I have just run into what I thought was a huge IDE bug or problem. I had placed several breakpoints in my web app to debug a few lines of code and test their values. When I ran the app, the IDE totally ignored the breakpoints. I then placed a message box in the code to tell me when I had reached a certain point, but the message box failed to pop up when I ran the app. Thinking that I had done something else wrong I remmed out the message box and deleted the breakpoints. When I ran the app again, the program stopped on the breakpoints and the message box popped up as originally planned. I was now convinced that I had gone quite bonkers and was getting ready to hit the forum when I thought that I had better clear the Safari cache just to be sure. I was amazed to see that testing of my app had returned to normal. The point of my conversation is that if you are getting weird results while testing your web app in a live browser, periodically clear your cache. It worked for me.
Interesting, I hadn’t thought of that. I’ve occasionally had REALLY strange things happen in my web apps and a restart would fixes it. I call it the “bat shit crazy fix”. Guess it wasn’t so crazy after all.
Although my wife and kids say that that’s really not proof. 
One should never consult the wife and kids on issues of sanity as they will always default to null.
CTRL+F5 will clear cache and reload the page in most browsers
Doesn’t do anything on OS X browsers 
So only 90% of the market then?
‘Clear caches’ appears only when you have the ‘develop’ menu up on Safari
If you only count only a decreasing segment of the overall market - sure.
iPads + iPhones + Macs outsell PCs by themselves. (could probably drop the Macs portion and they’d still outsell PC’s)
Safari runs on all of those as do several other browsers.
Toss in Android phones & tablets where I’d guess it also does nothing and PCs represent an decreasing segment of devices that consumers are acquiring.
Never mind the drop in PC sales that’s been ongoing for several quarters.
Even retiring XP hasn’t spurred renewed buying of desktops / laptops.
People are largely ignoring new PCs in favor of tablets - a market where Windows barely has a foothold.
If you look more broadly at “devices people are using”
http://www.gartner.com/newsroom/id/2793921
http://pocketnow.com/2014/05/06/q1-2014-shipment-numbers
430 million non-pc devices vs 75 million PC’s (worldwide)
So IF you look narrowly at “Desktops” Sure Windows is some 90%
If you look at personal computers (which I’d contend every phone & tablet is today) then its some fraction of the market - maybe 25% of the devices
And FWIW Apple is the ONLY vendor that reports actual sales - NOT shipments
SHIPMENTS aren’t a good measure - the Surfaces proved that when MS had to write down 1 billion on unsold units in the channel
They were “shipped” but not “sold”
The only number that matters IS sales - shipments never have
Because they don’t allow competitors to ship their products. Apple is the MOST closed shop in the market - it has it’s benefits I’m sure.
I see quite often that conversations run off the rails quickly. This tread is an example, within just a few replies a technical issue about breakpoints and debugging launched by @chris benton , goes to a discussion about marketing, sales and pc shipments. Why is that ?
Within this tread I am just interested in what is really going on around the problem Chris is reporting.
Hold shift and click reload 
[quote=120576:@Joost Rongen]I see quite often that conversations run off the rails quickly. This tread is an example, within just a few replies a technical issue about breakpoints and debugging launched by @chris benton , goes to a discussion about marketing, sales and pc shipments. Why is that ?
Within this tread I am just interested in what is really going on around the problem Chris is reporting.[/quote]
Thank you for re-centering the conversation. This perpetual “me got biggest shit” is puerile.
Back to business, what Chris discovered makes total sense : the caching mechanism in the browser is destined to prevent excessive download of resources. One can only suppose that from the browser point of view, if the UI page presented to it looks identical, it will use the cache instead and as it does, the server-side app has no chance to run accordingly.
This can be an issue. Would pushing something with control.refresh help ?
I guess I helped things go astray a little by going off track while responding to Bob Keeney’s input, but I have to say in response to Joost’s comment. I agree that all too many interesting topics take a turn and start discussing fishing tackle or some silly thing. Finding the reason that my web app started doing stupid things was a great victory for me and gave me the opportunity to contribute instead of asking . What made the whole caching thing confusing for me is that the problem did not appear on the first run of the app because it had to run at least once in the browser for it to be cached. I am surprised that there are not cautions in the web version help file.
File a feedback and I’m sure Paul will add the warning
shift-command-E (on a Mac) will clear the cache