- From: jugglinmike <notifications@github.com>
- Date: Wed, 21 Jun 2017 09:32:37 -0700
- To: w3c/permissions <permissions@noreply.github.com>
- Cc: Subscribed <subscribed@noreply.github.com>
- Message-ID: <w3c/permissions/pull/151/c310134576@github.com>
@raymeskhoury thanks for the feedback! > Did you have in mind practical impacts of this. Does it just "feel bad" or > does it mean that we can't test something that we otherwise would be able to > test? As I understand it, the "proactive API" would be specified to only modify an environment settings object. In other words, it would have no effect on any permission requests that were already awaiting user input. Such an API seems to be just as able to reach all the normative text, so it may be equally valid for the purposes of conformance testing. The API does has some bearing on the way tests are authored, though. It makes the "request permission" state non-recoverable. Application developers would have to maintain tests dedicated solely to verifying that certain user actions triggered prompting, but that went no further in an interaction flow. The degree to which this matters will vary between developers and projects, but I expect those that ascribe the Behavior-Driven Development will find it annoying to implement tests described by, "When I click the 'Join' button, the application displays the text 'Waiting' and nothing else happens." I decided to try the so-called "reactive" API for this initial patch in light of this and in deference to existing patterns in WebDriver (Simon shared the following in that document you've read--I'm including it here just to make the conversation easier to follow for others) > This most closely models the current webdriver commands. The example, the > User Prompt commands follow this style, and (obviously) so do the commands > for interacting with elements. That said, I'm not an implementer. If we can agree on an alternative that feels right for the various "drivers" and the spec editors, I'm happy to re-write any or all of this patch. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/w3c/permissions/pull/151#issuecomment-310134576
Received on Wednesday, 21 June 2017 16:33:33 UTC