This!

From a true scientist's point of view, full open source is critical. I avoid closed source to the very bitter end because I know that there's always another error/mistake to be found and that's less likely to found in closed source.

(Then there's a bunch of folks who just want to be first and get another manuscript accepted and who care zero about reproducibility and sustainable science.)

Thanks for the detailed explanation. Sorry if I mislead anyone (it was my honest belief).

Thank you for clarifying and correcting me on this one ... and I'm glad that I was wrong.

Shouldn't we call it "free trial" rather than "free"? As far as I understand it, we're going to have to pay up after the first year of "trial". This goes for everyone including us who got a full Windows installation.

Sorry, but the US is not efficient. It's a big patchy system with no long term plan.

But flashing it manual will wipe all the data whereas the OTA installation won't, so the former will be a more time consuming process in the end. That's why I'm waiting - simply too little time.

Everyone in the US, but in the rest of the world almost no one knows who Lou Gehrig was.

It's better to call deceases by their medical names and not based on some person's name.

Thanks. However, last time I played with Chrome Remote Desktop (on Windows) it required interaction on the receiving end (e.g. initiation or at least approval). I'm looking for something that requires zero interaction on the other end, i.e. "Just start the computer and I fix the problem for you". See, my mother sometimes get stuck when another windows gets in front of the web browser and she can do nothing, have an even harder time explaining the problem over the phone - it's at that level. So I need to be able to at a minimum log in remotely a view the screen - ideally take over a run it.

[Serious question] I'm currently helping my mother on Windows remotely via TightVNC and it actually works alright. As long as it boots up I can do most/all work from across the Atlantic. My mother does have to do anything - no "click to accept" or similar.

Now, what's a similar option for a Chromebook/ChromeOS? Anything that just work without my mother's interaction? I need to be able to get full control just as if I was in front of the machine.

Related, being able to remotely control the device is also why I've retained from getting here a, preferably an Android, tablet, because of the built-in security making remote access tricky. I understand that one get around it if one root it, but if imagine the user base doing that is so small that the quality of such software is just not there yet, or?

True that native code is faster than plain R, but from your benchmarking it sounds like you're not doing things properly in R. If you know your vectorization and how and when objects are copied/allocated, then R is super fast. Nearly as fast as native code, because you're effectively using R to call native code with minimal overhead.

People with shorter telomeres drink more sodas.

Solved a related issue a couple of years where we wanted audible feedback on the keyboarded being types. The use case was a complete mute person Text-To-Speech (TTS) generator to talk over phone/VoIP (in this case it was Skype). Without being able to hear the keyboard clicks the receiver of the call gets long periods of silence while the "speaker" is typing the next word. This silence is confusing and often leads to interrupts from both parts, similar to the problems one gets with walkie talkies. By adding virtual keyboard typing sounds, the other party (the "listener") knows that the "speaker" is typing and naturally waits. It was really helpful! (This was a Windows machine and the solution we ended up with was to use Sound Pilot [http://www.colorpilot.com/soundpilot.html].)

Kudos to IamTheRealMike for such an informative and well written summary.

From http://keepass.info/help/v2/autotype_obfuscation.html:

"The Auto-Type feature of KeePass is very powerful: it sends simulated keypresses to other applications. This works with all Windows applications and for the target applications it's not possible to distinguish between real keypresses and the ones simulated by Auto-Type. This at the same time is the main disadvantage of Auto-Type, because keyloggers can eavesdrop the simulated keys. That's where Two-Channel Auto-Type Obfuscation (TCATO) comes into play.

TCATO makes standard keyloggers useless. It uses the Windows clipboard to transfer parts of the auto-typed text into the target application. Keyloggers can see the Ctrl-V presses, but do not log the actual contents pasted from the clipboard.

Clipboard spies don't work either, because only parts of the sensitive information is transferred on this way.

Anyway, it's not perfectly secure (and unfortunately cannot be made by theory). None of the currently available keyloggers or clipboard spies can eavesdrop an obfuscated auto-type process, but it is theoretically possible to write a dedicated spy application that specializes on logging obfuscated auto-type."

Having access to the remote machine(s) will be very useful, because there will always be something that breaks and you'll need to admin that computer. I maintain a Windows 95 computer trans-Atlantic via remote VNC access. Basic setup:

Dynamic DNS:
1. Signup for a DynDNS.com account. Free.
2. Install the DynDNS client IP updater on the remote machine. Having it update its IP number at the DynDNS.com account. This will