Urs Hunkler

Hi Tim,

I am glad to give you a little info back after I have read so many helpful eclipse infos from you:

Select a word and move to the next occurrence with CTRL-k and to the previous with SHIFT-CTRL-k

Urs

"webapp" I have written in quotes because the Moodle interface on the "Touch" device looks and feels like other apps on the "Touch" devices. There is no specially developed app. The interface part is a theme with CSS hiding or changing Moodle page elements. So it's a special theme to adopt the look and feel of Moodle pages for the "Touch" devices.

The book module I have rewritten to deliver "Touch" specific data. Either the TOC or the page content and not both at the same time.

The iPhone simulator as a part of the SDK I will use to show Moodle in the iPhone over a beamer. And I will have my iPod Touch with me

It might be interesting to find out, if the ability of the "Touch" devices to store data in an own DB on the device can be used for working offline.

And with the Apple "Touch" products it's not only about technology and administration but also very much about the great user experience. The "Touch" products are addictive and therefore motivating

An early prove of concept I have designed/developed shows Moodle on a "Touch" device. As a "webapp" with the "Touch" look and feel - not as a normal Moodle site in the browser. I am going to show the prototype on the Moodle conference in Heidelberg next week.

Course:

Book (modified book interface):

Quiz:

Mauno, I decided to remove the Chameleon engine in the Custom corners theme. The CSS is somehow too complex to be easily handled in the Chameleon interface.

After further investigation I can show new facts and figures.

First the much slower page rendering on the PC in Firefox was caused by a broken extension. After I removed the extension page rendering takes only a little longer due to the slower CPU.

Second no need to worry about the theme "Custom corners" rendering speed. It's not significant slower than other themes. When files don't come from the cache the load times differ more. After some JavaScript enhancements it's performance got even better.

Page data

In my tests I wanted to know the page creating time and the time until the page is ready displayed in the browser. To get an overview I selected 6 pages: startpage, 2 course pages (one short and one longer with several images), a glossary, and 2 admin pages (frontpage settings and user settings.

Measuring

To get a good average I measured in Firefox:
_ 10 rounds loading the 6 pages with the standard theme as a reference
_ switch to the other theme
_ 10 rounds loading the 6 pages
_ closing the browser (to avoid slowing down due to eventual memory leaks)
_ 10 rounds loading the 6 pages with the other theme
_ switch to standard theme
_ 10 rounds loading the 6 pages with the standard theme as a reference

Therefore standard theme and the other theme alternate in the graphs.

The graph shows that there are only quite small differences in the loading time for different themes. The pages with a theme with several JavaScripts needs a bit longer, but not dramatically.

With text replacement filters active the server needs to work more, but the page load time is only minimal slower with the standard theme. (Shown in the middle before the numbers with the disabled cache)

The higher load times on the right show how much longer pages load when the browser cache is empty. (for example with first page loads or https connections).

(click to load big image)

Load times of the single pages comparing standard and custom corners theme

Quite interesting is also looking at the different generation and loading times of the 6 page types.

The graph for page generation time on the server with cache on and off

(click to load big image)

and the page loading time in Firefox with cache on and off

(click to load big image)