ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Donald <>
Subject Re: jakarta-avalon-logkit
Date Thu, 10 May 2001 00:02:56 GMT
At 04:34  9/5/01 -0700, Jon Stevens wrote:
>on 5/9/01 3:54 PM, "Peter Donald" <> wrote:
>> You must hate it when the truth gets in the way of your trolling. Have a
>> look at avalons list and you will see that the change was -1'ed and backed
>> out. 
>Peter. It was the fact that the code that broke other code went in in the
>first place.
>I only care about your concern about the concept of deprecation, not the
>fact that you were caught and told to fix things (or that you even provided
>a patch to Velocity to fix things).

I wasn't caught - the patch wasn't done by me it was done by someone else -
I was the one who -1 ed it. I "caught" someone else doing that - but that
doesn't quite fit your world view ... so you feel that it is appropriate to
construct another truth ... hmmmmm

>#1. At least, Conor, myself and Stefan all want a pluggable system. So, your
>arguing a point which is moot. This proposal isn't going to go through. You
>do not have the right to change my (or anyone else's) vote to -0 or anything
>else. I have provided an alternative solution which solves the problem. My
>-1 on a hard dependency on LogKit stands. End of discussion, you are not
>going to change my mind or my vote. Creating a small wrapper and interface
>is not an ugly design nor is it this huge amount of work. I have done it
>twice now and it was trivially easy and you can't convince me (or any other
>right minded person) otherwise.

you have neither the knowledge of ant or any of the logging toolkits to
make a judgement - so far the only person who has indicated they understand
the situation is Stefan.

Stefan and Connor also have not indicated they want a pluggable system,
both of them want to recreate a system inside ant and not use any external
toolkit. So you are wrong in that point aswell.

The only committer who has given any indication that they agree with you is
Christoph Wilhelms. 

And if you recall a while back that you agreed that straight out -1 votes
with no logical backing up was invalid ... except when it violates the
charter (The example you gave was -1 on making tomcat violate the servlet
spec). Unless you want to raise a vote to explicitly make "Using pluggable
log systems" part of ants charter then I am fairly certain my understanding
of apache process is the same as yours ... or at least how yours used to be. 

>#2. Deprecation is not about fixing what people catch you breaking. 

pfft - you know this as applied to me is false and yet you deliberatly go
out of your way to try to paint it on me. Why is that? 

>It is
>about taking an active stance to ensure that your code will work into the
>future. Sending patches to a group that is using your code isn't enough.
>What about all the tons of other people you claim are using LogKit or the
>people you don't even know about?

I moved everyone I knew who was using logkit to log4j. It was never
advertised at apache, always kept in a CVS with an alpha project which
explicitly listed a warning file in it saying "NO backwards compatability
is guarenteed - use at own risk". If that is not enough warning against not
using it then I don't know what is.

>If you come from a Perl or PHP background, JSP is a way to take
>your pain to new levels. --Anonymous


| "Faced with the choice between changing one's mind, |
| and proving that there is no need to do so - almost |
| everyone gets busy on the proof."                   |
|              - John Kenneth Galbraith               |

View raw message