These are chat archives for openseadragon/openseadragon

20th
Jan 2015
Christian Kütbach
@ckuetbach
Jan 20 2015 18:32
Hi there, thanks for the fast fix for #574
Christian Kütbach
@ckuetbach
Jan 20 2015 18:45
I think, there is now still the same issue for IE11 (even if IE10 works now)
Ian Gilman
@iangilman
Jan 20 2015 18:48
^ @msalsbery
Christian Kütbach
@ckuetbach
Jan 20 2015 18:50
InternetExplorer makes a lot of trouble. I had to spend the complete day, to find the issue in a C# app with an embedded IE11 webview no F12 tools, no console.log... was like developing for IE5.5
Ian Gilman
@iangilman
Jan 20 2015 18:54
:scream:
Christian Kütbach
@ckuetbach
Jan 20 2015 18:55
I have a question regarding the browser detection in openseadragon:
The is code, which will set the Browser and Brwoserversion
I am pretty sure the code is correct, but if one embeds a IE (like me) I get "Internet Explorer" instead of "Netscape" as appname. but the useragent does not contain a "MSIE" so the Browserversion is set to 0 ar NaN (not sure, about)
I changed my version of the browserdetection to use the rv11 regex (from the netscape section)
My questions are: Which problems will I get, if i switch to 1.2.1 with an unknown IE version?
Ian Gilman
@iangilman
Jan 20 2015 18:59
Good question! @msalsbery would know better than I.
Christian Kütbach
@ckuetbach
Jan 20 2015 19:00
and: Could i file an issue (with the version number), even if the problem will only exist in embedded webview?
Ian Gilman
@iangilman
Jan 20 2015 19:00
If you've got a detector that works reliably in both scenarios, seems like an improvement, so go for it.
Christian Kütbach
@ckuetbach
Jan 20 2015 19:00
right know, I use a patched 1.1.1 version, but I would be happy to switch to a released version
I think I have a detector (I only change the version number if not set and use the other regex)
Ian Gilman
@iangilman
Jan 20 2015 19:02
Cool. Please file!
...a pull request would be even better :-)
Christian Kütbach
@ckuetbach
Jan 20 2015 19:02
Yes, for sure.
Ian Gilman
@iangilman
Jan 20 2015 19:03
1.2.1 isn't released yet, but it's got some IE fixes that aren't in 1.2.0 (and presumably soon will have #574)
Christian Kütbach
@ckuetbach
Jan 20 2015 19:03
Tomorrow back in office (it 8PM here, and I happy not to see any IE)
Ian Gilman
@iangilman
Jan 20 2015 19:04
Cool :-)
Mark Salsbery
@msalsbery
Jan 20 2015 19:09
@ckuetbach I'll check the IE 11 docs quick to confirm and I'll add the same check for IE 11. I used the IE11 MSDN sample detection code originally. Apparently the embedded browser team doesn't read that ;)
Christian Kütbach
@ckuetbach
Jan 20 2015 19:10
No the embedded browser is rally different in many cases: An IE11 says Netscape (as appname), the embedded says Microsoft Internet Explorer
But, without Internet Explorer, life as webdev would be so boring ;)
Mark Salsbery
@msalsbery
Jan 20 2015 19:14
Indeed I see that now. :) Anyway, pointerEnabled is deprecated since it isn't necessary and thus was removed from the W3C pointer event model spec. So I'm hesitant to use it the same way as IE 10 check before looking into it more thoroughly. Sure, the check would work now on IE 11 and future IE versions, but other browsers that implement the pointer event model are not going to have that flag.
@iangilman I'm going to hold off on that change for IE 11, so the current PRs can be merged whenever (pending approval of course ;))
Christian Kütbach
@ckuetbach
Jan 20 2015 19:17
@msalsbery I think, there would be a check like window.PointerEvent && (window.pointerEnabled || $Browser != IE) ( don't know the browser detection variable)
Ian Gilman
@iangilman
Jan 20 2015 19:18
@msalsbery cool.
Mark Salsbery
@msalsbery
Jan 20 2015 19:18
Thanks @ckuetbach ! Yeah, I'm just hesitant to add browser detection. MS themselves has funky useragent strings to purposely prevent that very thing. Yet we're still forced to do it because of these things.
Really what that means is I need a few minutes to accept the fact that it's the only way :)
Christian Kütbach
@ckuetbach
Jan 20 2015 19:19
@msalsbery And the useragent wil change for IE11 and windows 10
Mark Salsbery
@msalsbery
Jan 20 2015 19:19
Yep, we already detect them
Christian Kütbach
@ckuetbach
Jan 20 2015 19:19
(It will be a like Webkit and not like Gecko)
Mark Salsbery
@msalsbery
Jan 20 2015 19:19
and yes they will be broken for Windows 10
Wow. Thanks for the buzzkill ;)
Christian Kütbach
@ckuetbach
Jan 20 2015 19:20
And to make it even worse: Spartan Browser
Mark Salsbery
@msalsbery
Jan 20 2015 19:20
I'm going to the bar
Ian Gilman
@iangilman
Jan 20 2015 19:21
LOL
Christian Kütbach
@ckuetbach
Jan 20 2015 19:21
:)
Mark Salsbery
@msalsbery
Jan 20 2015 19:22
Anyway @ckuetbach , if you need a solution quickly, take a look at the legacy.mouse.shim.js file in the openseadragon/tests folder. You can make a plugin just like that and tweak the MouseTracker feature detection to your needs without having to wait for OpenSeadragon releases
Christian Kütbach
@ckuetbach
Jan 20 2015 19:25
@msalsbery I use a patched 1.1.1 (because of another issue #508 ) So I patched my version and everything works for X-UA_COMPATIBLE: IE=10 and some other minor changes.
@msalsbery the good thing using an embedded IE to show a fixed location is, you have full control, the bad thing is the IE
Christian Kütbach
@ckuetbach
Jan 20 2015 19:53
@msalsbery I added the following check to the issue:
if ( window.PointerEvent && (window.navigator.pointerEnabled || $.Browser.vendor !== $.BROWSERS.IE ) ) {
Mark Salsbery
@msalsbery
Jan 20 2015 20:06
That should be cool, thank you! I'll add that to the code if you didn't already.