Please add topics for discussion (and your status if you’re a contributor).
No status this week for me on FuncUnit. Except that we are now already reaping the benefits of
sendKeys() here at MathWorks!
Work in progress: @Chinmay and I are discussing how to possibly incorporate touch event support (and pointer events too), right now we are still developing our test cases to show the problem statement. It seems there are several aspects to think about (such as whether we add touch event generation right into click(), leave click alone and develop new touch* methods, then later new pointer* methods, or do something else entirely). Hopefully we can share some kind of problem statement and proposal with you guys soon.
FuncUnit DnD does not work in some cases: Our users have found that HTML5 based DnD works where they were failed by the Selenium or FuncUnit DnD for some cases they looked at. I wonder if this group is open to updating FuncUnit DnD to use HTML5 based approach, or if there are oppositions to this idea. I can share a prototype of a tool they developed, for discussion. I will start by filing an issue where we can discuss this in detail.
QUnit upgrade: No status this week I would love to look at upgrading to
QUnit 1.23 (straight up to 2 seems more challenging in more than one way), but I don’t have the time right now due to overcharged workload. That is next on my list of priorities.
Sorry I can’t make it tomorrow due to meetings, but that is my not so short status/next steps.
- Got Steal’s testing passing in IE9-11, Edge, Safari, Chrome, Firefox, and iOS Safari.
- Got Steal set up with saucelabs which will run the above browsers on each PR.
- Merging all of the Steal 1.0 projects into their master branches.
- Releasing 1.0
- 1.0 article
- Filled out the “2016-11 DoneJS Report”: https://github.com/donejs/donejs/issues/753
- Fixed various documentation issues
- Conducted UX testing of the new CanJS site with three outside devs
- Opened an epic for the issues I’m creating based on the UX testing: https://github.com/canjs/canjs/issues/2862
- Worked on Bitcentive payouts component
- Fixed various other bugs on Bitcentive
- Continue working on Bitcentive
- worked with bitcentive team to get mvp bugs fixed and MVP working on staging
- will deploy to prod as soon as Travis node problems are sorted out
- Moving off open source work onto client work
- while working on Bitcentive, continue to improve
- start writing a blog post about mocking socket.io.
- Teamed with Ilya to update steal-socket.io to support methods and attributes that feathers-authentication requires to properly authenticate, disconnect, and connect a socket.
- Updated can-fixture-socket to support the same socket methods and attributes.
- Fixed login and logout flow in Bitcentive.
- Fixed authentication error in Bitcentive from Steal re-throwing errored promises. (Steal still re-throws)
- Updated can-connect-feathers to better handle JWT authentication.
- Refactor can-connect-feathers tests test fixtures in addition to the integration with the Feathers test API server.
- Get can-connect-feathers fully integrated into the canjs repo.
- CanJS 3.0 article released
- Finalizing DoneJS 1.0 plans
Rough meeting notes:
CanJS, StealJS, and DoneJS release schedule
- CanJS 3 announcement on Tuesday
- StealJS 1.0 release to npm sometime early next week
- Maybe announce at the end of next week or first thing the following week
- DoneJS effort: lots of projects already have prerelease versions that support CanJS 3, just need to update them with the latest
- Can the custom generator support wait until after 1.0?
Long-term fix for can-compile: https://github.com/canjs/can-compile/issues/56
- CDN is acceptable in some scenarios, but making it configurable would be helpful
- Can we ship the versions with the package? Not all of them, but say 2.3.28 to cover 2.3–2.3.28?
- Maybe ask someone using it to become a maintainer
Initial thoughts from the UX testing
- Epic for all the issues: https://github.com/canjs/canjs/issues/28622
- Open call for any feedback on those issues
- Please +1 ideas or leave comments if there are issues not being considered
- Briefly discussed Move the “Collections” and “The can Package” sections off the home page
- Briefly discussed Re(move) the API Guide page
- Users shouldn’t need a guide to use the site
- Idea: produce video tutorials to cater to a different learning style
– Chasen will (eventually) look into how we can get content on egghead.io, Lynda, etc.
- Question: what can we do differently to avoid needing to make big changes to our side
– Problem: for web apps you can prototype new features, but the content of the site is vital to gettting feedback
– Solution: conduct user testing as soon as the content is ready, preferably before launch
- Question: where do we get the users?
– Answer: meetups, elsewhere?