« first day (1176 days earlier)      last day (3298 days later) » 

12:29 AM
@MichaelE2 Well, well.. you should have known :-)
 
12:48 AM
@halirutan Yes, I suppose so. My practically -- oh what's that word for a memory like a photograph? photogenic? no, that's not it -- anyway, my infallible memory surely recalls the substantial content of every question whose format I improve. ;)
 
@MichaelE2 Hehe :-))
 
1:41 AM
@MichaelE2 That's the first I've heard that joke. I like it. The fancy word I hear these days is eidetic instead of photographic.
So I'm enjoying a recent game called Ori and the Blind Forest, but I'm afraid to play more than an hour every few days because good games aren't being produced fast enough. I'm enjoying sifting through Wikipedia categories though.
@halirutan I looked at a lot of half-timbered buildings today. Fachwerkhäuser?
Pretty charming.
 
2:25 AM
I really think computer algebra and such programs are making people stop thinking and to stop using their brain. I mean programs such as Mathematica and Matlab and such. I see questions being asked which make no sense at all, which clearly means the person does not have any idea what they are doing, and just want answer to something. Before computers, people had to think. That is why I think people 100 years ago and more were smarter than we are today
 
2:43 AM
@Nasser It's a broader phenomenon than computer algebra systems. Most people do everything they can to avoid programming because it requires unnatural attention to detail. They like to think a small number of programmers will be able to be "managed" into being 10x more productive, instead of just admitting that we need 10x more programmers. My last manager told me he was too stupid to program and quit instead of helping out by opening up a code editor again.
Everything became much simpler for me when I just accepted that thinking is computation. Some people put that realization off for a long time though.
 
3:41 AM
@MichaelHale Eidetic sounds familiar. I also know the term as it relates to Plato's theory of forms (eidos = form).
 
 
7 hours later…
10:28 AM
@MichaelHale Yes, this is the correct term.
 
 
5 hours later…
3:22 PM
@MichaelHale Just in case: You are aware that chat is being indexed by google, are you?
 
@SjoerdC.deVries That's fine. His boss quit after that, so now they are merging us with another group.
The first one's exact words to me were, "I used to be much smarter and happier when I was young and single like you. I have a short circuit in my brain now."
 
3:36 PM
They give me a lot of free time at the bank though, so I just keep a low profile and do what I'm asked.
 
3:55 PM
@MichaelHale IIRC the CEO for Microsoft Sweden said the opposite in a lecture here, he said that too many wanted to be programmers but that they needed to convince some programmers to choose a management path instead.
He himself had taken that route I think.
 
@MichaelHale I think this phenomenon is sometimes/often connected with what is known as Peter Principle. Everyone tries very hard to become richer and higher in hierarchy and end up doing things where they are not really good in.
 
@Pickett That's interesting. All of the managers I had at Microsoft were very good. I still chat regularly with one of them. Microsoft and other tech companies tend to have more programmers and fewer managers than 150 year old banks though.
 
@MichaelHale @Pickett The question is whether the managers really were excellent programmers in the first place. Being a good manager is pretty hard, because you need to have the knowledge and experience to decide when to trust the decision of people that you manage
 
Heh. Looked that as an anumation only after uploading it here. :)
 
@MichaelHale Right, forgot that you worked at Microsoft! Well, that's what he said, but of course the lecture was partly as a promotion towards graduates so the truth may have been "adapted" to the audience.
 
4:02 PM
@halirutan Very interesting. I guess that implies that if someone hasn't been promoted in a while they should be diverted to a different branch. The opposing force of course is that there are typically fewer positions at the top of hierarchical organizations, so traffic jams are perhaps inevitable.
 
it's random dots on Hilbert curve (fractional part of value moving along a straight line), but modified so that all dots have a minimum distance of 1 to neighbor, and rest having a(d-1)+1 distance, and all dots fitted to scale a so that all dots fit in whole Hilbert curve.
 
@MichaelHale I guess it takes courage to stand up and say that you are excellent at what you're doing and that you don't want to be promoted :-)
 
That's actually pretty annoying animated GIF.
 
@kirma It's more interesting now that I know what it is. So gradually filling up the Hilbert curve but weighted toward one end.
@halirutan I imagine in the future that any long-standing CEOs will have to be pretty good at utilizing computation or they will just get out-computed by a group of subordinates who then appear that they can make better decisions than the CEO.
 
@MichaelHale Subordinates who don't share their reports with the CEO won't last very long though?
 
4:16 PM
@MichaelHale In my eyes, in many situations the subordinates will always make better decisions, because they are closer to the problem/solution/software/framework. The managers need to be people that 1. have a very wide view of everything and they need to be superb in knowing when they should listen to their best people and when not. The worst thing you can have is a egocentric director that needs to make every last decision by himself.
Unfortunately, I have seen many of these bad examples.
 
@MichaelHale Those dots are picked from a normal distribution with rather arbitrary mean.
Happens to be slightly slanted in the regard of symmetry of the curve.
 
@halirutan I guess if someone runs into a ceiling caused by lack of positions above and not hitting the limit of their potential, that is something that causes them to go to another organization.
 
This is actually a quick hack for consideration of visualization for a friend, who wants to express whole IPv4 address space in a way that groups neighboring IP addresses from limited number of displayed addresses close to each other, but still at a separable distance for visualization.
 
@Pickett Well if there are 10 people in a meeting, one is the CEO, and two seem to always have really good ideas because they run lots of automated searches through various ideas at night, then I'd guess eventually the other 7 would want one of them to be CEO.
Maybe the incentives at that level are all really good so people don't do things like that.
@halirutan In my current group I feel the best strategy is to speak as little as possible even if I have an idea that might help. If I speak I will only create more work for myself. If I wait for the people above me to eventually come to the same conclusion based on lots of small reports they ask me to do over a long time, then I have lots of more free time and I end up getting the same promotion or raise without causing them extra stress from throwing extra ideas on them.
@kirma Ah, interesting. So using Hilbert curve to display 1D data in a 2D space while still preserving a distance relationship.
 
4:33 PM
@MichaelHale Something like that. It's not perfect, but I suspect it's better than "physics" models with repulsive particles and magical springs.
They always need lots of tuning.
In this case, erratic behaviours are pretty much predictable.
 
Right, this wouldn't require iteration.
 
4:47 PM
Or exhibit strange swings and stuff without damping...
 
 
1 hour later…
6:05 PM
The above for uniform distribution...
 
7:02 PM
posted on April 08, 2015 by Johan Rhodin

An important emerging standard has been rapidly adopted by industry: the Functional Mock-up Interface (FMI). It’s an independent standard allowing model exchange between different tools. We introduced FMI export with Version 4.0 of SystemModeler. Exporting your model as a Functional Mock-up Unit (FMU) serves many purposes. First and foremost, it can be used in other [...]

 
 
1 hour later…
8:06 PM
Anyone know the mma syntax for grabbing the upper/lower byte values in an integer?
 
8:23 PM
@dionys Do you really mean byte value?
 
@dionys As halirutan implied, there are many interpretations of that.
 
@dionys Anyway, I guess what you are looking for is IntegerDigits. With that you can easily turn a number into its binary digits and from this list you can take the any bit you like:
In[122]:= IntegerDigits[1234567, 2]

Out[122]= {1, 0, 0, 1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1}
 
thanks, that's what I was after.
 
:)
Then again, if you want actually bytes, you can use base 256 instead of 2.
 
8:47 PM
Thanks @kirma that's even better. Using BinaryRead and testing results ala: FromCharacterCode@Reverse@IntegerDigits[18755, 256].
 
@dionys OK :)
 
 
3 hours later…
11:42 PM
3 link robot arm demo is here, 12000.org/my_notes/mma_demos/robot_arm_RRR/index.htm did for class. It was fun. Next will do 6 links arm. Hope you like it. CDF and notebook are there
I wish M was better at 3D rendering. It still slows down with there are lots of 3D objects, even in V 10.1. I do not know how WRI handle 3D graphics, but it does not seem to be using GPU and other things that is supposed to speed rendering.
fyi, For fast 3D graphics, there is now new OPEN GL, just announced, that is supposed to be much faster than current openGL, here is the link (it is still new) en.wikipedia.org/wiki/Vulkan_(API) call Vulkan
 
@Nasser It doesn't display properly. Half of the interface is chopped off.
Anyone dare to guess how many months will go by until CDF Player 10.1 is realsed?
 

« first day (1176 days earlier)      last day (3298 days later) »