Thanks again, Justin. My responses are inline below. I rearranged your email such that your first finding comes last, since it is the most difficult to address :).
On Sun, May 29, 2016 at 11:23:25AM +1000, Justin Mclean wrote: [...] > The core release notes is out of data and talks about March release > with the next release being in April and it being the "second source > release”.The readme is also not up to date, and the supported board > list seem to be missing a few boards? The outdated release notes is definitely our screw up. However, the list of supported platforms in the readme file looks correct to me. The only discrepancy I see is that the Arduino 101 isn't mentioned in the list, but I think that is correct since this target is not fully supported yet. Is there another one that I missed? > The newt release notes are also out of date and refers to the "first > source release”. Darn. We should have caught this one as well. > I would suggest that the artefact names be renamed to include “mynewt” > not “newt” especially the core one for the next release. That sounds reasonable. I think we should just add "mynewt" to all three component names, but I am interested in hearing others' thoughts. > For core is there now any way we can improve on this situation? "This > product bundles additional files from CMSIS-CORE, but these files are > missing licensing information.” Yes, this one is indeed troubling. To provide some background: Recent CMSIS-CORE releases contain the BSD license text, but the older releases do not, so it is possible that the distributor simply forgot to add the license text in the older releases. Back when the CMSIS-CORE files were added to Mynewt core, we added versions which lacked the BSD license text. The CMSIS-CORE files that Mynewt bundles fall into two categories: 1. Revisions exist which are identical to those which were originally included in Mynewt; the only difference being the addition of the BSD license. 2. Functional changes were made by ARM mbed before the BSD license was added; the oldest revisions containing the BSD license text are different from those in Mynewt. The licensing issue was easy to solve for the "category 1" files: we just replaced our old copies of these files with the corresponding BSD-licensed revisions. The files in category 2 are the ones causing issues for us now. Since the versions of these files in Mynewt are not directly derived from revisions containing the BSD license, I don't know what we can reasonably claim. Intuitively, I feel like the Mynewt copies of these files are "more or less" derived from the revisions containing the BSD text, but obviously that is not a legal argument! Do you have suggestions on how best to proceed? (In case I gave the opposite impression, I agree that keeping things in their current state is not a solution). Thanks, Chris