Hi, I wonder if I could draw your attention to this vote ratification please? Nobody has raised any show-stopping objections to any of the content. On the other hand, nobody has voted yet. I have been reading all the helpful suggestions made in this thread and I have been reporting back on my activities in the light of those suggestions in my earlier posts to this thread.
In addition to my other reports in this thread I have been following up the MANIFEST suggestion for inclusion of the Application-Vendor-Id in the maven users group. So that means that for all the issues raised I believe I have either addressed and rectified them, or explained the current state, or made active steps to pursue improving that state where I understand the points raised, but don't currently understand how to rectify those points. Can you let me know if ratification of this vote is dependent on any further activity on my part please? Best Regards, Kelvin. On 01/11/06, kelvin goodson <[EMAIL PROTECTED]> wrote:
Robert, thanks for your comments. I believe the tag issue is now resolved, by earlier notes on this thread. (http://mail-archives.apache.org/mod_mbox/incubator-general/200610.mbox/[EMAIL PROTECTED] ) The CCLA issue has also been resolved by jeremy boynes (http://www.mail-archive.com/tuscany-commits@ws.apache.org/msg00018.html ) The MANIFEST requirement is an unknown quantity to me, but I will investigate. For the most part the instances of files that you reference which do not have license headers are that way intentionally since either they are generated or because they are present for verification of test execution, and must be equivalent to the data generated in the execution of the tests. This is true of all of the files in the test/resources hierarchy that don't have headers. There are some files that were in error in the samples hierarchy which I have fixed up in the trunk as you asked.I missed these when I ran rat against the source because we had a different source code structure at that time. The files in the model directory are generated and until now I had never looked inside them. If anything I I had assumed that thery were of a format that wouldn't take a comment. I see now that the .ecore file and the .genmodel files are generated XML. so i have added license headers to these two, but I can not modify the .mdl file without risking breaking it. I will investigate whether this file format has a comment syntax i can use to apply a header to the file. I looked at rearranging my source and binary distros in the way you suggested. The binary distro is the way it is (i.e. it unpacks into the working directory) because of maven assembly plugin defaults. I have tried to figure out how to reconfigure maven to put a base directory into the archive, but it wasn't obvious from the assembly plugin documentation and my attempts to play with the configuration failed. The issue I have over having a common base directory for the spec, impl and sample source distributions is that as I understand it the root directory of the distributions each need to contain a BUILDING.txt file. Having a common name for the root directory would cause one extraction to overwrite the contents of the first. I could move BUILDING.txt down to a 2nd level of directory, but that might not make it so obvious what to do with the archives. Any pointers as to how best to do this for the future would be most welcome, thanks. So the one remaining thing is the MANIFEST file issue. In the interest of expediting this release, would you be happy if I opened a JIRA for that issue and committed to resolve it before a further release? Best Regards, Kelvin. On 28/10/06, robert burrell donkin < [EMAIL PROTECTED]> wrote: > > On 10/25/06, kelvin goodson <[EMAIL PROTECTED]> wrote: > > The Tuscany PPMC has voted to release the SDO for Java API > implementation as > > part of the M2 release. > > In accordance with Incubator release procedures we are asking the > Incubator > > PMC to > > approve this release. > > reviewing http://people.apache.org/~kelvingoodson/sdo_java/RC5a/ > <http://people.apache.org/%7Ekelvingoodson/sdo_java/RC5a/> > (since the email doesn't include the explicit reference) > > major > ------- > > no major issues > > there are a few minor questions that need clearing up in 'important' > below. i'd be happy to see these issues resolved in the source without > recutting the release provided that the provinance of these files is > ok. (running RAT will give exact filenames.) > > important > ----------- > i can't find a tag in subversion. please take a tag next time (or > explain your tag naming system if i've missed it). > > the files in http://svn.apache.org/repos/asf/incubator/tuscany/java/sdo/sample/src/main/resources/ > > appear to be missing license headers. please conform that this is > either an oversight or that they are generated. > > the status file is worrying: there are two CCLAs pending. please > confirm that this is either an oversight or that these CCLAs are not > pertinent to this material. > > MANIFESTs are missing Implementation-Vendor-Id (yes, i know it's a > PITA and the various jarspecifications are a mess). best to add it if > you can do so without too much pain. > > files in http://svn.apache.org/repos/asf/incubator/tuscany/java/sdo/impl/model/ > > are missing license headers. please confirm that this was an oversight > or provide a reason why they don't need them. > > a few files in http://svn.apache.org/repos/asf/incubator/tuscany/java/sdo/impl/src/test/resources/ > > are missing license headers. please confirm that this was an oversight > or provide a reason why they don't need them. > > stylistic > -------- > > the download directories are a bit of a hotch-potch. the binary > unpacks to the current directory, sample unpacks to samples, sdo impl > unpacks to sdo, sdo api to sdo-api. it's best to have a naming plan > and stick to it. releases which unpack to the current directory > irritate me (and many other users). i prefer the unpack to directories > named after the release (this makes it easier to manage multiple > releases of the same product). > > - robert >
