Comments by "xybersurfer" (@xybersurfer) on "Computerphile" channel.

  1. Cooper Gates now give an example where that code would be required.
    1
  2. Cooper Gates interesting case. but i can't help but think that you could have used a (possibly rewritten) if-statement where you use "<" instead of "==". it would be more descriptive
    1
  3. Cooper Gates to me the second if-statement would be more descriptive. it shows what could happen after the cast: if (k < 0) return -1 * (int(-1.0 * k)); else if (float(int(k)) < k) return int(k) + 1; else int(k); i assume that the values of the float k are not too big for an integer
    1
  4. Cooper Gates i consider "!=" just as dangerous as "==" for comparing floats. i think "<" comparisons are usually the proper way of thinking here. in the real world you can't always expect numbers to have an exact value, or be able to catch the step at which a number has that value. the "<" comparison captures the thought that you are working with a range of possible values.
    1
  5. Cooper Gates to me that just shows that there is less margin for wrongly interpreting the comparison when reading the code.
    1
  6. even though i'm familiar with the stuff. i completely agree
    1
  7. ***** as a programmer i actually agree with you, on a philosophical level. binary is rather far removed from our intentions. but as you can see from the responses: people have a tendency to obsess over unnecessary details.
    1
  8. matsv201 that would be one way of looking at it. those components could be both binary and base 256 at the same time, depending on the level of abstraction you choose. just like a usbstick with 1GB could be seen as one big cluster having base 2^30.
    1
  9. matsv201 i don't disagree with you. if you choose to view things at the hardware level then that is what you see. my point is that it is not the only possible view. it's a bit unnecessary to argue about which view is better. i think it depends on the situation.
    1
  10. lulhoofd1 i don't deny that people can get used to binary. it holds no secrets to me either. but that doesn't mean it's the closest we can get to our intentions.
    1
  11. matsv201 you have shifted from addressibility to the inner workings of subsystems. yes when you go deep enough into IC's, you will likely, find only binary. the lines themselves are not binary though. a line can have more than 2 possible voltages. voltages exist on more of a continuous range. and those lines are made of atoms... i can shift the view too. but what's the point of all these details? what exactly is your philosophical point?
    1
  12. matsv201 you just don't seem to get the points i'm making. you have a hard time letting go of the details. you mentioned how systems are binary values. i mentioned that this is true but that this is just a single point of view. you changed your view to to a deeper lever where one only sees binary and i in turn showed you a deeper view where electric wires can be regarded as components, which are not limited to passing on only 2 voltage values but a continuous range to show you how pointless it is to senselessly pick a single view. you subtly introduce the assumption that the bit addressability equates to the "base". even though this is not necessarily the case.
    1
  13. yes, i also prefer the idea of accommodating the hardware to fit the conceptual ideas
    1
  14. i don't think the problem is whether it is persistent memory or not. because one could just leaving the machine running and it wouldn't make a difference. the problem is that memory could refer to state or RAM, combined with the fact that he refers to a computer with state as having no memory
    1
  15. people can make errors in thinking too. the proofs in Coq are readable to humans. but i do agree that they don't give much insight
    1
  16. virtual machines are not necessarily completely isolated from the host. there have been exploits
    1
  17. i disagree. statistical methods are not that interesting
    1
  18. Dylan Cannisi i like this video. maybe i wouldn't mind seeing 1 or 2 videos about statistical methods. but i consider what he is doing right here, more related to Theoretical Computer Science. of course depending on how it presented i could become more interested (i'm picky)
    1
  19. if there are multiple threads then there is thread switching, and if your thread was not the one running then it could have missed the signal. it's not so much about doing multiple things at the same time. it's more about the guarantee of not missing a signal. also by introducing multiple threads it becomes very hard to give guarantees about the time it takes the system to respond (a.k.a. real time). which is a lot of times important for embedded devices
    1
  20. Anders Jackson i know that. and the thought didn't escape me while i was typing it. but as ***** said. it would be silly to introduce a core at least every system that will be interacted with. it doesn't scale
    1
  21. cool. are these guys related to the Computer History Museum channel?
    1
  22. Lukas West oh i see. nice they, have a channel too. thanks
    1
  23. i don't think it would have helped. literally implementing the diagram would produce incomprehensible code, that would require this diagram to understand
    1
  24. i've learned in the past that it's better to not use "==" when floating point numbers are involved. it's better to use '<' and '<='... and also watch out for "-0.0" which is not the same as "0.0". MS Excel is also not free of floating point rounding errors
    1
  25. it's funny that you say that because London seems to have a lot of those openings compared to other parts of the world
    1
  26. +Strofi Kornego maybe it has changed, but up to about a year ago there were a lot of vacancies of functional programmers in London
    1
  27. Unpronouncable okay... let's assume that the definition of the problem didn't mention that there was only 1 barber and that you solved the problem for 2 barbers. i agree that we have to go with whatever the problem definition is. however, everything that is not in the definition makes the problem more broad (harder). you cannot pick a solvable case of a problem (and solve it), and then say that you solved the problem. you have to solve the whole problem definitively for all cases. (which here includes the case of 1 barber in town)
    1
  28. i don't see why the layout has to be repeated every time the book is read. it's not like the book is going to change
    1
  29. i think the professor picked an excellent example, but the explanation could have been better. he didn't have to mention the stack. he could have focussed on substituting the initial call into the resulting return statements.
    1
  30. i could not find anything about a killer door in germany
    1
  31. +Alexander Roderick nice try, but the problem as stated by +Noel Goetowski is not that they are competing
    1
  32. Nixitur i'm not sure i agree. the question is now: - did you prove that H is false? - or rather that the implication formula H → H+ you proposed is false? the statement ¬H+ → ¬H makes it very clear, that there will be disagreement if someone does not agree the original implication formula.
    1
  33. but H → H+ also holds when: H does not exist and H+ does not exist. according to the H → H+ = ¬H ∨ H+
    1
  34. but, i am saying that the implication is wrong. because your formula still holds when H does not exist and H+ does not. your formula holds as soon as H does not exist. there is not even any point in checking whether H+ exists
    1
  35. FernieCanto yes, as said. take: - H does not exist (H = false) - H+ exist (H+ = true) clearly this should not possible because H+ depends on H but when you fill them into the formula H → H+ = ¬H ∨ H+ = ¬false ∨ true = true ∨ true = true it says it's true
    1
  36. FernieCanto yeah it seems that there are some things i don't understand yet. i've struggled a bit with understanding the halting problem. as a student i have reproduced the proof but i never really felt like i understood it. i think i will re-watch the video a couple of times.
    1
  37. Nixitur i don't get how you can make such a strong claim
    1
  38. what is the chance of a cosmic ray flipping a memory bit?
    1
  39. poon5000 that's interesting. but it looks like they are using a more loose definition 2d sheet.
    1
  40. Haniff Din maybe this also depends on whether you think space is fundamentally analog or discrete. what's stopping you from suggesting it's 1d?
    1
  41. Richard Smith i also don't like theories that play semantic games. i'm undecided about whether digital computers can be sentient like a brain, i'm leaning more toward yes. and it seems like you are too if possibly a few changes are made. to me that sounds more like an optimization issue. it would seem that it is still a turing complete machine. i lean towards making things functionally equivalent. of course we may never know when they are functionally the same, and we might want to leave-out/improve some things
    1
  42. Haniff Din i also prefer the simplest representation. but i don't think we are talking about whether something is re-presentable as something else. there is a distinction between 2D and 3D before changing their representation. otherwise there would be no need to talk about changing representations at all. this seems like a completely different kind of discussion.
    1
  43. The Sapien then what is making current computers functionally identical to us, according to you?
    1
  44. Richard Smith to combine some of your ideas; i don't see a problem with clever programming if it is so clever that it can process the world in the same fundamental ways as humans
    1
  45. The Sapien i don't know any of course. i thought you had specific changes to current computer hardware in mind (not the software side).
    1
  46. what a dream job
    1
  47. great video by the way
    1
  48. ​ FichDich InDemArsch InvalidCastException comes to mind (you can write code that casts an object to anything). NullReferenceException also comes to mind, although that is more of a design flaw where strings and objects are always nullable, that i last heard they would be fixing. it's not as bad as dynamic languages like JavaScript though. i use .NET languages like C# regularly
    1
  49.  FichDich InDemArsch  JavaScript's aggressive type coercion is also well defined, but that doesn't make it type safe. i agree that an exception at runtime is better, than none at all (i absolutely hate dealing with silent errors. in fact i'm dealing with one right now!). but, i would prefer even more to be prevented from that mistake before runtime, if it's practical
    1
  50.  FichDich InDemArsch  language definition and type safety, are not the same concept
    1