Good enough means good enough. It's not 50%. It's not 80%. It's whatever is the powers that be decide is the appropriate trade-off between competing interests (i.e. time to market, cost of deployment, $ paid to developers, estimates of the cost of future bug fixes, ...). If the market doesn't accept the product, it only means that their definition of "good enough" was wrong, not that there isn't a different "good enough" that the market will embrace.

"Good enough" for the software that runs a nuclear reactor or the space shuttle is probably near 100%. Pace-makers, surgeries, and pharmaceuticals failing? It's been known to happen. Sure, maybe only in 1/10000 cases, and only if you don't have a pre-existing condition. So, good enough is maybe five nines there. "Good enough" for that $1 umbrella you bought at Wal-Mart probably means something entirely different.

I don't know how I used to get by before I knew find had an '-exec' option. I voted for "grep", but I hardly ever use it without find, or find without grep.

`find XXX -exec grep -Hni whatever {} \;`

Windows is a plenty good enough platform to deliver e-Books on. And Windows runs enough trendy OSS programming languages on (eg. python), that the kiddos could've gotten a kick out of it plus a crash course in programming. Also, there's a lot of educational software out there that only runs on Windows that these governments could potentially leverage on day 1.

Microsoft would've had plenty of incentive to keep it up to date, plus keep the price low. Besides, OPLC had potential customers who insisted that the device run Windows. What's Negroponte going to say - "no, your *country's* children can't have inexpensive educational tools because I insist that they run Linux!"? If the goal is to get educational materials into children's hands, sometimes compromises must be made.

I say this as someone who has a substantial amount of code on the XO - the XO failed largely on its own merits. The project was (in-large) done in a fishbowl and came in at 2x its promised price. Plus, we've seen time and again that top-down approaches toward helping poverty-stricken areas (unfortunately) seldom work.

Their working with MS amounted to, what - adding a $2 MD card reader to the XO? I think that's all that it changed in terms of the hardware plan, anyway.

I have an XO. The card reader is bloody useful, if you ask me.

even iCab -a browser developed by one person- beat them to full compliance by months

That's disingenuous. The version of iCab (4.6) that passed the Acid3 test uses the same WebKit rendering engine as Safari and Chrome. And it beat Safari 4 to market by 1 day.

Not so much. You probably don't find Qt to be too inefficient, but Arthur does double buffering by default too. And for good reason.

Personally, I blame GTK2's obsession with double buffering everything.

Clearly, the Mozilla developers just forgot to call 'gtk_widget_set_double_buffered(false);'. That's what's been gumming up the works. Much appreciated, and thanks for your informed opinion on the matter!

My laptop: Fry
Wife's: Leela
Wife's old laptop: Amy
Printer: Zoidberg (dispenses ink)
Router: Bender ("bends" packets)
OLPC XO Laptop: Kiff (both small and green)
Car: Planet Express Ship (with which the 2006 Honda Civic shares a striking resemblance)
Cat: Zapp (cavalier, not too bright, doesn't wear pants).

I've been told by wifie that future pet names will include "Nibbler" and "Scruffy".

They updated the version of WebKit that they're using to one that passes the ACID3 test. That's something.