Defect 1788, solves many IE7 (actually IE < 9) compatibility problems #2
Conversation
…IE7, which in turn made the dataset Edit page impossible to work with.
…d of horizontally because of missing support for inline-block in IE <= 7, fixed with a css workaround.
…e.html for IE < 9, to properly show HTML5 section elements which corrupted the layout on those browsers. It must be included in <head> tag because it is said to be required to come before the <body> tag: http://code.google.com/p/html5shiv/
…footer element in all browsers, and fix the footer appearance in IE <= 8. This not only fixes IE7 behavior (the first half of the footer gray, the second half white), but also shows footer containers stretching properly to content in Firefox with Firebug element highlighting feature. Side effect: there may be less white space below the three boxes in a empty homepage.
|
@zydio another big thank-you for what look like a great set of fixes. We'll review this asap and merge. Re html5shiv I'm somewhat surprised we weren't already using that and that should definitely go in (I think there must have been an oversight when we merged the move to the blueprint theming!). Re issues commenting on trac I think I have now fixed that (authenticated users had not been granted relevant permissions). |
[#1788,master][s]: solves the vast majority of layout and misc issues encountered in IE < 9 (and hence IE7 as well). The first two commits, the "light" ones: - some Javascript errors which prevented the dataset page to work correctly - minornavigation links layout. The third commit also tries to deal with the major layout issue: the missing support for HTML5 elements which breaks the header and footer layout (as shown in the screenshot in the ticket #1788 ). It does so by conditionally including html5shiv for IE < 9, which currently seems to be the best way (besides removing all HTML5 elements from ckan). The fourth commit sorts the last annoying visual problem: the footer in IE <= 8 shows half grey and half white, with lists overflowing in the white half. Reworked a bit the html layout and some css properties, and the layout now seems to work better (even analyzing the page with Firebug).
|
Thank you for the prompt feedback @rgrp . There seems to be a problem anyway: I think someone did a Reset (to a commit six days ago) to the master branch after your merge, and the fixes are nowhere to be found now. |
|
@zydio I just noticed this too. What's really weird is I also can't see the commits in master in your repo either (the commits exist e.g. zydio@c102a46. I'm a bit concerned about the reported filesystem issues on https://status.github.com/. We're going to investigate and get back to you ... |
|
Thank you for the update @rgrp . |
@rgrp I guess I now understand why user posting was disabled on ckan trac... |
|
@zydio for spam reasons I've disabled general editing but given you edit / create / etc permissions. |
|
@zydio that's really awesome again! Much appreciated as I was about to go mess around with the DB. In the mean time I've given you permission to do stuff as you have discovered! |
Did somebody ask for translations?
[1602] Replace v2 of OGL with v3
Pull of ckan master in this branch
few minor fixes
…ile-fixes to dev-env * commit '8b198c84173863bfb7dc0e9f907d2c3b2467cf95': Fix some problems I encountered when rebuilding the CKAN docker container. This updates PIP from 1.5.6 to 9.0.1, and also updates setuptools to 23.x to be compatible with the newer pip. Also delays the symlink to `bin/paster` until _after_ the paster binary has actually been installed.
…olumn-types Qol 6127 clean up formatting
I have faced, and been asked to fix some IE issues in CKAN... and those I'm requesting to pull should solve the vast majority of layout and misc issues encountered in IE < 9 (including IE7, obviously).
The first two commits, the "light" ones:
The third commit also tries to deal with the major layout issue: the missing support for HTML5 elements which breaks the header and footer layout (as shown in the screenshot in the ticket #1788 ). It does so by conditionally including html5shiv for IE < 9, which currently seems to be the best way (besides removing all HTML5 elements from ckan).
The fourth commit sorts the last annoying visual problem: the footer in IE <= 8 shows half grey and half white, with lists overflowing in the white half.
I reworked a bit the html layout and some css properties, and the layout now seems to work better (even analyzing the page with Firebug).
I've spent some more words on the last two commits as the first one involves adding a new script for IE browsers (which, as noted in the commit must be in the tag), and the second make a moderate change in the layout (which is fine in my view, and doesn't seem to affect badly any page in any browser).
PS: I've marked everything as related to the Track defect #1788 which I've found searching for the problems I was solving in Track. I'd have commented that defect, but I don't seem to be able to, even if registered.