[thelist] Web Page Analyzer -- "incredibly useful tool"?

[Clive R Sweeney]

Tim Beadle said the following on 8/2/2004 3:50 AM:

>I emailed Andy King (of WSO/Speed Up Your Site) and he said that at the
>time (March 19, 2004), images included by CSS (e.g. Image Replacement
>images) were not being counted in the manifest, but that the next
>version would fix that.
>
>Looking at the version history [0] it seems that these bugs were fixed,
>but if you're still seeing differences, I would urge you to report them
>to Andy King, using the contact form [1]. He's a responsive chap, and is
>trying to provide a Very Useful service to the community, so anything we
>can do to help him improve the service will be good for everyone.
>
I agree that the Web Page Analyzer should be a very useful service, but 
I'm definitely experiencing some strange glitches when I use it. I've 
tried a few other sites and come up with more strange results. For 
example, the evolt.org home page (http://www.evolt.org/) is reported by 
the Analyzer to require a download of 293KB, but when I check my cache 
the files for the same page total 49KB.

Here are some excerpts from the Analyzer's report for the evolt.org home 
page:
"TOTAL_SIZE - Warning! The total size of this page is 300027 bytes, 
which will load in 59.99 seconds on a 56Kbps modem. Consider reducing 
total page size to less than 30K to achieve sub eight second response 
times on 56K connections. Pages over 100K exceed most attention 
thresholds at 56Kbps, even with feedback. Consider contacting us 
</contact/> about our optimization services....
"IMAGES_SIZE - Warning! The total size of your images is 255200 bytes, 
which is over 30K. Consider optimizing your images for size, combining 
them, and replacing graphic rollovers with CSS."

In this case, it seems to report the correct number of images (they're 
listed individually in the report), but the file sizes are way off. For 
example, the "spacer.gif" is supposedly 31850 bytes, although the one I 
see in my cache is 49 bytes.

So it would seem that there are different problems happening here. Of 
course, not every report from the Analyzer is so far off. I've mentioned 
three that were significantly different from my own results 
(Designshift, Stopdesign, and evolt.org), but I checked 4 or 5 others 
that seemed fairly accurate.

-- 
Clive R Sweeney
Designshift | Durham NC