Metadata
- Source
- FLUID-2244
- Type
- Bug
- Priority
- Major
- Status
- Closed
- Resolution
- Fixed
- Assignee
- N/A
- Reporter
- Justin Obara
- Created
2009-02-17T10:15:25.000-0500 - Updated
2010-12-14T11:31:17.241-0500 - Versions
-
- 0.8
- 1.0
- 1.1
- 1.1.1
- 1.1.2
- 1.2
- 1.2.1
- Fixed Versions
-
- 1.3
- Component
-
- Pager
Description
Pressing tab, after selecting a page, places focus on the first page link. Focus should be moved to the next page link
Steps to reproduce:
1) Open the renderer version of pager
http://build.fluidproject.org/fluid/sample-code/pager/renderer/pager-render.html#
2) Click one of the page links
3) Using the keyboard, tap the "Tab" key
Notice that focus is placed on the first page link, instead of the next one.
Environments
FF2, FF3, Opera 9.6, Safari 3.2 (Mac OS 10.5)
Safari 3.2 (Mac OS 10.4)
FF2, FF3, IE6, IE7, IE8, Opera 9.6 (Win XP)
FF3, IE7, IE8 (Win Vista)
IE6 (Win 2000)
Attachments
Comments
-
Antranig Basman commented
2009-05-15T13:48:33.000-0400 Fixed at revision 7176 - new "delegate plugin" and global focus manager
-
Justin Obara commented
2009-05-19T09:07:41.000-0400 This issue has been fixed for this page:
http://build.fluidproject.org/infusion/integration-demos/sakai/html/pager-site-setting.html#but not for this one:
http://build.fluidproject.org/fluid/sample-code/pager/renderer/pager-render.html# -
Antranig Basman commented
2009-06-03T13:53:47.000-0400 This kind of problem affects almost all of our keyboard accessibility cases. In general, clicking something with a mouse, and then expecting that keyboard focus has also followed this selection does not work (cf "accessible tabs", etc.)
-
Justin Obara commented
2010-05-07T13:17:26.371-0400 a11y issue
-
Justin Obara commented
2010-11-19T14:41:02.675-0500 Try preventing the default action of the link, which is adding # to the end of the url. If this doesn't work, we may need to reassess for bug parade
Also this issue actually appears to be working in the current trunk, but regressed in the aegis branch
-
Justin Obara commented
2010-11-19T16:08:17.424-0500 "Bug Parade Infusion 1.3"
-
Justin Obara commented
2010-11-30T12:32:05.460-0500 As part of the commit to FLUID-3828 & FLUID-3829 at r10313, the default action of the links were prevented. This didn't however solve this issue. It may have something to do with the id. Previously the pager had been using the same type of page link to represent all of the page links. Now it has switched to using the disabled type of page link for the current page. This is what is causing the issue. It seems thought that even switching the disabled page link to have the same markup as the other page links doesn't help, which is what leads me to suspect the differing id as the culprit. The id's on the standard page links look something like "page-link:link1" whereas the current page link id's look like "page-link:disabled1". If the current page were using the same type of page link as the others, the id would stay the same after navigating to that page, and focus would remain.
-
Justin Obara commented
2010-11-30T12:37:20.115-0500 I've confirmed that it is related to the id.
In fluidRenderer.js
line: 1446
lastId is set to the id of the last focused elementline: 1472 - 1477
The element identified by lastId is found. If it exists, focus is called on it. -
Harris Wong commented
2010-12-01T11:55:03.610-0500 Instead of disabling the current element with the "page-link:disabled", we now apply the currentPage style onto the current element. Also removed all "flc-pager-pageLink-disabled" from the Pager-tests.html. Lastly, the pager tests are also adjusted respectively.
-
Harris Wong commented
2010-12-01T12:52:47.623-0500 Replaced patch.
-
Justin Obara commented
2010-12-01T17:02:20.146-0500 Committed Harris' patch ( FLUID-2244-patch.txt ) which solves the issue of the changing id's by removing the notion of a disabled link. The code had recently been changed to use disabled links for the current page, which resulted in the id of the element changing, and the lack of focus after rerendering.
-
Michelle D'Souza commented
2010-12-14T11:31:10.819-0500 Functionally tested by Mike, code reviewed by me. All systems go.