I think it is very unfair to qualify our reaction on this matter as condescending or trivializing [the requested change]. If my memory serves me correctly, this is the first flame war on this mailing list, ever. I am actually not sure what to make of it.
Hmm. I didn't realize this was a flame war; the ones I'm used to have more shouting and profanity...
But really, Ceki, I'm neither for nor opposed to this change. However, both sides' reactions seem silly to me; I won't get into that.
What I will get into is the way you're expecting people who want answers to wait for your ApacheCon presentation when many of the people involved in this discussion will not be able to attend. I can't speak for Endre or anyone else, but to me one of the most frustrating things about this "flamewar" has been your saying "We'll have something will make it simple and efficient to fly simply by flapping your hands very soon, just give us time". That's a vague claim, and do you blame anyone for being a bit doubtful that this new feature is really going to solve this problem? (A problem which, I might add, isn't so complicated, yet you want them to wait for a whole new, untried feature like domains before you give him any solution?)
Here's what I'd like to see:
1. Some code that demonstrates the domains approach. A single class with logging done in this fashion so we can see what it looks like would be nice. I'm sure Endre would like to see something a little more concrete than a vague description of what you're implementing.
2. At least a discussion on performance ramifications. It doesn't have to be a in-depth analysis or benchmarks or anything, just tell us what you think /might/ make it slower and what /might/ make it faster.
3. Configuration details, if those will be different.
4. Backwards compatibility information. I haven't seen much of a request for forty different TRACE levels (TRACE00 - TRACE39), but I have seen a bunch of bitching for a *single* TRACE level. I don't understand why the opinions of one group of people -- unaffected by the change -- should override those of the other group of people -- who would be totally affected by the change. (I would probably have just created a patched distribution by now, but that's irksome as well; you have to keep tracking the project and merging in changes if you want them, since you can't just download the latest release.)
5. Also, if you left TRACE out in the domains version, which sounds like the current plan, would there be a way to use trace-level debugging with the ease that we already have for the other pre-defined levels? I.e., somehow I get the feeling that domains aren't going to be quite as simple as a log4j.debug() is or as a log4j.trace() statement could be.
I'm not for or against either of these things being discussed. It just sounds like both of you are on the major defensive and major offensive at the same time. Why don't we instead try to work out some specifics, instead of talking about vague things like "you're a bad developer because you don't write the code your userbase wants" and "you're a bad user because I don't like the feature you want".
-- Toby Butzon <[EMAIL PROTECTED]> ILC R&D Co-op / Georgia Tech CS / KKPsi Iota Sp03 678.904.2413 work / 678.362.6483 mobile Fifteen Piedmont Center, Suite 700 / Atlanta, GA 30305
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
