I had something similar to your experience at an airport. I saw 5 computing devices (not counting phones): a Macbook, two iPads, and what looked like an Android tablet, plus my own Linux laptop. Not a machine in sight running Windows. That would have been unimaginable as recently as a year ago.
According to the stats I have seen, this is the year that the number of Linux systems in use (in the form of Android) will exceed Windows systems. In current sales Android is already blowing out Windows and, for that matter, everything else on Earth. The year of desktop Linux has arrived, but not in the way everyone was expecting. High fives, for now at least.
The existence of things like the Nitdroid project and Jolla/Sailfish, plus the fact that N9 matched or possibly outsold the Lumia crud despite the massive disparity in corporate support, shows how good the technology was. The last time I saw this was...ooh it hurts to type this name...the Amiga. Commodore went under, not due to poor demand for the Amiga, which was profitable to the very end, but due to massive losses in PC clones that led them to credit default with their suppliers. They couldn't get parts to build Amigas anymore.
I hope that the prevalence of open source these days gives the keepers of the Meego flame more success then the Amigans had.
I would make a stronger point about the removal of the floppy: Apple has an indifferent attitude toward removable media generally. The iPod, as hardware, would make a perfectly serviceable external hard drive, but software support for this was thin, and we also have the absence of SD slots. The organizing principle seems to be to have iTunes managing data instead of a file manager. iTunes knows more about the (meta-)data and can handle it with more aplomb than a general purpose file manager. That comes at the cost of limiting its scope (taking things out again) to AV media and rinky-dink apps, and keeping removable media at arm's length (copy and import rather than open up and look inside); using iTunes to manage CAD files or source code would be ridiculous.
On Gnome, I don't have much complaint. With Ubuntu and a fast connection, it is very easy to add in stuff as you need it, "Golly, I need XYZ...sudo apt-get install XYZ". My bigger beef is with KDE, which seems to be built by people who have spent too much time looking at Macs in the Apple store and not enough time trying to use Macs to do actual work. More broadly, I find that everyone in the industry apes Apple too much; Apple has their niche, and serves it well, and profitably, but it can never be more than a minority of the market. Workhorse products serving the bulk of the market will be made by others.
That is sort of the crux though. Steve Jobs takes things out: the SD slot, the keyboard, the command prompt, the floppy drive. Taking stuff out makes it possible to simplfy the user interface. If anyone else did this, the product would fail in the market for being feature-poor, but he could get away with it because he was trusted (by some, not all) to do it well. Granted, some of the stuff Jobs took out should have been left in, the lack of fan on the early Macs comes to mind. That is the charisma, the RDF, or maybe just street cred for having introduced the Mac. Whatever it is, it accrued to him personally, not to Apple as an organization which went adrift whae he was gone.
That also explains the stupid holy wars that surround Apple:
- Look at Apple's new product! It is so innovative and easy to use! (Because he took stuff out and simplified)
- It is not innovative! Everything there was already done by others. (True, but that is not what makes it easy to use)
- Apple's stuff is merely a toy! (It is lacking certain features, but that is the price of simplicity. You are not the intended market.)
- Apple is an industry leader! Everyone copies them! (Now everyone else can take those things out without being accused of being feature-poor.)
I would say the walled garden metaphor misses the point, or at least it gets taken too far. A better way to think about it is delegation and trust. With Linux, you can, in principle, trust no one and collect all the cource code yourself and compile it. The next step is to delegate the job of compiling code to a distro, trusting them (or for Slackware, Pat) to use only open source software in keeping with the distro's community values. Various other relationships can go from there: you presumably trust your employer-provided laptop not to spy on you, and they trust you not to improperly divulge company information on it.
I would say that Steve Job's real innovation was to be cognizant of this delegation of trust, and he was good at getting people to buy in: Call it Charisma or the RDF. Apple products are offered as a package deal, take it or leave it. If you want the nice ease of use and aesthetic package, you have to play by Apple's rules. No modding, no tinkering, no exceptions, no whining. I never cared for the deal and always walked away, but I would admit that it is appropriate for some (most notably the very old and very young).
Yahoo, well, can't say I have an opinion of what it is like now, haven't used it since 2004. Does it still exist?
A few months ago, I did an analysis of the list of parents' email addresses from the school drama club. Yahoo was the most common provider with about a third of them. After that was the local DSL or cable ISP. Third place was people using work email addresses. GMail was fourth at about 15%. Then Hotmail/Live. Only about 1%, myself and one other person, were using a paid 3rd party service. For the record, I use usermail.com and am very happy with them, and seldom use my Gmail account.
What conclusions can be drawn from that? I would say Yahoo-email's first-mover advantage is much more durable than Friendster's or Myspace's was. Or perhaps Facebook's is. I would think that a survey of younger people would have fewer yahoosiers and more GMailers
But what does it mean to be "backed by something real"? Very few currencies are redeemable for specie anymore. And being backed by a commodity isn't enough: Confederate bonds backed by cotton circulated in Europe until WW1, but their value tracked the prospect of redemption more than the value of cotton. Government currencies are backed by courts, who pronounce judgments and settle debt in terms of fiat currency: For a stark example, US account holders on e-gold.com quite emphatically cannot claim their gold, but only a court's judgment of its dollar value. And having courts enforce the value of a currency is not always enough either. The Zimbabwean dollar lost its status as money despite draconian efforts of the government to enforce it and ban alternatives.
But government courts are not the only institutions that can enforce the value of a currency. Rai stone coins on the island of Yap don't have much intrinsic value, and the one that sank to the bottom of the ocean has none but circulates as money just like the rest of them. They are not backed by anything other than tradition and social norms. If social norms are enough to turn a hunk of limestone into money, it could happen to bitcoins, too.
Your larger point that money is a social phenomenon is valid. I like to describe money as "a claim on the labor of others", so money doesn't make sense in isolation.
Even so, that leaves unanswered how we keep score. A physical token of a well known luxury material? A book entry in a government-chartered bank? Or a cryptographically-verified transaction history of proof-of-work hash values? When confidence in banks rose, fiat money replaced gold coinage. The real bitcoin experiment is whether confidence in crypto systems is enough to sustain the bitcoin economy without the help of government courts. If the bitcoin economy matures from the mostly speculative transactions we have seen so far to a more stable role as an abstract claim on others' labor in a diversified economy, then it has as much claim to the title of "money" as anything else. I am not ready to place better then 50-50 odds that will happen, but it could.
That concurs with my experience when I have to dial-up around the Great Firewall (which is blocking news from Egypt these days...don't want to give the natives any restless ideas...). I typically have to fall back to 9600 or 2400 bps, which is good enough to make an independent check on SSH key MD5 sums, at least. Modems signals are pretty standard down to 2400 bps, but the slower 1200 and 300bps standards are different between Europe and North America. Russia used North American standards, because the Soviets of the 80's were more accustomed to stealing American technology.
And I wish dial-up were more obvious to the Ubuntu packagers. Recent versions of Network Manager won't recognize Gnome PPP as a network connection and Evolution refuses to fetch mail, thinking it is offline. Arrrgh!