There are some reasons for not committing this directly: o Norman recently developed and added the functionality, he tested it and it was all good. o Now, I did some more than minor refactorings including changing the component API and thus the integration of the components. o Currently, I am not able to test the Bayesian management, including my new code, by myself. (time and workplace restrictions)
I think it would be unfair to unintentionally break the working code from Norman for the pure reason of (subjective) beauty. Committing untested code is not acceptable. So I chose to put it in the JIRA until myself, Norman or others have time to pick it up. I only tested the availability of the new BayesianAnalyzerManagement JMX component and that part seems to works fine. Should've added these notes to the patch in the first place...
So feel free to just commit it using the "JAMES-590" identifier in the log so that it will automatically appear in the JIRA issue.
Are you saying, JIRA parses commit logs and links them to issues? Bernd On 9/5/06, Stefano Bagnara <[EMAIL PROTECTED]> wrote:
Bernd, Imho you should directly commit this kind of patches. You are a committer (a PMC one) and we use the Commit-Then-Review approach with our trunk. So feel free to just commit it using the "JAMES-590" identifier in the log so that it will automatically appear in the JIRA issue. This make it easier to review either the diff and the final result without having to apply patches manually. This is not a critic, I just want to let you know that I trust you and I think you produce good quality code, so feel free to commit to trunk (we can always revert it later if anyone has problem with the code) Stefano Bernd Fondermann (JIRA) wrote: > [ http://issues.apache.org/jira/browse/JAMES-590?page=all ] > > Bernd Fondermann updated JAMES-590: > ----------------------------------- > > Attachment: bayesian_jmx.patch > > reduced the number of thrown excpetions to one: BayesianManagementException. (this is best practice. components should not put the burden of handling internal errors on the caller) > > made management functionality available for JMX. > >> Add commands to RemoteManager to corpusfeed JDBCBayesianAnalyzer >> ---------------------------------------------------------------- >> >> Key: JAMES-590 >> URL: http://issues.apache.org/jira/browse/JAMES-590 >> Project: James >> Issue Type: New Feature >> Reporter: Norman Maurer >> Assigned To: Norman Maurer >> Priority: Minor >> Fix For: 3.0 >> >> Attachments: bayesian_jmx.patch >> >> >> We should add commands to RemoteManager to allow an admin to corpus feed JDBCBayesianAnalyzer. This whould allow an new user to train the spamfilter with an corpus of ham or spam without sending each spam or ham with mail --------------------------------------------------------------------- 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]
