My old S60 series Nokia - it has offline maps, with driving instructions (and voice guidance) and a working GPS. I got a car-window mount and a recharger for that (cost about â 10) and now it serves as a navigator in my car. I connect it via USB every few months to load in the latest map data, but other than that, it now lives in the glove compartment when not in use.

I got my MSc 10 years ago. Anyway, I'm not sure how these would apply to your situation (wanting both MSc and PhD), but basically I immediately enrolled for post-graduate studies afterwards. I didn't pick any topic for dissertation or anything like that - for the next 7 years I basically did all the required coursework. Typically it was like 1 or 2 seminar courses each semester, so it was typically one afternoon a week. I didn't really have a study plan as such - seminar courses were interesting for their own sake because the courses were structured around stuff that people were really doing.

For example, one course was about P2P networking, DHT algorithms and so on, and a group came in with a variant of Bittorrent that organised the chunks so that BT could be used for streaming. (This was like 2005-2006). At the time, it was a very good way to keep oneself informed on the latest developments without sacrificing too much time.

Anyway, in 2009, I had basically done all the courses, and partially by chance I got laid off with a rather nice severance package, at which point the professor who's going to oversee my defense on Friday called me and asked if I'd like to come and work in his team as a researcher and create the actual thesis. At the same time an ex-colleague (who had been at the same company) got in touch and basically told me that the academia is going to do squat with the CCIE - they have this little network shop going on and if I like, I could work for them as much as my time would permit. The arrangement worked out rather beautifully....for example, I could incorporate real-world problems from our customers to my research work as well.

So the only real advice I can give - unless you are really going for full-steam career in academia and a tenured professor option - stay in touch with industry somehow (part time, consultancies, even participate in trade shows if nothing else). In academia, it is really easy to seal yourself to that ivory tower that has no bearing on real-world issues and lives from one grant money to the next. Even in completely academic circles, it's possible to maintain that touch - besides everything else, for me it meant regular trips to IETF, which is not something you can exactly classify as a conference. (And at least I got my name as an author on a single RFC to show for it).

I'm going to defend my PhD dissertation on Friday, actually. Anyway, the problem is that in academia, there's sometimes a big disconnect between what's happening in "the real world".

Anyway, the way I've handled it is that I've basically kept my feet in both camps. Throughout the research work I also worked part-time for a consultancy (now full time since I'm done). I'm also a CCIE. PhD alone might mean "an absent-minded professor" to a recruiter, but combined with credentials from industry side I've at least so far gotten the feeling that it's a big selling point. Which might mean more $$$$.

Most clients will NOT be happy if a hot sweaty engineer turns up on a bike (even if he did then do an excellent job because he wasn't scared of climbing through a few ducts to find issues).

Depends, if the hot sweaty engineer knows what he is doing. A bit off topic, but kind of fits...

My colleague in sales told of an engineer he used to work with in his previous company - the engineer took a vacation - went basically off the grid for 5 days, surviving by hunting and fishing in the wilderness. Afterwards, he was driving home on the early hours of Monday morning, intending to shower, change and go to work. However, he got a call stating that he'd be immediately needed at customer site to give out technical details on implementation - no time to freshen up.

So a guy who has been better part of a week in a forest, basically with one set of changes in clothing shows up, unshaven, hair in tangles, and reeks of gutted fish...and completely unprepared gave such a presentation that the customer was sold on the solution. Afterwards the sales guy heard comments that the technical presentation was clearly the most convincing of the ones seen so far...and only presentation tool he used was a whiteboard.

So, sometimes, it's the substance that matters. Same sales guy has told me to preferably show up to meetings with technical or geeky T-shirts, or the customer wouldn't get the impression that we actually have the skills (his view is that if I'd show up in a suit, the customer would get the impression that we are just some sleazy people in suits who are all glitter and no competence). I have no objections...He actually told me the previous anecode as his reasoning why.

Have to answer with a mix.

I telecommute from home, and for such purposes I have a Cisco/Tandberg E20 videophone on my desk, and I use that to make both regular and video phone calls. I use e-mail a lot, including mailing lists. I use Lync, Jabber and Skype for instant messaging and desktop sharing. I use my cellphone when I'm not at my desk. And I visit office like once a week for face-to-face meetings (and customers whenever they need, of course, but the travel expenses are of course reimbursed).

The only problem with such multifaceted reachability is that it's hard to disconnect completely from work during vacation season. If someone at work insists on bothering me during holidays, even for 5 minutes, it still brings the work-related matters to the front and essentially bringing me back from vacation-mode.

I mean, Lync (with federations and so on) pretty much covers every "locally admistered" instant messaging + file sharing + desktop sharing + voice + video needs you have...and it works across organization boundaries. So what is Yammer for?

Since you included fantasy. Works perfectly. My friend is currently reading it one chapter a night for her 7-year-old - and he just loves it.

On the sci-fi-side, others have mentioned lots of examples - but you could try giving him book versions of Star Wars and Empire Strikes Back - and then the Timothy Zahn's Thrawn trilogy. Not so serious but the stories flow smoothly. And they have space battles with lots of turbolasers.

This is only true if your definition of "full-featured" does not include KMS or complete XRandR support.

They added XRandR 1.2/1.3 support in 302-series.

See e.g. http://www.phoronix.com/scan.php?page=news_item&px=MTEyMDk

Still have my N900. It has it's flaws, but unfortunately there is no real replacement. The repositories are full of interesting software, I can run even stuff like Wireshark and the like (which is kind of nice when you are a network engineer and occasionally need to debug stuff over mobile network).

N9 has no keyboard, and is pretty much a dead end since Meego is pretty much dead with Nokia going the way of Windows Mobile.

Anyway, one day the hardware is probably going to die - is there *any* phone that would come even close? With Android I need to pretty much tie myself to a google account - sure, I can root Android phone but then I pretty much have to start scaveging the net for software. It's been pointed to me that there are several alternates to Android market, but no one bothers to list examples.

From what I've seen on my wife's phone about software for Android - it looks like *everything* is adware. For Maemo everything seems to be based on common open source software. I can run stuff like Stellarium, Ur-Quan Masters, ScummVM, AbiWord, Gnumeric and stuff like that...Heck, N900 was the first phone where I could get IPSec VPN to work properly - I can just get the normal openVPN up and running. Oh, and sharing your 3G net connection over WLAN works too - there's of course Joikuspot, but for other systems it costs money, with N900 I could just set up ethernet bridging + ad-hoc WLAN like with a common Linux box.

It looks like I could get most of these functionalities on an Android phone, but I'd have to find ad-loaded alternatives, tie the phone to a Google account - including having to sync phonebook and calendar with Google...kinda feels like step backwards.

Get a Palo alto firewall. You can filter by application, and even make firewall rules like "allow reading of facebook, but disallow posting", or even "disable attachments".

Of course, you didn't exactly specify budget...

I had two old computers - a 386SX, 16MHz, and a 486DX2/66 not so long ago. I actually booted them up occasionally - for the express purpose of running old games on them (DOS-based).

However, when I moved last summer to a new apartment, I was able to dump them for good. DOSBox pretty much has removed the need to keep real hardware around for nostalgia. Only thing I wish is that they'd get the Roland MT-32 support to work *completely* - especially the sound effects still could use some work.

It looks more like Don Martin to me...

Interestingly, the specifications (below the image of the processor box) say "Multi-core: Dual core", not just the item name...

Like I said in grandparent, I have been able to make using Latex tolerable by using Lyx. Only problem is that if there's some newfangled class that doesn't have a corresponding Lyx layout I have to somehow try to make one...and that's not really all that easy.

What's unfriendly about the following command?

\cite{some_book}

The fact that if I typo it to \cite{some_booky} it doesn't compile. And unless I rigorously recompile after every edit, I might not even catch that. Worse if the brace is missing.

What's unfriendly about the following entry?

It's in a separate file, for starters.

Also, you need to run Latex *twice* to get it working properly (the first time generates the .aux and then you can do it again). Oh right, creating a makefile is apparently easy for everyone.

And like I said, if you forget a comma at the end of the line, it doesn't work.

Oh, I guess I could just use readily-made citations that I can copypaste in from ieeexplore and the like? Well, guess what, the readily available bibtex exports are crap. For example, the bibtex containing all RFC:s (http://tm.uka.de/~bless/bibrfcindex.html) have all sorts of stuff in them that shouldn't be included (including standardization status and what RFC's it obsoletes). When I wrote my latest paper to Elsevier that included lots of RFC references, I basically had to run that .bib through a bunch of perl scripts with lots of regexps to get rid of all the cruft. Same has happened with most other readily made citations. At least the Word's XML has enough of the damn fields that you can pick'n'choose what to include in the reference. With bibtex, I have basically resorted to turning everything into @MISC.

So easy....not.

Only problem I have with Word's citation mechanism is that there isn't an easy way to get citations directly that format, but I have been using Bibutils (from http://sourceforge.net/p/bibutils/home/Bibutils/ ) to get back'n'forth between various formats.

I'll give Latex that it produces the most neatest documents there are, but to get that far you end up fighting all sorts of indicate details far too much. Don't even get started on how to create a new document class - if your text doesn't quite work with any of the provided classes and you'd like to create your own styles, good luck.