data oriented design .. from cppcon
741 views
Skip to first unread message
ymo
Mar 18, 2015, 3:30:50 PM3/18/15
to mechanica...@googlegroups.com
Not sure if you have seen this but most of what he is saying applies to java just as well !
https://www.youtube.com/watch?v=rX0ItVEVjHc
https://www.youtube.com/watch?v=rX0ItVEVjHc
Martin Thompson
Mar 28, 2015, 2:43:48 PM3/28/15
to mechanica...@googlegroups.com
Love the point that the compiler can only solve problems in the 1-10% problem space. The 90% problem space is our data access which is all about data structures and algorithms. The summary is he shows how instruction processing can be dwarfed by cache misses. This resonates for me with what I've seen in the field with customers in the high-performance space. Obvious caveat is applications where time is dominated by IO.
Follow this with, "...any one who says premature optimisation right now can leave the room - that is the most abused quote of all time."
As he begins closing he makes the comment - "the bad thing is we have a culture that thinks it is a good thing to hide the real problem, we pile more and more stuff to hide how memory is accessed, how memory is arranged, to hide the most significant problem we have to deal with..."
The final quote he gives to Christer Ericson:
"Design patterns are spoonfed material for brainless programmers incapable of independent thought, who will be resolved to producing code as mediocre as the design patterns they use to create it."
On 18 March 2015 at 19:30, ymo <ymol...@gmail.com> wrote:
Not sure if you have seen this but most of what he is saying applies to java just as well !
https://www.youtube.com/watch?v=rX0ItVEVjHc
--
You received this message because you are subscribed to the Google Groups "mechanical-sympathy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mechanical-symp...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Rajiv Kurian
Mar 28, 2015, 6:11:53 PM3/28/15
to mechanica...@googlegroups.com
MIke Acton is always entertaining. Another awesome quote by him: “Reality is not a hack you're forced to deal with to solve your abstract, theoretical problem. Reality is the actual problem.”
I can only imagine what kind of vitriol he would have for typical Java code after his C++ rant.
Jan van Oort
Mar 28, 2015, 6:18:24 PM3/28/15
to mechanical-sympathy
Totally second the last thought. I teach software design and UML to audiences of "engineers" with major companies in Europe, the ones that produce medical, automotive etc. etc. software your life depends upon everyday without you or I giving it a second thought.
You wouldn't believe how many of these "engineers" try to fit their software to existing design patterns, instead of doing it the other way round. Recently, in the UK, I was teaching with a major Producer of Medical Contraptions. An "engineer", after four days of teaching, showed me how proud he was to have his design fit the ( Abstract ) Bridge pattern. He did not *at all* have the problem the original design pattern was thought to solve: maximally decouple a concept from its implementation. He had, however, retrofitted his design to make it "fit" to the ( Abstract ) Bridge pattern....
This is code you and I are at risk to be directly exposed to when getting an MRI scan.
Wow.
0 new messages