Running Redis on the same server still adds overhead to the system, to store an object in a map that's on the Java heap doesn't require any inter-process communication or serialization. Plus it's another point of failure, what if the Redis process stops but the Java process is still running? If Redis is remote, what happens if the network goes down? Keeping everything within the JVM avoids having to think about those failure scenarios.
I don't know how Mailinator is coded, but I'd guess he hasn't written anything too custom for storing key/value pairs in memory, there are plenty of Java caching implementations, the key thing he has done is work out a way of reducing the duplication, which Redis won't provide out of the box.
Mailinator is a great example of having the minimal set of features required and no more. There's no logins, no setup steps and no guarantees. This allows Mailinator to run on such minimal hardware, adding disk persistence would add complication without making the product better, by definition the kind of emails you send to Mailinator aren't important, so if they're lost during the occasional restart it doesn't detract from the service.
The polymer notes[1] we use in Australia (and also print for several other countries) have been around for about 18 years now and have very low rates of counterfeiting due to the number of security features built in[2].
The notes do cost more to produce, but the extra durability means that they last about 4 times as long (from memory) in circulation, so overall the cost difference is marginal for the government, but is an additional deterrent for counterfeiters.
There's also tools such as this http://tools.pingdom.com/ from Pingdom that'll show the same figures from an independent network link.
From there, look at how long it takes to receive the HTML (not including any of the images, CSS or JS), if that's slow then it's likely that the web server is taking a long time to render the templates server side (it could also be due to a really slow connection...)
If most of the time is spend downloading a lot of images, CSS or JS then YSlow will highlight this and the areas that should be improved. YSlow will also highlight any issues you might have with not setting cache headers correctly.
If the site doesn't contain excessive images, CSS or JS, then the connection might be slow, the 500kb background image will give you a decent indication of the connection speed, though something bigger would be better.
Also, if you have access to the logs then most web frameworks will log out the time taken to render each page on the server.
There's a similar geographically accurate map on wikimedia, but without the simplifications, such as making the section of the circle/district line through from Notting Hill Gate to Edgware Road straight. Removing some of the curves and the other changes, while relatively small, really makes the map easier to read.
http://upload.wikimedia.org/wikipedia/commons/0/08/London_Un...
The 500MHz processor in the original Pre really does hold back WebOS. Running an overclocked custom kernel, I have mine set at 1GHz, makes a huge difference to the usability and surprisingly doesn't reduce battery life by much (which is partly due to the custom kernels clocking the CPU back to 500MHz when the screen is off).
It looks like HP/Palm have learned from this and aren't going to make the same mistake twice, the more recent videos of the Touchpad show how much they've optimised the UI to remove the lag:
http://www.precentral.net/webos-3-0-optimizations-make-smoot...
He works around this by reading the email himself first, clicking the option to always display images from the sending email address and then starring it so the email is always at the top of his inbox even though it has been read.
I suppose an attacker could bypass this method by turning off images in their browser though.
I think part of the reason that dropbox's video works so well is it shows you how dropbox solves a problem, if they just had a technical demo of how their stuff synchronises files then techies would see it and realise how it was useful, but a person on the street probably wouldn't.
I can see where your app would be good, but I'm a developer and I have friends with English as a second language. If you had a video that showed someone catching up with a person they'd met on a holiday in another country or something similar then I think it'd put the problem you're solving in context better than the demo.
Getting your first job in a field can be a pain, if I had a dollar for every time I saw an ad for a graduate position that wanted experience that you could only get by already having a similar job, I'd have been able to retire early...
Companies look for experience since it demonstrates your ability, do you have a design portfolio, even if it's just mock up sites you've done in your own time to learn?
I've done a fair bit of hiring for the company I work at (unfortunately we're not hiring at the moment) and if a candidate with little experience can show us the kind of things they can build we regard it more highly than someone who has had a job for a year but hasn't shown much initiative beyond being given a task list and doing it.
I don't know how Mailinator is coded, but I'd guess he hasn't written anything too custom for storing key/value pairs in memory, there are plenty of Java caching implementations, the key thing he has done is work out a way of reducing the duplication, which Redis won't provide out of the box.
Mailinator is a great example of having the minimal set of features required and no more. There's no logins, no setup steps and no guarantees. This allows Mailinator to run on such minimal hardware, adding disk persistence would add complication without making the product better, by definition the kind of emails you send to Mailinator aren't important, so if they're lost during the occasional restart it doesn't detract from the service.