I just sent this to microsoft... and I figure it's a good idea so I'm posting it here aswell :]

Since I still have to use IE for work (web designer..) I have an excuse for giving Microsoft good ideas too ;]


-- Max's Microsoft push follows... ---

We all know that the page download progress bar in IE5 and IE5.5 is practically useless. The blue bar at the bottom of the screen will grow regardless of data being recieved. This is also a source of confusion among customers, which working for an ISP helpdesk has taught me. People are quite right to presume that data is being recieved when the bar is growing but in fact there is no correlation.

My suggestion to fix this issue is this:

When first visiting an URL the bar will grow according to a presumed amount of data on a page (including images), say for example 100K. The amount of data recieved will be shown as a numeral on top of the bar, x.xK. The bar will not grow if data is not recieved.

When the page has loaded the size of the page will be stored in the browser history. The next time the page is visited the assumed size will be 110% the size of the page, thus minimizing the probability of overshoot.

If the page is larger than the default or assumed amount the blue bar will simply stop growing when the whole area is filled, but the numeral size representation will still allow the user to see that data is still being transfered.

Hopefully this suggestion will be incorporated in IE at some point.

Sincerely,

Max Romantschuk

--- EOF ---


Wow... you bothered to read it all, nice one :]

This would be more appropriate to submit to the mozilla folks, since embeddors have very little control over the calculations used. As a web designer, you may know that, as defined in the HTTP protocol, the server will send a "Content-Length" header for every URL requested. Therefor, assuming the server sends valid information (and I don't know of an instance where it wouldn't) all it would take to know the exact amout of data in the entire "page" would be adding all this information together.

The problem is that the browser will parse the HTML page as it is loaded, and begin downloading the other content (images/css/etc) immediately. This makes it impossible for the browser the know the size of the page until it has completely received the HTML file.

The most accurate solution, would be to display the total progress (xx.xK/xx.xK), and the corresponding progressbar, but update the totals with each concurrent URL request. This would probably lead to the progressbar that jumps all over the place as it displays the progress of the HTML, then the HTML+first image, then the HTML+first image+second image, ad infinitum...

The most practical solution would probably be displaying the progressbar showing the status of the HTML file, then resetting the progressbar when that has finished to show the progress of the rest of the content. This may confuse users, however, if their progress bar suddenly jumps from 100% to 20%.

-- Jeff

isn't that how it works now? (displays progress of the html, then of the images) I know the current kmeleon bar jumps around madly when a page is loading.