In the past we used to make changes to release and merge those into dev after
releasing. This has the benefit of protecting the release from untested changes
in dev and theoretically saving some time as some of the previous release steps
can be re-used.
Since there hasn't been a lot of activity on dev since RC6, there's not
significant risk in simply creating another RC from dev. It would be easier for
me to follow the usual steps rather than figure out which steps are needed. So
I say commit your changes to dev, and I'll cut an RC7 from that.
Thanks.
On 2021/08/22 02:23:54, Greg Dove <greg.d...@gmail.com> wrote:
> I have made changes locally which I believe address the things I was
> concerned about. Can someone please advise : should these changes be pushed
> to the 0.9.8 release branch or to develop? I think it is supposed to be the
> release branch, and they will later be merged back to develop, but I just
> want to verify that before I do anything.
> I should be able to push these tonight once I can be sure which branch I
> should commit them to.
>
> The changes do the following:
>
> -For 3 zero-byte font (.ser files) and 1 binary vector image file
> (.afdesign in frameworks\themes\JewelTheme\src\main\resources\afdesign) -
> exclude these from RAT reporting in the ApproveRoyale script - remove
> unnecessary noise
> -Include compiler's RELEASE_NOTES.md in release staging
> -update top level RELEASE_NOTES to include a 0.9.8 section (summary of asjs
> and compiler changes)
> -update royale-typedefs RELEASE_NOTES to include a 0.9.8 section ('no major
> changes')
> -minor changes to asjs RELEASE_NOTES and compiler RELEASE_NOTES (one
> addition to compiler, mostly fixing spelling errors/typos for the remainder)
>
> Thanks,
> Greg
>
>
> On Thu, Aug 19, 2021 at 12:21 AM Piotr Zarzycki <piotrzarzyck...@gmail.com>
> wrote:
>
> > Hi Josh,
> >
> > Thanks for explanation. My expectation was just that if I run approval
> > script with switch for JS only tests - I would get JS only binaries after
> > final build. - This is not a blocker but everyone need to be aware that
> > build producing JS-SWF version.
> >
> > wt., 17 sie 2021 o 18:46 Josh Tynjala <joshtynj...@bowlerhat.dev>
> > napisał(a):
> >
> > > In your screenshot, the left side looks like a JS-only distribution
> > (except
> > > for the extra player and air directories that are there, for some reason)
> > > because frameworks/libs contains KeepsFBFromHanging.swc, while the right
> > > side looks like a JS+SWF distribution with all of the framework SWCs.
> > >
> > > It's worth mentioning that a freshly downloaded JS+SWF binary
> > distribution
> > > is not currently supposed to contain playerglobal.swc, and you are
> > expected
> > > to add it manually (or another tool like Moonshine would do that). It's
> > > quite possible that the ApproveRoyale script doesn't do that for you, and
> > > you still need to do it manually.
> > >
> > > Technically, we are able to build a playerglobal.swc/airglobal.swc from
> > the
> > > asdoc XML using the playerglobalc tool I created, but that is not yet
> > > included in our distribution.
> > >
> > > --
> > > Josh Tynjala
> > > Bowler Hat LLC <https://bowlerhat.dev>
> > >
> > >
> > > On Tue, Aug 17, 2021 at 7:59 AM Piotr Zarzycki <
> > piotrzarzyck...@gmail.com>
> > > wrote:
> > >
> > > > I went trough Approval script and everything goes fine. I have even
> > > tested
> > > > Maven artifacts to build our Apache Royale application - this one also
> > > went
> > > > fine. I have faced some wall when I wanted to try binary distribution
> > > which
> > > > was produced after build in Moonshine. When I try build Royale
> > > application
> > > > I'm getting following error:
> > > >
> > > > "This SDK does not contains playerglobal.swc in
> > > > frameworks/libs/player/11.7/playerglobal.swc."
> > > >
> > > > Folder frameworks/libs - contains bunch of swcs "Collections.swc",
> > > > "Basic.swc" etc.
> > > >
> > > > When I download nightly produced by Maven [1] I see in folder
> > > > frameworks/libs/:
> > > >
> > > > frameworks/libs/air
> > > > frameworks/libs/player/20.0/playerglobal.swc
> > > >
> > > > Screenshot [2]
> > > >
> > > > Why they are so big differences ? I used following command to run
> > > approval
> > > > script: ant -e -f ApproveRoyale.xml -Drelease.version=0.9.8 -Drc=6
> > > >
> > > > Any thoughts on this ?
> > > >
> > > > [1] https://nightlies.apache.org/Royale/Royale-asjs/
> > > > [2] https://paste.pics/b034e5b11b9342184899e8f29b9887d8
> > > >
> > > > Thanks,
> > > > Piotr
> > > >
> > > > pon., 16 sie 2021 o 23:47 Andrew Wetmore <cottag...@gmail.com>
> > > napisał(a):
> > > >
> > > > > This is great news
> > > > >
> > > > > On Mon., Aug. 16, 2021, 3:22 p.m. Josh Tynjala, <
> > > > joshtynj...@bowlerhat.dev
> > > > > >
> > > > > wrote:
> > > > >
> > > > > > Thanks again for the hard work, Yishay!
> > > > > >
> > > > > > I ran the approval script successfully, I confirmed that the
> > > > royale-asjs
> > > > > > works correctly for powering code intelligence in vscode-as3mxml,
> > > and I
> > > > > can
> > > > > > compile and run asconfigc for Node.js. Looks good to me.
> > > > > >
> > > > > > --
> > > > > > Josh Tynjala
> > > > > > Bowler Hat LLC <https://bowlerhat.dev>
> > > > > >
> > > > > >
> > > > > > On Mon, Aug 16, 2021 at 1:15 AM <apacheroyal...@gmail.com> wrote:
> > > > > >
> > > > > > > This is the discussion thread.
> > > > > > >
> > > > > > > Thanks,
> > > > > > > Yishay Weiss
> > > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > >
> > > > Piotr Zarzycki
> > > >
> > >
> >
> >
> > --
> >
> > Piotr Zarzycki
> >
>
