I'm very much in favor of simply making it a requirement that if you want to
compile log4j (as oppsed to simply using it) then you have to have ant
installed. It's no different than the requirement that a Java compiler is
installed or that cvs is installed to get things from the CVS repository,
and it GREATLY simplifies many things (totally removes the need to provide
your own shell or batch scripts, doesn't require redownloading something
that the person probably already has installed, etc.). Then, again, you can
have support libraries (those that can be redistributed) "gettable" from an
ant task. Makes things much simpler and faster for both the user and the
maintainer.
-Jim Moore
-----Original Message-----
From: Ceki Gülcü [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, March 14, 2001 6:01 AM
To: LOG4J Users Mailing List; LOG4J Users Mailing List
Subject: Re: Build log4j is a PITA
At 10:00 14.03.2001 +0100, Anders Kristensen wrote:
>Robert Burrell Donkin wrote:
>>
>> the main advantage of this is that the build doesn't depend on the user's
classpath environmental variable. this makes the build
>> much more portable.
>
>But then log4j has to ship with all libraries it depends on - which is
>quite a few. And also, there is no guarantee that the versions of these
>libraries match what log4j users use themselves, i.e. log4j compilation
>could work and then using resulting jar files might fail because they
>depend on newer versions of libs.
Absolutely. That's the down side of adding libraries to the distribution &
CVS. The plus side is that the user is freed from fetching these libraries
herself.
To mitigate these concerns, we could perhaps start by including only ANT and
an XML parser in the distrib+CVS. That way the time between a retrieving the
log4j and building could be just a few seconds instead of a few minutes.
Comments? Ceki
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
