DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://nagoya.apache.org/bugzilla/show_bug.cgi?id=15729>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=15729 StarTeam rootLocalFolder should be java.io.File ------- Additional Comments From [EMAIL PROTECTED] 2003-09-01 18:46 ------- Now, that at last, I have some free time, I am taking a fresh look at this bug. If I may summarize the discussion to this point: Lee Marlow (supported by Jose Alberto Fernandez) wants the rootlocalfolder attribute to be set relative to basedir when not specified as an absolute path. They point out that this gives more conceptual unity with the rest of ant. Stefan Bodewig worries that this will break backward compatibility with users who expect it to be relative to current working directory. I, Steve Cohen, point out that while backward compatibility will break, Stefan incorrectly understands part of what current functionality is. While it is true that if rootlocalfolder is specified as a relative path, that will currently be interpreted relative to CWD, there is a further twist that if NO rootlocalfolder is specified, rlf will be the default local directory stored in starteam. This is primarily useful in the StarTeam GUI itself, and losing this functionality in ant would be no great loss, IMHO. (It could be remedied by adding yet another attribute to the task, possibly). But Stefan is correct, there would still be a significant backward compatibility issue when relative paths are specified. I wasn't thinking about THAT issue at the time of my earlier comments. However, I am inclined to press forward despite this. Current functionality is a mess: 1) if an absolute path is specified, use that. 2) if a relative path is specified, it is relative to CWD 3) if no rootlocalfolder is specified, the starteam default is used. The problem is that 2 is undocumented (no documentation is provided on what happens when a relative rootlocalfolder is specified). Additionally, its lack of documentation makes it inconsistent with 3. User is arguably entitled to assume that not specifying the attribute would be the same as specifying a relative ".". In my use cases, it would make no difference. I always specify rootlocalfolder as an absolute path (perhaps unconsciously aware of the relative path ambiguity here). So what I propose to do is that whenever a relative path is specified, print a WARNING message indicating that the meaning of this attribute has changed from what it meant in 1.5. Something along the lines of WARNING - a relative rootlocalfolder is now interpreted relative to basedir and not to the current working directory as it was in Ant versions prior to 1.6. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
