- From: Andreas Tolfsen <ato@mozilla.com>
- Date: Tue, 18 Oct 2016 14:07:56 +0100
- To: public-browser-tools-testing <public-browser-tools-testing@w3.org>
- Cc: Jim Evans <james.h.evans.jr@gmail.com>
Jim Evans <james.h.evans.jr@gmail.com> writes: > > On Oct 18, 2016, at 08:09, Andreas Tolfsen <ato@mozilla.com> wrote: > > > > Selenium also does implicit wait polls when clicking, sending keys > > to, and getting the text of elements. > > I hate to contradict you, Andreas, but at least some parts of > Selenium only apply implicit waits to finding elements. This is > especially true of the IE WebDriver implementation. > > I can't speak for other Selenium-compatible implementations like the > legacy Firefox driver or the Chrome implementation. But a blanket > statement that Selenium uses implicit waits on anything other than > findElement(s) is inaccurate. OK. I implemented implicit waits for element interaction commands in OperaDriver, and from Clay’s Edge bug report it seems to exist also in chromedriver. I don’t know what FirefoxDriver does, but the conclusion I’m drawing from this is that implicit waits are _not consistently_ used in Selenium, apart from in the element retrieval commands which is what the spec is currently mandating. Does this mean we should treat the EdgeDriver bug as things working as expected? I personally think it’s a reasonable middle ground to implicit waits to do them only for element retrieval if that is what users are expecting.
Received on Tuesday, 18 October 2016 13:08:27 UTC