I hope their signatures are more up to date than their ads.
Ich weiß von nix, sicher hast du das schon alles im Auge?!
- Die einen zerstören aus religiösen Gründen alter Kulturgegenstände.
- Die anderen sichern aus wirtschaftlichen Gründen das Ölministerium vor/statt den Museen.
Ja, ich sehe einen graduellen Unterschied.
THE TEAM | ZDF http://the-team.eu/social/index.php?consent=true
Cool article, though I only skimmed the details by now. It's been quite a while I heard something from the "formal proof of software correctness" faction ( )
- Author was frustrated that their fix hasn't been adopted by Java community, but unsafely been fixed by increasing a certain memory allocation. Unfortunately it is not apparent to me, if this "other way to fix", is correct and if it might be a legal memory vs. CPU trade, as the proposed fix complicates a certain condition to be check quite a few times.
- The (for me) striking argument for practical relevance of ongoing research of sort algorithms was by R. Sedgewick. Heonce mentioned, that if both memory and CPU follow Moores Law, then the operation "sort a whole memory full of data" will take longer and longer due to the overproportional growth of sorting time.
But from the API perspective caller is not allocating anything: Given an array a, the highlevel call is as simple as Arrays.sort(a). (I agree this is surprisingly terse for a Java API. )
So as I understood, the authors are really worried about the (repeated) waste of internal memory. However this memory is small, a few hundret bytes per call and, unless you are short of memory, this allocation merely translates into CPU time in form of garbage collection effort. This has to be compared to the effort in term of additional checks during sort.
What is more efficient might very well depend on the task actually done, i.e. lots of times sorting of comparably small arrays is on one end and rare sorting of huge arrays on the other. But such decisons when providing a service library are always based on lack of information about later use case.
- Friedrich-Alexander-University, Erlangen-NurembergMathematics
Ingress Troll :-)=
- Schaeffler Grouppresent
Zoomeffekte: iOS 7 macht Anwender seekrank - Golem.de
IPhone- und iPad-Benutzer, die auf Apples neues mobiles Betriebssystem iOS 7 umgestiegen sind, berichten von Symptomen der Reisekrankheit. S
Dilbert comic strip for 09/26/2013 from the official Dilbert comic strip...
The Official Dilbert Website featuring Scott Adams Dilbert strips, animation, mashups and more starring Dilbert, Dogbert, Wally, The Pointy