Howdy, I couldn't agree any more... +1 to everything RBair said ;) As an aside: one way to encourage submissions and ensure they're not dropped is to tell people to enter them as Enhancements in Bugzilla. We've been doing that with some of the jakarta-commons projects I'm contributing to, and it works well.
Yoav Shapira Millennium ChemInformatics >-----Original Message----- >From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] >Sent: Thursday, January 30, 2003 6:16 AM >To: Log4J Developers List >Subject: RE: logging.apache.org > > >I figure now might be a good time to throw in my two cents (in review, it >looks more like 25 cents ;-). I'm one of those developers that's out here >on the mailing list silently reading all the posts etc. I have to agree >with Mark on several of the points he made here. > >Within the past couple of months I submitted a potential change to the >DOMConfigurator. I just wanted to be able to run an XML string into the >DOMConfigurator rather than always having to have a physical XML file. I >haven't heard anything about it since I submitted it. Why haven't I heard >anything? I don't know. It doesn't bother me that much, but... > >Its important to consider that programmers are proud people. We tend to >think rather highly of our own intellect, even when we are fools. In an >open source environment, when a developer who thinks highly of his work is >snubbed (or thinks he has been), he is quite likely to take his talents >elsewhere. The process needs to be such that each contributor feels like a >part of "the team". > >Personally, I feel that logging isn't exactly a niche product. I feel that >logging is an integral part of writing solid enterprise level code. It is >absolutely imperitive that we have logs of what our applications are doing >in an enterprise setting. Even in a desktop environment I think logs are >extremely important. What desktop app have you ever written that didn't >have some homegrown custom logging setup? Nearly every program I've ever >written has written to the console or a flat file for lack of a better >place. Logging is as integral to development as unit testing. EVERY >programmer has ALWAYS written debug information to the screen during debug >or development. Whereas JBoss only matters to j2ee and increasingly to >general middleware apps, logging matters to everybody. > >So what can be done for the "Mike McAngus' of the world"? Too many >committers can make a mess of things, it is true. But, can there be a >difference between committers on an architectural level and committers on >an "internal" code level (ie: fixing comments in the code, optimizing a >method, etc). > >Another project I've been associated with had committers (2 really active >committers, but 4-5 with commit priviledges), and then had team members, >and then the community at large. The community basically just offered up >ideas, but no real code changes. The team members and committers would >provide the real code changes, and the committers would, well, commit them >(or not). This organization provides a bigger surface area for interacting >with the community because the community can interact with team members (of >which there could be 20 or more) rather than just with committers (2). > >There are plenty of us out here who'd be more than happy to offer our >services as team members. This would provide the ability to get a lot more >done much more quickly. > >There are SO MANY THINGS that can still be done! While the overall >architecture is set and is really quite good, there are a zillion tools and >things to be improved. What about a GUI for specifying the log file >settings? What about improving the "chainsaw" or other gui's for reading >log files? What about more and better appenders? What about "plugging" >log4j into 1.4 logging? What about doing more for saving and loading >settings and files in databases? What about ...? There are hundreds of >ideas that people have, but don't know how to submit the ideas, or don't >have the time to write the code themselves and so don't submit. > >If log4j is going to survive, indeed thrive, then the contributions MUST >increase. Log4j will die against 1.4 logging if it does not continue to >make strides. A healthy contributor base is essential. I am also in doubt >that Sun has many (if any) engineers actively working on their logging API, >but that isn't going to stop the vast majority of java programmers in real >world situations from adopting it. Sun, IBM, and others have made very >substantial investments in Java and you can bet that they are going to >support their package. > >Richard > >PS> +1 for logging.apache.org (in case anyone cares ;-) > > >--------------------------------------------------------------------- >To unsubscribe, e-mail: [EMAIL PROTECTED] >For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
