.html".

[Log4cpp-devel] the state and future of log4cpp

[Bastiaan B. <bas...@li...>]

Hi log4cpp developers and users,

Lately people have been wondering whether log4cpp is still an active
project. 
So, is the log4cpp project dead?
No, no, no, it's not dead, it's, uhm, dormant, yeah that's it! It's just
resting for a while....

The are two reasons for this 'sleep':

1) log4cpp has matured enough to satify the logging needs for our
inhouse software development at LifeLine Networks. So, at work I don't
have the time to work on improving log4cpp. Especially since most C++
programs we have will be phased out in favour of Java implementations
(using log4j).

2) In the past few years I've acquired a distinct dislike to C++ for
robust platform independent programing. I won't bother you with a long
rant on what sucks about C++ (I suppose you already know), but for me it
has taken away the joy in working on log4cpp (and other C++ projects).

In short both fun and profit have gone, and the itch as well...

So what's next?
Well, one option would be to simply let the project fade away and die,
and have everyone migrate to log4cplus or log4cxx.
Of course there is another option: one or more log4cpp developers or
users will enthousastically jump forward to take over maintainership of
the project. Yes, you could be the lucky winner! Reply now! :-)

So, if your interested, just apply for the job through the mailing list
to have everything as transparent as possible. Preference will be given
to current project members.

Cheers,

Bastiaan Bakker
resigning project admin for log4cpp

PS.
I'm not leaving earth, so I will be reachable for the new maintainer to
help getting things going again.

PPS.
Recently Apache has formed the Logging Service Project
(_http://logging.apache.org/_). The log4cpp project has been invited to
participate in this project. The current LGPL license of log4cpp is
incompatible with the Apache license of the project. However if a new
maintainer would like to have log4cpp join the Logging Service Project,
LifeLine Networks and me are willing to relicense log4cpp under the
Apache license. NB. Even though we do hold copyrights to the bulk of the
code, some parts were contributed by others, retaining their copyrights,
so you'll need to get their permission as well. (grepping for
'copyright' in the sources should get you a list of other copyright
holders).

Re: [Log4cpp-devel] Is this list dead?

[Simon J. J. <sj...@er...>]

Hi Tony,

Thanks for your reply.

First off I should point out that I use log4cpp in some projects already 
and I like it a great deal. I think it serves an extremely valuable role 
for C++ programs.

However, I am really concerned about the state of the project for the 
following two reasons:

1. There does not seem to be any active or even routine support going on. 
It might be the case that log4cpp's feature set is largely complete, but 
there are 30 or so open bugs on the bug list (some assigned, some not) as 
well as 13 support requests, 7 patches and 7 feature requests. Almost all 
of these are more than 6 months old and some changes, particularly for MSVC 
compilers, appear to be critical.

2. Far and away the biggest problem I've had with log4cpp is the fact that 
it has virtually no documentation. Although I've been learning it slowly 
(mostly from log4j), this really limits peoples' understanding of the 
capabilities of the library.

Do you know if there are there any plans to address these issues?

Cheers,

Simon



#################################################################
#################################################################
#################################################################
#####
#####
#####
#################################################################
#################################################################
#################################################################

Re: [Log4cpp-devel] Is this list dead?

[Tony Yat-T. C. <dra...@as...>]

Hi,

I am not if there is any active development on the log4cpp.

However, given log4cpp is a logging class for C++, it is already very 
mature and is usable in any commercial deployment.

If you run into any problem with using log4cpp, I am sure someone in 
this list would be glad to answer you.

Regards,
Tony Cheung

Simon J. Julier wrote:

> Given the lack of activity in the project area, the CVS and responses 
> to queries on this list, is it fair to assume that log4cpp is not been 
> actively worked on or maintained?
>
> Cheers,
>
> Simon
>
>
>
> -------------------------------------------------------
> The SF.Net email is sponsored by EclipseCon 2004
> Premiere Conference on Open Tools Development and Integration
> See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
> http://www.eclipsecon.org/osdn
> _______________________________________________
> Log4cpp-devel mailing list
> Log...@li...
> https://lists.sourceforge.net/lists/listinfo/log4cpp-devel

[Log4cpp-devel] Is this list dead?

[Simon J. J. <sj...@er...>]

Given the lack of activity in the project area, the CVS and responses to 
queries on this list, is it fair to assume that log4cpp is not been 
actively worked on or maintained?

Cheers,

Simon

[Log4cpp-devel] a few questions

[<al...@cw...>]

Hi,

I'm new to this list, and a little new to log4cplus...

I had one weird problem with log4cplus - whenever i created an appender
(rolling file - but i don't thing it matters which) with a name longer
than 15 chars, and tried to print it's name:
printf("name %s\n", logger.getName().c_str());

I got an exception, why is that? Is there some way to change the "max
appender name"?

And i was wondering if a rolling file appender can also be periodical?
Is there an appender which can be both daily and limited to size. It
doesn't seem complicated to write one, but maybe there is one already.

And another question is: is there an appender that writes to the same
place as windows's OutputDebugString() ?

Thanks a lot.
Alice :)

[Log4cpp-devel] Boost thread support and getThreadId

[Simon J. J. <sj...@er...>]

I'm using log4cpp with BoostThreads under linux, cygwin and native win32 
(after manually hacking the configuration files). However, I get large 
numbers of "getThreadId function defined but not used" warnings when 
compiling with gcc because the function body is declared in the header 
file. Following the pattern from PThreads and OmniThreads, would it be 
possible to move the function body into a new source file? I have done the 
following, which works on all of my target platforms:

1. In threading/BoostThreads.hh remove the function
body and only leave the declaration:

std::string getThreadId();

2. Create a new file (e.g., BoostThreads.cpp) with the contents:

#include <log4cpp/threading/Threading.hh>

#if defined(LOG4CPP_HAVE_THREADING) &&
defined(LOG4CPP_USE_BOOSTTHREADS)

namespace log4cpp {
namespace threading {

std::string getThreadId() {
return std::string("not implemented");
};
}
}

#endif // LOG4CPP_HAVE_THREADING &&
LOG4CPP_USE_BOOSTTHREADS

Cheers,

Simon

P. S. I had originally posted this on the patches list on sourceforge but 
it looks like that list hasn't been touched for over a year. Is that correct?

Re: [Log4cpp-devel] no rootCategory?

[Aaron I. <ai...@ya...>]

setAdditivity(false)

--- Scott M <sm...@ya...> wrote:
> Is it possible to have no rootCategory? If not, can it at least be a
> noop
> (don't log anything)?
> 
> Thanks,
> 
> - Scott
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: SF.net Giveback Program.
> Does SourceForge.net help you be more productive?  Does it
> help you create better code?   SHARE THE LOVE, and help us help
> YOU!  Click Here: http://sourceforge.net/donate/
> _______________________________________________
> Log4cpp-devel mailing list
> Log...@li...
> https://lists.sourceforge.net/lists/listinfo/log4cpp-devel


__________________________________
Do you Yahoo!?
Exclusive Video Premiere - Britney Spears
http://launch.yahoo.com/promos/britneyspears/

[Log4cpp-devel] no rootCategory?

[Scott M <sm...@ya...>]

Is it possible to have no rootCategory? If not, can it at least be a noop
(don't log anything)?

Thanks,

- Scott

[Log4cpp-devel] [Fwd: Log4cpp Visual Studio .NET support]

[Bastiaan B. <bas...@li...>]

Hi,

Tony Novak sent me this archive with project files to enable building of
log4cpp with VS .NET 2002 and 2003. Anyone interested in checking it and
adding it to CVS?

Cheers,

Bastiaan

Re: [Log4cpp-devel] PropertyConfigurator from within library?

[Scott M <sm...@ya...>]

Uh, anyone out there?

- Scott

--- Scott M <sm...@ya...> wrote:
> I'm building a library (UNIX .so, but that shouldn't matter) which uses
> log4cpp. Clearly I want my library to work with old apps which don't use
> log4cpp, as well as new log4cpp-based apps.
> 
> How do I configure log4cpp for my library? I'd like to have a separate
> lib.conf file, which I load at lib load time. Problem is it looks like
> PropertyConfigurator is static (global) per process, so any other
> libraries
> or the app itself has the potential of overwritting my settings. Is this
> true?
> 
> Do multiple PropertyConfigurator::configure() calls (within the same
> process) complement each other? That is, they are additive, so each
> component can configure itself without fear of corrupting other
> components'
> log4cpp setup?
> 
> Thanks,
> 
> - Scott
> 
> __________________________________
> Do you Yahoo!?
> Yahoo! SiteBuilder - Free, easy-to-use web site design software
> http://sitebuilder.yahoo.com
> 
> 
> -------------------------------------------------------
> This SF.Net email sponsored by: Free pre-built ASP.NET sites including
> Data Reports, E-commerce, Portals, and Forums are available now.
> Download today and enter to win an XBOX or Visual Studio .NET.
>
http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01
> _______________________________________________
> Log4cpp-devel mailing list
> Log...@li...
> https://lists.sourceforge.net/lists/listinfo/log4cpp-devel


__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com

[Log4cpp-devel] PropertyConfigurator from within library?

[Scott M <sm...@ya...>]

I'm building a library (UNIX .so, but that shouldn't matter) which uses
log4cpp. Clearly I want my library to work with old apps which don't use
log4cpp, as well as new log4cpp-based apps.

How do I configure log4cpp for my library? I'd like to have a separate
lib.conf file, which I load at lib load time. Problem is it looks like
PropertyConfigurator is static (global) per process, so any other libraries
or the app itself has the potential of overwritting my settings. Is this
true?

Do multiple PropertyConfigurator::configure() calls (within the same
process) complement each other? That is, they are additive, so each
component can configure itself without fear of corrupting other components'
log4cpp setup?

Thanks,

- Scott

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com

[Log4cpp-devel] New Feature Thoughts

[Eric B. <eri...@pr...>]

Log4cpp is working really well for me. There are a few features I'd like to
see added. If I get a chance, I'll try to add them myself and submit
patches, but these are my thoughts:
- It looks like shutdown() is supposed to be called to clean up memory and
resources for Category Objects? I'd like this to happen automatically
somehow.
- I'd like the PatternLayout be able to support putting the thread name and
thread number in the output. I can add this for Win32, but don't know posix
enough to do it there.
- On Win32, build libraries for different compiler settings and generate
#pragma comment(lib, ...) to get library linking automatically. (See
<boost/regex/v3/regex_library_include.hpp for example.) (Be sure to include
#pragma comment(lib, "wsock32.lib") too.)

Is there any talk of adding log4cpp to boost?

Eric

RE: [Log4cpp-devel] Logging Architecture

[Steve O. <Ste...@Ne...>]

From a performance perspective, having a hierarchy in your log4cpp
categories is a problem, Ex: "root.cat1.cat2". The performance hit is not
significant for GUI code, but for a tight loop where performance matters, it
can be a problem.

Calls it isDebugEnabled(), which have to jump up a category hierarchy, begin
to add up.

SteveO

-----Original Message-----
From: Colin Law [mailto:c....@el...]
Sent: Wednesday, August 13, 2003 3:51 AM
To: log...@li...
Subject: [Log4cpp-devel] Logging Architecture


Hi,

I have been using log4cpp for a while now, but I have never been sure on
the 'best' way to organise log categories within my classes. Should I
define a log category within each class, and get one category per class,
or get a global instance per implementation file, or possibly create a
shared instance to be shared amongst a group of classes?

I guess there are probably different techniques for different scenarios,
but I would be interested to hear how other people are doing this.

~Colin.



-------------------------------------------------------
This SF.Net email sponsored by: Free pre-built ASP.NET sites including
Data Reports, E-commerce, Portals, and Forums are available now.
Download today and enter to win an XBOX or Visual Studio .NET.
http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01
_______________________________________________
Log4cpp-devel mailing list
Log...@li...
https://lists.sourceforge.net/lists/listinfo/log4cpp-devel

[Log4cpp-devel] Logging Architecture

[Colin L. <c....@el...>]

Hi,

I have been using log4cpp for a while now, but I have never been sure on
the 'best' way to organise log categories within my classes. Should I
define a log category within each class, and get one category per class,
or get a global instance per implementation file, or possibly create a
shared instance to be shared amongst a group of classes?

I guess there are probably different techniques for different scenarios,
but I would be interested to hear how other people are doing this.

~Colin.

Re: [Log4cpp-devel] Help - Gobal Initialization

[Aaron I. <ai...@ya...>]

Hi Eric,
I suggest maintaining a global Category pointer rather than a global
Category reference.  The pointer can then be initialized within
DllMain.  As for the "exception or poor memory reference", please post
details so I can offer suggestions.
-Aaron

--- "Eric Brown (Propel)" <eb...@pr...> wrote:
> I'm trying to use log4cpp inside a dll. I don't want to use global
> initializers like:
>     Category& logger = Category::getInstance("bla");
> because it gets invoked before my dllmain. Furthermore, it seems to
> be causing some kind of exception or poor memory reference or
> something that I can't detect in my context. (I'm not sure what the
> stupid Windows loader is doing, but it is doing something annoying.)
> If I call getInstance() in a function, I have to call it in every
> function as I can't figure out how to assign a category to a global
> variable after my dllmain is called because Category::operator= is
> declared as private and not implemented as far as I can tell.
> 
> I'm really stuck here. Otherwise, my tests of log4cpp seem to
> indicate that other than the code size it adds to my dll, it is
> great.
> 
> Thanks,
> Eric
> 


__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com

[Log4cpp-devel] Help - Gobal Initialization

[Eric B. \(Propel\) <eb...@pr...>]

I'm trying to use log4cpp inside a dll. I don't want to use global =
initializers like:
    Category& logger =3D Category::getInstance("bla");
because it gets invoked before my dllmain. Furthermore, it seems to be =
causing some kind of exception or poor memory reference or something =
that I can't detect in my context. (I'm not sure what the stupid Windows =
loader is doing, but it is doing something annoying.) If I call =
getInstance() in a function, I have to call it in every function as I =
can't figure out how to assign a category to a global variable after my =
dllmain is called because Category::operator=3D is declared as private =
and not implemented as far as I can tell.

I'm really stuck here. Otherwise, my tests of log4cpp seem to indicate =
that other than the code size it adds to my dll, it is great.

Thanks,
Eric

Re: [Log4cpp-devel] log4cpp problems

[Robert J. S. <r.s...@ch...>]

Op zondag 13 juli 2003 01:24, schreef David H. Lynch Jr.:
>     Atleast for me logging is a global function for the whole
> application. Therefore I need a global logger/category beyond the scope
> of of a single function.

Define this:

    Category&
    log()
    {
	static Category& l = Category::getInstance("my_category");
	return l;
    }

and then use it like this:

    log().debugStream() << "message";

I have such a log() function defined for every class, most of the time as a 
free function in the anonymous namespace and sometimes as a virtual 
protected/private member function.

>
>     I have been unable to declare a category object outside the confines
> of a function. 

You should declare a reference not an object, see above.

> Either there is not an appropriate constructor or the
> resulting program faults during initialization with a bad memory access.
> I can declare a global pointer to a category object, and assign it a
> value inside main, but some of the methods generate exceptions that do
> not appear to be documented when used in the pointer form cat->method()
> instead of cat.method();
>
> The problem here is more likely to lie with my early c++ skills rather
> than log4cpp but a clue would be appreciated.

It seems so, but until you post some code/error messages we won't know.


>     I have done numerous searches looking for documentation or sample
> applications that use log4cpp and there is a dearth of documentation
> available. I tend to find I learn more easily for examples, but the
> examples I have found are primarily the test samples that come with
> log4cpp, which are nice but either confine a category to the scope of a
> function or pass it as an argument neither of which are practical and
> violate the precept that the debugging is independent of the operation
> of the program being debugged.

I dont know what you mean by this precept (in what way independent?). Maybe 
that you want to strip all the logging code in release mode?

Logging != debugging. If only want to produce debug messages and want to strip 
this completely from your code in release mode then perhaps its beter to look 
for a macro based solution. This is not what log4cpp is about.

[Log4cpp-devel] log4cpp problems

[David H. L. Jr. <dh...@dl...>]

    I am new to log4cpp and relatively new to C++. But the capabilities
of log4cpp look very useful as aides to the software I develop.
    However I have been unable to figure out how to get log4cpp to work
as I need.
 
    Atleast for me logging is a global function for the whole
application. Therefore I need a global logger/category beyond the scope
of of a single function.
 
    I have been unable to declare a category object outside the confines
of a function. Either there is not an appropriate constructor or the
resulting program faults during initialization with a bad memory access.
I can declare a global pointer to a category object, and assign it a
value inside main, but some of the methods generate exceptions that do
not appear to be documented when used in the pointer form cat->method()
instead of cat.method();
 
The problem here is more likely to lie with my early c++ skills rather
than log4cpp but a clue would be appreciated.
 
    I have done numerous searches looking for documentation or sample
applications that use log4cpp and there is a dearth of documentation
available. I tend to find I learn more easily for examples, but the
examples I have found are primarily the test samples that come with
log4cpp, which are nice but either confine a category to the scope of a
function or pass it as an argument neither of which are practical and
violate the precept that the debugging is independent of the operation
of the program being debugged.
 
    

[Log4cpp-devel] VS .NET 2003/NTEventLogAppender Redux

[Amit V G. <am...@pe...>]

Has anyone been able to use NT event logging in log4cpp with VS .NET 2003 ?

There was a posting a month ago about problems getting NTEventLogAppender.dll
to build with VS .NET (VS 7). Has anyone been able to get it to work since ?

If you have been using log4cpp NT event logging I would appreciate hearing of
what was required to get things working...

Amit
-------------------------------------------------------------------------------
| Amit V. Garde      | am...@pe... | +91-20-567-8900 x517           |
| Persistent Systems | Bhageerath,402 Senapati Bapat Road,Pune 411 016, India |
-------------------------------------------------------------------------------

[Log4cpp-devel] Release plans

[Bastiaan B. <Bas...@li...>]

Hi,

Since Brian asked, and I've been silent about my log4cpp plans for too
long anyway here's they are:

1) Make a final release of the 0.2.x branch which fixes gcc 3.x
compilation issues. After that this branch will be closed.

2) Release 0.3.5 in the next few weeks, containing Dirks patches for
MSVC DLL builds and other bugfixes. Focus for further 0.3.x releases
will be stablization, multithreading code cleanup, bugfixes, and
platform support. 

3) Branch off 0.3.x and push the trunk to 0.4.x. 0.4.x will be the place
for new API incompatible stuff e.g.:
 - introduction of the new naming scheme (Logger, Level, etc.) 
 - proper cleanup of log4cpp objects through reference counting.
 - extensibility of configurators for custom appender classes.
 - SocketAppender for log4j interaction.
 - faster, inlineable logging methods.


Ad 1): This has been completed this morning: 0.2.8 is on SourceForge
now. 

Ad 2): I plan to release 0.3.5rc1 somewhere next week, and promote that
to 0.3.5 if no issues show up.

Ad 3): I'll let that coincide with the release of 0.3.5 for convenience.
But if you have plans, ideas, suggestions, etc. for the future of
log4cpp development, feel free to discuss them now. CVS access will be
open for people with good ideas (and code of course).

Cheers,

Bastiaan 

-- 
Bastiaan Bakker <Bas...@li...>
LifeLine Networks bv