Overview: May 2020 Sprint

[mike-thompson-day8]

Isaac and I are doing a three-week-long sprint on re-frame - starting 4th May. Isaac's time is being paid for by ClojuristTogether funding. My time is paid for by Day8. So between us, we have about 6 man-weeks.

At a high level, the strategy is:

• do the low hanging fruit (there are a few smaller, easier issues)
• do a 1.0.0 release (it is about time)
• create a grown-up website and fix up documentation which has fallen a bit into disrepair. This is a bit scary because it could take an infinite amount of time, so how do we limit it?
• attack three of the bigger, meatier issues: BranchScope, Multiple frames on a page, views. All three of these issues are interrelated.
• do a release

This will likely take longer than six man-weeks. And then there will be necessary adjustments to re-frame-10x. To quote a line from the movie Get Shorty "This is only the second day, and we're already two weeks behind." But, one step at a time, let's just see how we go.

I will update this issue with new thoughts as we go. We will probably engage with the community via the clojurians #re-frame channel. If you are interested in the process, please keep an eye out there.

[mike-thompson-day8]

Progress on the early, low-hanging-fruit step:

• thoroughly review all the existing issues and PRs.
• close any issues which need it
• assemble the candidate list of issues we identify as "low hanging fruit"

Progress ...

Closed with explanations:

• Reaction instance gets reused after dispose #553
• dispatch-n calling dispatch-n evaluates events in a non-obvious order #572
• Subscription behaviour using a keyword as the computation fn (correct but confusing) -> doc ? #555
• FAQ Entry suggestion: wire up re-com model atoms with your re-frame handlers #112
• Investigate microtask queueing once it's implemented #481
• Preserve stack frames to show where event came from #164
• dispatch-later doesn't allow dispatch-n #554
• Use of goog.async.nextTick prevents reagent from rendering #480

Picked Fruit, so far:

• Add docs for :<- #544
• Docstrings not available for some re-frame.core functions #456
• Create API documentation for core.cljs #216

Issues arising:

• Missing Docstrings #588

Building a grown-up website #582:

• A review of options and likely paths has commenced.

[mike-thompson-day8]

• some distracting issues with re-frame-template, shadow-c;lj and Lein-Shadow. See npm dependency is culled out when I run lein dev re-frame-template#128 and https://gitlab.com/nikperic/lein-shadow/-/merge_requests/3
• some distractions with re-frame-10x (almost there)
• Implement a way to add global interceptors #570 - Implement a way to add global interceptors
• about to cut a new release which includes all the low hanging fruit

New website progress

• learning new static site generator (good decision so far, but there's always stuff to learn)
• we've spent far too long getting Klipse going (to supply a page of live coding)
• a GithubActions based build system for new website
• a LOT of rewriting, writing and thinking
• integration with API docs

New docs include the following:

• Reusable Components . Addresses issue Composing "complex" components #264
• On Dynamics
• Subscriptions. Previously a neglected area.

Hoping to deploy the new website over the coming weekend.

We're down to about 18 issues and 3 PRs. When we started, I think it was 30 issues and 9 PRs.

[mike-thompson-day8]

The new re-frame website is now up at: http://day8.github.io/re-frame/

The re-frame logo has been tweaked. See it here:
http://day8.github.io/re-frame/re-frame/

And a release candidate for 1.0.0 is also out

Also, we're now down to about 16 issues and 2 PRs.

Now we begin on the meatier issues.

[mike-thompson-day8]

The task of addressing meatier issues is now slightly delayed.

• new FAQ item added which closes Race condition in simple example app #538
• Add re-frame to cljs-oss/canary #478
• in light of negative feedback we have received, we are reworking the new website. Goal is to present code earlier and require less initial reading.

We are down to 13 open issues and 2 PRs, which is a numerical measure of some success.

[mike-thompson-day8]

Many improvements to the docs have been made recently. Too many to mention but some are notable:

• added a new FAQ on laggy input
• added a new FAQ on field focus
• added to existing FAQ on global interceptors
• Additions made to External-Resources
• completely reworked Infographics for Dominoes 1,2 3
• completely reworked Infographics for Interceptors. See also other explanations added to that tutorial.

Also:

• closed Possibility to use without Reagent? #590
• created Improve exception debugging by removing catch and rethrow from re-frame's router #627 and developed a corresponding PR Improve event exception debugging #628
• we continue to wrestle with getting a v1.0.0 out. React native niggles still exist. See v1.0.0-rc2 throws error on React Native Expo #604 v1.0.0-rc2 doesn't work on Webworker #614 Fixing running re-frame on webworker #615.

Numerically, we're at 13 issues and 3 pull requests.