Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
cpgxiii
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
30 ms
·
331.
▲
by
cpgxiii
6y ago
The problem with OpenCL, just like OpenGL, is that it relies on vendor support for essentially the full stack. That means if a single vendor doesn't support the feature, it's basically impossible to ship any code that uses that
332.
▲
by
cpgxiii
6y ago
Note that the article covers the 747-400, which first flew in 1988. Using 3.5" floppies for updates made sense then, and aircraft avionics tend not to be updated unless absolutely necessary.
333.
▲
by
cpgxiii
6y ago
There are a number of interesting what ifs to consider. Generally, scale back the size of the heavy bomber force and spend the resources on other things Absolutely, at least for the RAF. Coastal Command was denied long-range bombers for f
334.
▲
by
cpgxiii
6y ago
There were plenty of other reasons to attack from below - interest in the tactic predates the war - but I don't think the tactic would have been anywhere near as effective if RAF heavy bombers had been equipped with belly turrets (and
335.
▲
by
cpgxiii
6y ago
interviews with aircrew who had been shot down and (miraculously) survived and made it back to England revealed that most never knew what hit them The particularly depressing part of this story is that for months RAF crews were reporting
336.
▲
by
cpgxiii
6y ago
It's easy to look back with the benefit of hindsight to say that doctrine needed to change faster, but the realities of the war made that a challenge. The lessons of 1940 should have been "build long-range escort fighters ASAP!&
337.
▲
by
cpgxiii
6y ago
US had such fundamental problems with bomber design that you could say they were of doctrine in nature The B-17 "Flying Fortress" had lots of guns and a crew of 10, but couldn't defend itself against German fighters These