Slashdot Mirror


User: HuguesT

HuguesT's activity in the archive.

Stories
0
Comments
3,087
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 3,087

  1. Re:I was in the same situation once on No RIF'd Employees Need Apply For Microsoft External Staff Jobs For 6 Months · · Score: 1

    So then what happened to your dad?

  2. Re:I was in the same situation once on No RIF'd Employees Need Apply For Microsoft External Staff Jobs For 6 Months · · Score: 1

    Won't work, because most employees would rather use the cash than save it. Then society would have a massive influx of destitute, retirement-age people, which *would be* a problem. It's been done in the past with catastrophic results.

  3. Re:I disagree on Math, Programming, and Language Learning · · Score: 1

    Your reply is very interesting and outlines the fact that one should not stop learning after school, but instead revisit past and new subjects with a different eye and different tools. Also motivation is the prime mover here.

  4. No math compiler on Math, Programming, and Language Learning · · Score: 1

    Computer programming can be seen as more rigorous than mathematics because if the written program is not correct, the executable will not run; whereas a mathematical proof may contain elements that are not completely described but part of mathematical lore. However we do not possess a compiler for mathematics. Conversely language may be more abstract than mathematics because language, in addition to mathematics, may express information that is not mathematics, e.g. poetry, imagerie, etc. However mathematical abstraction is also very rigorous, which is not the case of poetry or other literary constructs.

    Mathematics is unique in requiring both a high level of abstraction and rigour at the same time, yet this must be performed without any artificial help like a debugger or compilers. In addition, creative mathematics require a high level of intuition and the capacity to concentrate on a specific problem for long periods of time (months, sometimes). Altogether, mathematics requires specific talents that are fairly rare and not necessarily found in programmers or writers.

    Fortunately we are not all alike.

  5. Side effect of grant structure on Elite Group of Researchers Rule Scientific Publishing · · Score: 3, Interesting

    Grant money is given preferably to teams that already publish a lot. Even "starting grants" in the EU require a single principal investigator (PI) with a lot of well-cited publication under their belt. This can only be achieved if the PI has done their initial research in a well-heeled lab, with a well-known head of the lab who is well-connected, and so on. This encourages a pyramidal structure with a lot of grunt students at the bottom, supervised by post-docs, supervised by assistant professors, and so on. Success encourages visibility, which encourages grants, which ensures money, which ensures good grunt students can be hired, and so on.

    This is not the only possible successful structure, but one of the most common. A single researcher, however brilliant, cannot usually keep up with the outpouring of landmark papers the pyramidal structure can achieve. On the other hand, if everybody does their job, meritocracy in the pyramidal structure ensures that the best grunt students get promoted to post docs, and so on, usually in a different pyramidal structure.

    The big drawback of the pyramidal structure is that the prof at the top usually doesn't know exactly what is going on at the bottom, even though they put their name on most of the papers that the structure produces.

    Disclaimer: I'm a tenured prof. I do have a reasonable number of students, but I work with them directly. All my students are co-supervised with at least one other prof. Occasionally I do have a few post-docs but the structure is always collaborative. This is not the standard but this works well enough also as long as there isn't any ego-driven fights in the lab. This means choosing your collaborators well. I've made a few mistakes, but so far so good.

  6. Re:Well, of course on NSA Considers Linux Journal Readers, Tor (And Linux?) Users "Extremists" · · Score: 1

    If you look at just about anybody's success story, the first thing that is of utmost importance is being in the right place at the right time. In other words, luck. The American dream has always been a dream. I'm not convinced that anything much has changed in the last 70 years about this, i.e. since about the end of WWII. Sure hard work is a factor but by no means the only one.

  7. Not all libraries. OpenCV for instance.

  8. Re:Python is better overall but R is more like SAS on Ask Slashdot: Switching From SAS To Python Or R For Data Analysis and Modeling? · · Score: 1

    Inefficient (two interpreters), inelegant (two syntaxes), and there is usually no point to it. Both languages are roughly as capable.

  9. Re:I always knew on Ikea Sends IkeaHackers Blog a C&D Order · · Score: 1

    Only the really cheap stuff is particle board. They also have plenty of real wood stuff. What they really do well is the non-particle, real wood, cheap shelves like the Ivar brand. These are everywhere and there is essentially no alternative.

  10. Re:In civilized countries... on Starbucks Offers Workers 2 Years of Free College · · Score: 1

    Only of those Europeans that have rich families. The European universities are full of European students too (and very few North Americans), usually the not-so-rich kind.

  11. Re:In civilized countries... on Starbucks Offers Workers 2 Years of Free College · · Score: 1

    Most prestigious, most awash with money, yes. What befuddles me is why these super-rich universities don't simply select the very best students all over the world (including the US), and don't offer them affordable tuition. They would be even better. As of now, most US universities simply perpetuate a rich class divide.

  12. Re:In civilized countries... on Starbucks Offers Workers 2 Years of Free College · · Score: 1

    The US to the rescue! the dream of all countries mired in anti-democratic squalor. If you look around the list of recently US-"liberated" countries, even as far back as the 1950s, it could perhaps bring you back to reality.Getting the US attention is more a curse than a blessing.

  13. Re:Legacy file systems should be illegal on One Developer's Experience With Real Life Bitrot Under HFS+ · · Score: 1

    I'm sorry, citation needed. It does have compression and encryption, and you *can* make a RAID 0 or 1 with it, so it has some (weak) redundancy support, but I have never heard about it having deduplication. ZFS is the only filesystem I know that has deduplication.

  14. Re:Is it a Complete Set? on Lego To Produce Three Box Sets Featuring Female Scientists · · Score: 1

    As long as the US remains attractive for immigrants, you are correct, this is a life choice. However realize that population in any given country needs to be replaced. Having (too) many kids in Uganda do not compensate for the (dramatic) lack of kids in Japan. It's not a simple matter of shifting kids around, which is never simple to begin with.

    So in short you may choose not to have children, but somebody will have to pay your pension eventually. It can help if this is someone you now well.

  15. Re:Ellsberg got a fair trial on Daniel Ellsberg: Snowden Would Not Get a Fair Trial – and Kerry Is Wrong · · Score: 1

    It is because it is a more complex issue that first thought.

  16. Re:R & D in America on MIT Used Lobbying, Influence To Restore Nuclear Fusion Dream · · Score: 1

    Quadtree are an approximation technique widely used in imaging and computational geometry. Did you look on Google Scholar/Web of Science or just in patents?

    A light search returned these links:

    http://link.springer.com/chapter/10.1007/3-540-47789-6_106 (sorry paywalled)

    https://diglib.eg.org/EG/DL/Conf/EG2002/short/short90.pdf

    I'm not sure if this is what you are looking for. In discrete geometry (construction of a Voronoi tessellation on pixel data), it is often more efficient to used an Euclidean distance function, which is linear. Indeed constructing the quadtree plus using it for the computation takes more time.

  17. Re:MIT sure has fallen far on MIT Used Lobbying, Influence To Restore Nuclear Fusion Dream · · Score: 1

    Is the project poorly managed now? Do you have inside information?

  18. Re:non-locality or GTFO on Electrical Control of Nuclear Spin Qubits: Important Step For Quantum Computing · · Score: 1

    No, we really do have working quantum computers in the lab right now. They are fully capable of running Shor's algorithm. It's just that they can factor at most a number like 15.

  19. Uber is a multinational skirting regulations on Virginia DMV Cracks Down On Uber, Lyft · · Score: 1

    In many cities around the world, taxis are heavily regulated. Among these regulations are a fixed number of license plates, and the costs of these plates (or equivalent medallions, etc). This means that in many instances there aren't enough taxis to go around because these numbers were fixed a long time ago and may not be have been updated to meet demand. This benefits most the taxi operators and to some extent the drivers themselves because a high demand drives the price of the fare up. Also a business with low competition is always more comfortable to run. Customers hate it but are used to this situation.

    Now Uber and others have sought to change the game, first by ignoring regulation and getting self-employed people to drive their own car to ferry people around. This is very good to some extent because taxi business in a lot of places is over-regulated and does not meet demand. Also the Uber et al have a nice online presence and at this stage at least do provide a useful service, so why not.

    However, Uber fares are not cheap, this is not "sharing", this is a business. The self-employed individuals driving the cars may be putting themselves at risk: with their rides, the regulatory authorities, in case of accident, with other regulated taxi drivers, etc. We are still in a "honeymoon" period but this is sure to end. Uber has become much to big to be ignored, and so will soon have to fight for its own existence, in a lot of places all at once. I'm not sure their (huge) valuation will be enough.

  20. Re:Since when does Qt "work" with OS X? on Apple Announces New Programming Language Called Swift · · Score: 1

    Dear Troll,

    You are wrong and you know it, but please continue.

  21. Re:Ellsberg got a fair trial on Daniel Ellsberg: Snowden Would Not Get a Fair Trial – and Kerry Is Wrong · · Score: 1

    Exactly. Thank you. Right now Snowden is innocent under the law.

  22. Re:Ellsberg got a fair trial on Daniel Ellsberg: Snowden Would Not Get a Fair Trial – and Kerry Is Wrong · · Score: 1

    Actually there would be a trial, where you, the murderer shooter, would be first assumed innocent of any crime. Then facts would be brought up to ascertain whether or not you were allowed to invoke self defense when you shot that murderer.

    In Snowden's case, this is completely hopeless. Everybody has their media-hammered opinion that he is guilty. And since he ran away, everybody is convinced he is a coward too and does not even deserve a second glance. This opinion is everywhere in this discussion.

  23. Re: Ellsberg got a fair trial on Daniel Ellsberg: Snowden Would Not Get a Fair Trial – and Kerry Is Wrong · · Score: 1

    Nothing is illegal for the NSA then. The law is a human construction, it is deeply flawed, not a hallowed text. In this case there should be some checks and balances over what the NSA is allowed to do and how they go about doing it. There isn't, and Snowden was right to expose that fact. One can express some pride at the extent and depth at which the NSA has been able to spy on foreign people (including friends and allies), one can also be frightened.

  24. Re:Ellsberg got a fair trial on Daniel Ellsberg: Snowden Would Not Get a Fair Trial – and Kerry Is Wrong · · Score: 5, Insightful

    Jury nullification almost never happen, and in this case the jury would be carefully selected to completely exclude this remote possibility. Like it or now, your attitude summarizes a lot of what we hear about Snowden. Almost no one on the radio, newspaper, etc says anything about the crimes that Snowden has uncovered, and that will go unpunished, because, you know, everything is about Snowden.

    The best option for Snowden is to hide and wait for the American people to realize he was right all along, and a true hero like Ellsberg, and certainly not submit himself to some kangaroo court. If that never happens, the American people deserve their fate.

    Almost no one like heroes that disturb the comfortable status quo.

  25. Re:Ellsberg got a fair trial on Daniel Ellsberg: Snowden Would Not Get a Fair Trial – and Kerry Is Wrong · · Score: 1

    Let's face it though. He is guilty

    That's it. Enough said, thank you.