OK, good to know it works for other people. I've been trying to adjust the build scripts to create this package automatically. The current problem I'm wrestling with is in the Flash Builder launch configs. The launch configs are used by Flash Builder to update a new Flex project with a valid FlexJS application template. Flash Builder currently can't create a new FlexJS project with the right template. We also use them as a convenience for launching Cordova apps.
Right now, the Installer (actually the Ant script) modifies the launch configs to replace a token in the scripts with the installation folder. Launch configs are copied from the installation folder to the workspace folder so when you launch them, they have lost where they have been copied from so it can't find other files in the SDK, which is why the installer replaces a token to point to the SDK folder. Then, related, as we were preparing our quarterly board report, some PMC members asked how many FlexJS downloads there have been. Consider that if we distribute this non-Adobe packaging as a zip file on Apache mirrors and it becomes popular, it will totally skew our download statistics because we can't get good numbers of downloads from the mirrors and even then, it can be copied internally. So, launch scripts that work out of the box and good download analytics are two things we lose by going with this new packaging. Many folks don't use Flash Builder so requiring Flash Builder customers to run some script to update the launch configs may be a reasonable solution. But it would be a bummer to lose track of successful installs. So, we could go back to using the Installer to install this new package. The install should go fast and won't ask for accepting Adobe licenses. Or we could try to create something that the Installer runs after you unpack the zip. Maybe leave a file in the zip that the compiler knows about and after that file gets 30 days old it asks you to run the Installer to 'register' your SDK and that will fix up the launch configs and hit our analytics. We could have the compiler try to hit the analytics, but I don't want to go down that road right now. Thoughts? -Alex On 9/10/17, 8:10 AM, "PKumar" <prashaku...@gmail.com> wrote: >after removing that line, everything is fine. Thanks > > > >-- >Sent from: >https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fapache-fle >x-development.2333347.n4.nabble.com%2F&data=02%7C01%7C%7Cc9f4a45d0efb475da >70208d4f85e1630%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C6364065303683 >06769&sdata=f%2Bl76Ga%2F%2BjXuaGZlzHiPnoZYP8po1epdVszWY8IJyjg%3D&reserved= >0
