1. Things that are wrong with the new Oxford web site http://www.ox.ac.uk/

• Breaking OSM licence: http://www.ox.ac.uk/applications/dynamic/map.rm?postcode=OX1+3JP&location=St%20John%27s%20College&id=483 - no attribution

  • (this has now been corrected)

• Input possible through URL query string; used unchecked in title, e.g. http://www.ox.ac.uk/applications/dynamic/map.rm?postcode=OX1+3JP&location=unwanted%20data&id=483

  • This isn't checked at all, allowing cross-site-scripting injection of JS, etc. Eg http://www.ox.ac.uk/applications/dynamic/map.rm?postcode=OX1+3JP&location=%3Cscript%3Ealert%28%22Hello%22%29%3B%3C%2Fscript%3E&id=483 (this has now been corrected)

• Can't get to full form for contact search except by doing a simple search for "people"
• Department/college A-Z: have to guess the letter of the college you want rather than being able to browse them all
• 'This is Oxford' on the home page - what on earth is this for? What sort of information am I expecting to see?

• pages aren't valid xHTML, '&' needs to be changed to '& amp;' amongst other URL form GET problems

• About page claims "The University uses Google Search Appliance, accessible through the search box in the header of all pages. Internal users may also use ht://dig to search Oxford's site including intranet pages, a link to which is also provided on search results pages." When no link to ht://dig is on the search results pages.
• CSS is invalid, references non-standard opacity property, and non-existent vertical-text property
• Visually hard to navigate - uniform mass of small pale-blue text on dark blue background.
• no direct links to key resources such as Webmail, Weblearn, OLIS, Oxlip - navigating to these resources is now too hard. (Weblearn and Webmail have now appeared as Quicklinks)
• Some "Quicklinks" (e.g. library) merely duplicate items on main page - implying that they are hard to find on the main page.
• Try pressing [Go] without selecting a Quicklink!
  • a branded list of Quicklinks would be a more obvious response than an unbranded blank page.

• having  alt=""  might satisfy an automated accessibility check but isn't very helpful for disabled users looking at images.

  • especially when the image is the only content of an href!

• The 'Skip to content' link http://www.ox.ac.uk/#primary-content goes nowhere (on the front page, it's ok elsewhere.)

• 'news' is difficult to find from front page
  • all the links on the front page point to 'unfriendly' coldfusion URLs (then rewrite to 'friendly' URLs)

• Many many broken links - for an illustration of just how emabarrassing this is, see http://www.google.co.uk/search?hl=en&q=university+of+oxford&meta=&btnG=Google+Search

  • It may also cause problems for Google (local and public) when both forms of a URL get indexed
• on a 1024x768 screen, the A-Z of colleges boxes take up so much space that you can't see the results underneath without scrolling
• [if you turn off Javascript, the headings on the home page pop up a big picture you cannot remove (and then hitting the back button does not take you to the home page, but a previously visited site)] - actually, you can remove it by clicking the 'close window' button on the top RH corner.

• attachment:BugRiddenScreenShot.jpg screen shot showing lots of bugs (FF 2.0.0.8 on Win XP)

• Why weren't stakeholders given a preview of the new page and a chance to comment?
• The index page references 73 images in its source, very few of which are visible when you initially view the page (as they are hidden by css). This is an unfortunate blend of the disadvantages of images (the 73 images total nearly 2 mb) without the benefits (the visual richness which images give a page). The heavy size of the page will make it very hard for people on low bandwidth links to access the page. This is bad for any page but doubly so for an index page, as it is a common point of entry.
• [There is no mention of 'Information for Staff' on the front page] - actually there is, it's the 'Staff' link under the 'quick links' drop-down. Not obvious but there.

• Links to http://www.ox.ac.uk/aboutoxford/maps/map.pdf are broken -- this was one of the most popularly linked pages on www.ox.ac.uk!

• NO COLDFUSION ERROR HANDLING: (in certain applications) e.g. http://www.ox.ac.uk/applications/dynamic/map.rm : it does seems to exist here though: http://www.ox.ac.uk/track.rm but not in certain parameter contexts : http://www.ox.ac.uk//track.rm?id=

• (minor) http://www.ox.ac.uk/enterprise/innovation/intellectual.html has an unnecessary '>' on the first line of text

• IE7 freezes when clicking any of the main headings (Admissions, Divisions, Colleges, etc) on the homepage. Image animation is displayed but overlaying text never appears. Browser then becomes unresponsive. Tested on Vista and Windows XP.

• Try browsing http://www.ox.ac.uk/index.html with images off.

• The use of img tags without an src attribute is not only invalid xhtml, but results in a "missing image" icon being displayed in IE6. (eg at the bottom of this page: http://www.ox.ac.uk/this_is_oxford.html)

• On my screen (Mac Pro, using Firefox or Safari), the "blue" of the web site is quite different from what I would think of as Oxford blue, and certainly different from the previews I get (using Mac Preview) of the .eps files in the Downloads section.

• The raw code confirms the colour difference - http://www.ox.ac.uk/css/front/structure.css includes background: #002D62 (#header-wrapper), whereas http://www.ox.ac.uk/staff/branding_toolkit/the_brand_colours/index.html states #002147 should be used.

• Colleges A-Z shows Queen's College. Correct title is The Queen's College.
• It takes a good fifteen seconds to load on my PDA, and the layout thereon is horrible. However, I'd like to speak in favour of its design when viewed in a more normal browser: it's simple, sleek and sophisticated.
• in Opera Mini (on a Nokia 6230i) the homepage's "Skip to content" link appears as the last link on the page.

• Many of the colleges are positioned in the wrong place on the individual (Open StreetMap) college maps:

  • All Souls, Brasenose, and Corpus Christi, appear in Christchurch [sic] Meadows
  • Green, Kellogg, Somerville, and St Antony's all appear on Winchester Road instead of Woodstock/Banbury Roads
  • Harris Manchester, New and St Cross all appear in a field outside Kirtlington
  • Nuffield is a couple of streets away from where it should be
  • Oriel is in the same place as Pembroke