Hi Vincent,

We actually don't have any trouble detecting quirks. The issue is that users of browsers we don't "officially" support don't like the warning dialog we put up when loading a DHTML OpenLaszlo application.

jim

On Jun 21, 2007, at 10:14 AM, Vincent de Phily wrote:

On Thursday 21 June 2007, Jim Grandy wrote:
On Jun 18, 2007, at 2:59 AM, Vincent de Phily wrote:
Konqueror pretty fast too, if only there wasn't that annoying
nag-screen/browser-detect telling the user "run at your own risks"...

If someone would like to propose an alternative mechanism that still
gives reasonable feedback about supported/unsupported status, I'm
sure we can muster resources to get it implemented...

Disclaimer: I dont know what browser quirks you are up against, or how you detect / work around them, but here's the best browserquirk design I know :


* have a global JS array keyed by quirk name, eg 'lacks_array_push_method' * detect the individual quirks as needed. Test for the quirk itself, not the
  browser.
* If you really cant test for the quirk (doesn't happen that often), consider
  working around it in your code. Otherwise, use a browser detect.
* In your code, test look up the specific key in the JS array when needed. * If a needed feature is not available, tell the developper via an API or a console warning. Dont tell the user. The develloper knows best to either
  tell the user, use a fallback, or let the user try anyway.


I'd be happy to know about quirks you have trouble detecting.
--
Vincent de Phily

Reply via email to