On 26 July 2016 at 20:15, David Costanzo <[email protected]> wrote:
> I find the Xerces release process to be confusing, but I think I've pieced > together some things. Please let me know where I'm incorrect: > > 1) Michael Glavassevich is the project lead and dictates how and when > Xerces is released. > I agree. Michael as a Xerces PMC chair and as one of the primary contributors of XercesJ2, has taken this responsibility always. I believe, he should take this responsibility in future as well. > > 2) Xerces has two separate code bases, "trunk" and "xml-schema-1.1-beta". > These are released as separate JARs and have a different set of bugs, > features, and fixes. > Yes this is correct. As far as I know, the codebase on "trunk" only has the code for XercesJ's XML Schema 1.0 processor, and is required to be compatible with JDK 1.3. The codebase on the branch "xml-schema-1.1-beta" which the latest version on the site, is very current to the XML Schema 1.1 processor. It is fully compliant to the XML Schema 1.1 spec I think. > > 3) Whenever Xerces is released, both the "trunk" and "xml-schema-1.1-beta" > are released with the same version number. > Yes, this is correct. This has been true upto the 2.11.0 release. > > 4) All of your (Muluk's) fixes were made in "xml-schema-1.1-beta" and so > would be included in any Xerces release, but would go into the > "xml-schema-1.1-beta" JAR, not the mainline Xerces JAR. This is correct, to an extent to the ways in which Xerces releases have been occurring till now. > To get your fixes into the mainline release, the xml-schema-1.1-beta > branch would need to be merged into trunk, but there are no near-term plans > to do this. > I'm not sure, whether there aren't any near term plans to do such a thing. Just to answer your question... > Is this correct? > Yes that's correct, as I've answered your question. > > Muluk Gandhi wrote: > (I guess, there's some problem with your mailer. It seems to receive Muluk instead of my name Mukul. But its ok.) > When I tagged the issues in JIRA, I had carefully checked where each fix > was committed so that the auto-generated release notes[1] wouldn't include > fixes that aren't in the mainline version, which is what people will > download when they download 2.12.0. I think, I should apologize of my 2.12.0 tagging actions to my fixes, which seem to have disturbed your view of the auto-generated release notes[1] (I've also seen the details on this link while writing this mail, and can understand how it appears and what are its contents). I did it out of my understanding. I can revert that action (or you may also do that). But doing that, would send so many mails to this list members. Please advise. > That said, since I'm not a Xerces developer and I have no problem with you > following whatever process you like. > I'm sorry, I'm not the only stakeholder in this work. Other Xerces committers and PMC members might have to intervene in this. At this point, I wish not to take any decision on this and I think I must not. > Are you suggesting that the schema 1.1 branch gets merged back into the > mainline code? I'm not suggesting that. I would like to support and adhere to the process as already mentioned on this list about this many times before. I think, someone else might also be able to say something on this. > > [1] > https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=10520&version=12336542 > > <[email protected]> > > -- > Regards, > Mukul Gandhi > <[email protected]> <[email protected]>
