I'm guessing cloud cover. Which would tend to track with the underlying land... Just like the clouds over much of southern Europe around 0:21 and 7:03 in the video.
Not to mention there's no guarantee it's actually ADT and not somebody casing houses and installing fake or compromised equipment. Not like it takes a lot to fake plolo shirts and a folder.:-) If you wanted their services, you'd call and schedule so you could have more assurance of identity.
Shared memory parallel codes (OpenMP) could benefit, though. Many (originally single-threaded) or homemade scientific applications run in this space: get some parallelism for relatively little work (insert pragmas, be careful to be thread safe, and test test test), without all the extra work of redesigning those simulations for efficient message passing.
You certainly find problems where it is much better bang for the buck to throw an expensive processor and OpenMP ( O($10^2 to $10^3) ) at a problem than to throw specialized MPI development effort ( O($10^4 to $10^6) ) at that problem. Especially when the first step to any hybrid OpenMP-MPI code is to work on single compute node performance with OpenMP, before connecting nodes with MPI.
We taught a little of this in some of our engineering courses this Spring. Suppose gene A down-regulates gene B, and gene B down-regulates gene A. Then if (A,B) is your network state, (1,0) and (0,1) are stable states, and all intermediate states go back to one of these. This is a bistable toggle. It's a way to write a bit of data to a cell.
Now, add two more genes: A promotes P which blocks A. B promotes Q which blocks B. This turns the system into a biological oscillator. Now you have a system click with tics (A up and B down) and tocs (A down and B up). Fun stuff.
I took a brief look through the paper. Table 3, glia (rightmost columns) seems to sum up this study nicely. Control group had 817 mice, 3 malignant brain tumors. Highest dose had 409 mice, 3 with malignant brain tumors. Not a significant difference in this entire table at any dose in any sub-population, even at p = 0.05 levels.
Table 2 focused on schwannomas, and they had to dig deep to male mice at highest exposure (n = 207) to get a significantly significant (at p = 0.05) difference. We're talking 3 / 207 male mice with malignant schwannomas at highest exposure. The control males had no cases (n = 412), but we're really in the weeds here where a stochastic variation of +/- 1 mouse makes a huge difference in their tallies. No other significant difference in any other dose in any other sub-population in any other table in this paper.
Kaplan-Meier survival curves (Figure 3 g-h) look just about identical for all doses: we're not seeing a big difference in survival times at any doses. And there's no effort to estimate error bars for those curves. That's a hint about (lack of) replicates.
From what I can see, there was exactly one replicate for each group / arm (e.g., mice exposed to a specific dose). This is not good, because technical and biological variability can cause flukes and false differences. 1 technical replicate per arm: if a technician had a bad day or screwed up a protocol when the exposing the mice to the highest dose, your one measurement set could be off. 1 biological replicate per arm: a weird batch of mice, or a batch of sick mice, etc., could throw off your one measurement set for the arm. Most cell line experiments we've worked with have at least 3 technical and biological replicates, in very controlled culture conditions. You'd be amazed at the variability, even in "identical" cells.
Oh, and read the neat Nature story (summary) where the sex of the scientist performing the experiments on mice can cause statistically significant differences. Because the male and female scents in our clothing can actually induce stress hormone changes in mice. Experiments are sensitive. Replicates are a good thing.
I was confused on that, too, and neurobiology is pretty far from my regular work. Part of the problem is that there's a lot going on here: plasticity and differentiation (cell adaptations, transformations from one phenotype to another, differentiation, etc.), cell proliferation, tuning of connections, etc. It's a messy area with lots of new and sometimes contradictory data coming out.
From what I understand, there's a lot of plasticity in the brain as an adult: far more than was originally appreciated. I thought that we "knew" there were no new brain cells, and then we "knew" there were, and now we may "know" there aren't. That's the nature of evolving science, as others point out. And imprecise science communication--and imprecise English--doesn't help, either. Even biologists can get tripped up: we talk about tumor growth and cancer cell population growth, but we really primarily are talking about cell division there.
And there are all sorts of neat surprises. It was found that glial cells (a type of brain cell responsible for maintaining brain structure) can transdifferentiate into endothelial cells (which make blood vessels).
See this PNAS paper and this one. This has all sorts of implications for gliomas and other brain cancers. And God only knows what other surprises are waiting to be found.
I suppose that's another reason they looked at the "new neuronal cell" marker: to see whether other cells could become new neuronal cells by transdifferentiation or other plastic processes. Biology is weird.:-)
There's a big difference between growth of cells (they get bigger) and proliferation of cells (they divide to create new cells), and so it's important to be careful. (I've been working on modeling both cell growth and cell division for a good while, mostly in cancer and a little in tissue engineering and synthetic biology. e.g., here.)
I looked at the study. They stained for Ki-67, the gold standard immunohistochemical marker for cell division. Cells that are actively cycling--in late G1, S, G2, and M phase, and a smidgen of G1 phase after division because Ki67 protein doesn't instantaneously degrade--stain positive for this marker. In particular, it is a nuclear marker, so the stain is localized to the cell nucleus, and the stain is very definitive. It's one of the easiest immunostains to do image processing on, because you can do nuclear segmentation, then analyze the colors in the segmented nuclei to see if they stained positive or negative for Ki-67.
And that Ki-67 marker was virtually non-existent in the region of interest in all the samples above 13 years old. See Figure 2. This is *the* universal gold standard marker for cell division used across pathology and experimental biology. So yes, the study indeed found no proliferating cells in the GCL. And then they used this "young neuronal cell" marker (DCX+PSA-NCAM+ cells) to further confirm what they already saw in Ki67.
Also, the Nature link is the *summary* of the paper, and not the actual paper. It's pretty common for the big journals to ask for a non-involved scientist in the same field to write a summary and commentary when a potentially controversial or significant paper comes out. Here's the actual paper.
Just please make sure that the measure of activity is not just by commits, code changes, etc. Some projects are just "feature complete / done" and shouldn't be de-listed or downranked in searches.
Moreover, it's a tradeoff. If the battery loses capacity a little faster but charges in 18 minutes instead of 90, then it may not matter if you have to charge more often.
It may go straight to File 13, but I wrote in too.
Hello, I'm really disappointed that Disney is pulling its content from Netflix to force customers to purchase yet another subscription service. As more and more companies like you do this (e.g., CBS All Access), you are fragmenting the online market, and making it harder and costlier for customers to legally access content. We were willing to cut the cord from cable providers and buy access to 2-3 services like Netflix, Amazon, and Hulu to get access to most content at lower prices. But this fragmentation is quickly driving costs above traditional cable access. This is unsustainable. We don't plan to buy the extra subscription to Disney when it's off Netflix. Maybe we'll start watching more Dreamworks and Warner Bros. movies.
Its built-in linux-like commands don't appear to work as expected, I guess due to aliasing. "ls" is fine. But with any arguments, and I get stuff like this:
PS C:\mingw\bin> ls -l Get-ChildItem : Missing an argument for parameter 'LiteralPath'. Specify a parameter of type 'Sys again. At line:1 char:4 + ls -l + ~~
+ CategoryInfo : InvalidArgument: (:) [Get-ChildItem], ParameterBindingException
+ FullyQualifiedErrorId : MissingArgument,Microsoft.PowerShell.Commands.GetChildItemCommand
I've been happy enough with cmd + mingw/msys. That got me ls, make, and all the same goodies with the same syntax as linux. But if Windows PS is going to have "ls", it should work like "ls" or get out of the way. (Just like they evidently do in linux and macos powershell.)
Thanks for your reply. It's an interesting discussion, and indeed, I get a little fed up when even in academia, translational medicine morphs from meaning "translating theory into practice" to meaning "getting patents and making profitable startups." It's needed, but it can sometimes distort the field and culture when it becomes an ends and not a means.
I'm a little curious as to your definition of techie, because a lot of the discussion really boils down to how you define a techie.
I find my techie friends are more informed and intellectually curious than any other group. I'd put them at about equal to my academic friends in this area.
When you say this, this makes me think your Venn diagram for "technies" and "academics" has no intersection. But ask most any grad student, postdoc, or faculty in an engineering, CS, or applied math department (and increasingly many biology departments), and they'll likely regard themselves as techies.
It seems to me a reasonable definition that a techie is a technophile, particularly one who loves, uses, and improves technology in their daily work and hobbies. But if you restrict your label to Silicon Valley and tech startup types, you're lose most of the amateur techies, the open source people, and the citizen scientists.
Again, it's an interesting discussion, so thanks. (And great to meet you here on Slashdot!)
Most of us who entered science and academia did so to make the world a better place, and many of us are techies. You'd be amazed at home much coding and tech is required for pretty much every area of science today.
We're writing open source software to solve real problems in science and engineering. We're spending the last of our startups on open access for our papers because it's the right thing to do. We're contributing to open data repositories because sharing data makes all our work better. We're writing free content on blogs, code tutorials, and MOOCs for public outreach, because we view our roles as educators seriously.
Most people in academic endure years of low pay and job uncertainty as postdocs and entry-level faculty--and defer or postpone indefinitely having children and buying that starter home--rather than faster and better-paying paths in industry, IP law, and mathematical finance because we do want to make the world a better place, and we're actively working on it.
So, while I agree with your general feeling, take a look around, and you'll see more techies trying make a difference that you might have realized.
When we get the right to demand money for whatever we deem is our interlectual property, the next step is to demand that companies like google are not allowed to make their own decisions in regard to what they return as search results. We'll put that under something like "no discrimination" or something.
Sounds like Spain tried to do this almost verbatim:
The Spanish Newspapers Publishersâ(TM) Association (AEDE) is now asking that the Spanish government and EU competition authorities stop Google News from shutting down its operations in the country, âoeto protect the rights of citizens and businesses.â
The media lobby group announced that an end to Spanish Google News would represent âoenot just the closure of another service given its dominant market position,â identifying that the closure would âoeundoubtedly have a negative impact on citizens and Spanish businesses.â
I am not a legal expert but I believe their plan to produce a nuclear-armed spacecraft violates the Outer Space Treaty (to which Russia is a signatory) and specifically Article IV which says "States Parties to the Treaty undertake not to place in orbit around the earth any objects carrying nuclear weapons or any other kinds of weapons of mass destruction, install such weapons on celestial bodies, or station such weapons in outer space in any other manner." (which sounds like exactly what Russia wants to do)
Then again, with the way the Russian economy is these days, I dont think they have the funds to actually build or launch this thing so it wont matter...
Sounds like it's planned to only ascend when needed, so nothing stationed in space. (And in any event, not in orbit until the point where treaties are moot.)
I read it too, and I fail to see the breakthrough. There are plenty of pseudo random number generators, such as the Mersenne Twister, with very long periods, so just occassionally XORing even a poor quality random number into the feedback loop, is enough to make it completely unpredictable.
Mersenne Twister is pretty much the standard for simulating a uniform distribution in a lot of scientific computing. These depend not only upon unpredictability (useful for avoiding biases, and clearly important in the security realm), but also upon properties of the uniform distribution.
But when we test it out, we find it's still not as great as we'd like: look at a histogram of outputs, and you'll see that until you get really large numbers of function calls, the histogram isn't particularly uniform. (In particular, numbers near the bottom and top of the range don't get called quite as often.) This means that simulation properties that rely upon uniform distributions over both long and short time periods may be thrown off, and short- and mid-time simulation results may well stem from the MT rather than from the mathematical model. Moreover, low-probability events may have artificially smaller probabilities in the simulations (because of the non-uniformity of the distributions near the bottom and top ends of the range).
Over very short numbers of function calls (a few hundred to a few thousand), the outputs can even tend to cluster in a small neighborhood. So suppose that you are simulating a tissue with many cells, and calling MT as your uniform distribution to decide if the cells divide or stay dormant (each with independent probability p, so each cell divides if PRNG/max(PRNG) < p). The math says that for a uniform distribution, you don't need to worry about what order you evaluate your decision across all the cells. But if the PRNG outputs cluster over several sequential calls, then a neighborhood of cells may simultaneously divide if they are all evaluated close to one another sequentially. In analyzing the spatial behavior of such a simulation, you may draw incorrect conclusions in smaller spatial structures that, again, derive from non-uniformity of the PRNG, rather than problems with predictability. (And then you may accept/reject a particular hypothesis or mathematical model pre-maturely.)
So, there's definitely more to it than just unpredictability, depending upon where the code is being used.
Have you tried mingw-w64 / msys? The CLI tools integrate pretty well (make, ls, tail, head, etc.), and they have a package manager to grab these things.
I'm not in academia, but I've published a bunch of (mostly IT security) research to be freely read by the public under my own copyright or the copyright of a company that's hired me. My serious question is: what is to prevent individual researchers from just publishing what they have as a PDF or WordPress article on a random site on the Internet? (e.g. are there rules in their contract that says they can only publish through so-and-so service, who has the copyright of academic research, etc.)
In part, this is what preprint servers like arxiv and bioarxiv are for.
However, there are deeper-rooted, cultural issues at play here. Academics are rated on their job performance (for keeping your position, finding tenure-track positions, and later attaining tenure) based upon their peer-reviewed publications. Traditionally, this has meant going through the private, paywalled journals.Likewise, getting grants requires publications in peer-reviewed journals, rather than just posting online.
Now, posting in open access journals (like the PLOS family of journals, PeerJ, etc.) helps here, since at the least the access isn't paywalled. But now the academic / lab itself has to pay a much larger publication fee. (Often on the order of $1500 per article.) Moreover, many of said tenure review panels and grant review committees judge you not just on whether you've published, but where. Impact factor matters, and that again tends to steer people towards glammy, paywalled journals like New England Journal of Medicine (which just made a big kerfluffle about research parasites), Nature, Science, etc.)
So, there's a lot going on here. And even the scientists who want to just post preprints and move on are facing tremendous pressures.
How does that actually happen with a smart phone? I mean it's not like you're going to download ISOs to your phone and...
Mobile hotspot.
For when you don't want to pay the outrageous wifi charges at a hotel conference center, or Starbucks' connection is flaky, or Charter / Comcast / Timewarner crashes for the evening and you have work to do, or... etc.
A heat pipe in a smartphone! That must be one hot SOC!
Although some of these SOCs reported heat problems, the heat pipe here might have as much to do with the water-resistant enclosure (can't vent air past a heatsink as readily in w watertight case...)
Slashdot Mirror
You bet! Blew me away first time I saw it. Biology gets crazier the more you look. :-)
You can actually see mitochondria crawl inside cells.
Here's an example, imaged over about 10 minutes: https://twitter.com/MAG2ART/status/1087386722667761665.
Here's another gorgeous video I just found: https://www.youtube.com/watch?v=x5IxkI6tkn0
I'm guessing cloud cover. Which would tend to track with the underlying land ... Just like the clouds over much of southern Europe around 0:21 and 7:03 in the video.
Not to mention there's no guarantee it's actually ADT and not somebody casing houses and installing fake or compromised equipment. Not like it takes a lot to fake plolo shirts and a folder. :-) If you wanted their services, you'd call and schedule so you could have more assurance of identity.
Shared memory parallel codes (OpenMP) could benefit, though. Many (originally single-threaded) or homemade scientific applications run in this space: get some parallelism for relatively little work (insert pragmas, be careful to be thread safe, and test test test), without all the extra work of redesigning those simulations for efficient message passing.
You certainly find problems where it is much better bang for the buck to throw an expensive processor and OpenMP ( O($10^2 to $10^3) ) at a problem than to throw specialized MPI development effort ( O($10^4 to $10^6) ) at that problem. Especially when the first step to any hybrid OpenMP-MPI code is to work on single compute node performance with OpenMP, before connecting nodes with MPI.
Maybe they can call them General REpair Micro LocomotIve automatoNS. Everybody wants to see gremlins in the engines. :-)
We taught a little of this in some of our engineering courses this Spring. Suppose gene A down-regulates gene B, and gene B down-regulates gene A. Then if (A,B) is your network state, (1,0) and (0,1) are stable states, and all intermediate states go back to one of these. This is a bistable toggle. It's a way to write a bit of data to a cell.
Now, add two more genes: A promotes P which blocks A. B promotes Q which blocks B. This turns the system into a biological oscillator. Now you have a system click with tics (A up and B down) and tocs (A down and B up). Fun stuff.
You're exactly right.
I took a brief look through the paper. Table 3, glia (rightmost columns) seems to sum up this study nicely. Control group had 817 mice, 3 malignant brain tumors. Highest dose had 409 mice, 3 with malignant brain tumors. Not a significant difference in this entire table at any dose in any sub-population, even at p = 0.05 levels.
Table 2 focused on schwannomas, and they had to dig deep to male mice at highest exposure (n = 207) to get a significantly significant (at p = 0.05) difference. We're talking 3 / 207 male mice with malignant schwannomas at highest exposure. The control males had no cases (n = 412), but we're really in the weeds here where a stochastic variation of +/- 1 mouse makes a huge difference in their tallies. No other significant difference in any other dose in any other sub-population in any other table in this paper.
Kaplan-Meier survival curves (Figure 3 g-h) look just about identical for all doses: we're not seeing a big difference in survival times at any doses. And there's no effort to estimate error bars for those curves. That's a hint about (lack of) replicates.
From what I can see, there was exactly one replicate for each group / arm (e.g., mice exposed to a specific dose). This is not good, because technical and biological variability can cause flukes and false differences. 1 technical replicate per arm: if a technician had a bad day or screwed up a protocol when the exposing the mice to the highest dose, your one measurement set could be off. 1 biological replicate per arm: a weird batch of mice, or a batch of sick mice, etc., could throw off your one measurement set for the arm. Most cell line experiments we've worked with have at least 3 technical and biological replicates, in very controlled culture conditions. You'd be amazed at the variability, even in "identical" cells.
Oh, and read the neat Nature story (summary) where the sex of the scientist performing the experiments on mice can cause statistically significant differences. Because the male and female scents in our clothing can actually induce stress hormone changes in mice. Experiments are sensitive. Replicates are a good thing.
Thanks, Bruce. That was my first question at this story, and I appreciate hearing it from your expertise!!!!
No problem, and thanks for reading and replying.
I was confused on that, too, and neurobiology is pretty far from my regular work. Part of the problem is that there's a lot going on here: plasticity and differentiation (cell adaptations, transformations from one phenotype to another, differentiation, etc.), cell proliferation, tuning of connections, etc. It's a messy area with lots of new and sometimes contradictory data coming out.
From what I understand, there's a lot of plasticity in the brain as an adult: far more than was originally appreciated. I thought that we "knew" there were no new brain cells, and then we "knew" there were, and now we may "know" there aren't. That's the nature of evolving science, as others point out. And imprecise science communication--and imprecise English--doesn't help, either. Even biologists can get tripped up: we talk about tumor growth and cancer cell population growth, but we really primarily are talking about cell division there.
And there are all sorts of neat surprises. It was found that glial cells (a type of brain cell responsible for maintaining brain structure) can transdifferentiate into endothelial cells (which make blood vessels). See this PNAS paper and this one. This has all sorts of implications for gliomas and other brain cancers. And God only knows what other surprises are waiting to be found.
I suppose that's another reason they looked at the "new neuronal cell" marker: to see whether other cells could become new neuronal cells by transdifferentiation or other plastic processes. Biology is weird. :-)
There's a big difference between growth of cells (they get bigger) and proliferation of cells (they divide to create new cells), and so it's important to be careful. (I've been working on modeling both cell growth and cell division for a good while, mostly in cancer and a little in tissue engineering and synthetic biology. e.g., here.)
I looked at the study. They stained for Ki-67, the gold standard immunohistochemical marker for cell division. Cells that are actively cycling--in late G1, S, G2, and M phase, and a smidgen of G1 phase after division because Ki67 protein doesn't instantaneously degrade--stain positive for this marker. In particular, it is a nuclear marker, so the stain is localized to the cell nucleus, and the stain is very definitive. It's one of the easiest immunostains to do image processing on, because you can do nuclear segmentation, then analyze the colors in the segmented nuclei to see if they stained positive or negative for Ki-67.
And that Ki-67 marker was virtually non-existent in the region of interest in all the samples above 13 years old. See Figure 2. This is *the* universal gold standard marker for cell division used across pathology and experimental biology. So yes, the study indeed found no proliferating cells in the GCL. And then they used this "young neuronal cell" marker (DCX+PSA-NCAM+ cells) to further confirm what they already saw in Ki67.
Also, the Nature link is the *summary* of the paper, and not the actual paper. It's pretty common for the big journals to ask for a non-involved scientist in the same field to write a summary and commentary when a potentially controversial or significant paper comes out. Here's the actual paper.
Just please make sure that the measure of activity is not just by commits, code changes, etc. Some projects are just "feature complete / done" and shouldn't be de-listed or downranked in searches.
Moreover, it's a tradeoff. If the battery loses capacity a little faster but charges in 18 minutes instead of 90, then it may not matter if you have to charge more often.
It may go straight to File 13, but I wrote in too.
Its built-in linux-like commands don't appear to work as expected, I guess due to aliasing. "ls" is fine. But with any arguments, and I get stuff like this:
PS C:\mingw\bin> ls -l
Get-ChildItem : Missing an argument for parameter 'LiteralPath'. Specify a parameter of type 'Sys
again.
At line:1 char:4
+ ls -l
+ ~~
+ CategoryInfo : InvalidArgument: (:) [Get-ChildItem], ParameterBindingException
+ FullyQualifiedErrorId : MissingArgument,Microsoft.PowerShell.Commands.GetChildItemCommand
I've been happy enough with cmd + mingw/msys. That got me ls, make, and all the same goodies with the same syntax as linux. But if Windows PS is going to have "ls", it should work like "ls" or get out of the way. (Just like they evidently do in linux and macos powershell.)
Thanks for your reply. It's an interesting discussion, and indeed, I get a little fed up when even in academia, translational medicine morphs from meaning "translating theory into practice" to meaning "getting patents and making profitable startups." It's needed, but it can sometimes distort the field and culture when it becomes an ends and not a means.
I'm a little curious as to your definition of techie, because a lot of the discussion really boils down to how you define a techie.
When you say this, this makes me think your Venn diagram for "technies" and "academics" has no intersection. But ask most any grad student, postdoc, or faculty in an engineering, CS, or applied math department (and increasingly many biology departments), and they'll likely regard themselves as techies.
It seems to me a reasonable definition that a techie is a technophile, particularly one who loves, uses, and improves technology in their daily work and hobbies. But if you restrict your label to Silicon Valley and tech startup types, you're lose most of the amateur techies, the open source people, and the citizen scientists.
Again, it's an interesting discussion, so thanks. (And great to meet you here on Slashdot!)
Most of us who entered science and academia did so to make the world a better place, and many of us are techies. You'd be amazed at home much coding and tech is required for pretty much every area of science today.
We're writing open source software to solve real problems in science and engineering. We're spending the last of our startups on open access for our papers because it's the right thing to do. We're contributing to open data repositories because sharing data makes all our work better. We're writing free content on blogs, code tutorials, and MOOCs for public outreach, because we view our roles as educators seriously.
Most people in academic endure years of low pay and job uncertainty as postdocs and entry-level faculty--and defer or postpone indefinitely having children and buying that starter home--rather than faster and better-paying paths in industry, IP law, and mathematical finance because we do want to make the world a better place, and we're actively working on it.
So, while I agree with your general feeling, take a look around, and you'll see more techies trying make a difference that you might have realized.
... this could handily digitize 1 LOC.
More seriously, this could be fantastic for opening up old archives and making searchable.
Sounds like Spain tried to do this almost verbatim:
source: [thestack.com]
Sounds like it's planned to only ascend when needed, so nothing stationed in space. (And in any event, not in orbit until the point where treaties are moot.)
Mersenne Twister is pretty much the standard for simulating a uniform distribution in a lot of scientific computing. These depend not only upon unpredictability (useful for avoiding biases, and clearly important in the security realm), but also upon properties of the uniform distribution.
But when we test it out, we find it's still not as great as we'd like: look at a histogram of outputs, and you'll see that until you get really large numbers of function calls, the histogram isn't particularly uniform. (In particular, numbers near the bottom and top of the range don't get called quite as often.) This means that simulation properties that rely upon uniform distributions over both long and short time periods may be thrown off, and short- and mid-time simulation results may well stem from the MT rather than from the mathematical model. Moreover, low-probability events may have artificially smaller probabilities in the simulations (because of the non-uniformity of the distributions near the bottom and top ends of the range).
Over very short numbers of function calls (a few hundred to a few thousand), the outputs can even tend to cluster in a small neighborhood. So suppose that you are simulating a tissue with many cells, and calling MT as your uniform distribution to decide if the cells divide or stay dormant (each with independent probability p, so each cell divides if PRNG/max(PRNG) < p). The math says that for a uniform distribution, you don't need to worry about what order you evaluate your decision across all the cells. But if the PRNG outputs cluster over several sequential calls, then a neighborhood of cells may simultaneously divide if they are all evaluated close to one another sequentially. In analyzing the spatial behavior of such a simulation, you may draw incorrect conclusions in smaller spatial structures that, again, derive from non-uniformity of the PRNG, rather than problems with predictability. (And then you may accept/reject a particular hypothesis or mathematical model pre-maturely.)
So, there's definitely more to it than just unpredictability, depending upon where the code is being used.
Have you tried mingw-w64 / msys? The CLI tools integrate pretty well (make, ls, tail, head, etc.), and they have a package manager to grab these things.
In part, this is what preprint servers like arxiv and bioarxiv are for.
However, there are deeper-rooted, cultural issues at play here. Academics are rated on their job performance (for keeping your position, finding tenure-track positions, and later attaining tenure) based upon their peer-reviewed publications. Traditionally, this has meant going through the private, paywalled journals.Likewise, getting grants requires publications in peer-reviewed journals, rather than just posting online.
Now, posting in open access journals (like the PLOS family of journals, PeerJ, etc.) helps here, since at the least the access isn't paywalled. But now the academic / lab itself has to pay a much larger publication fee. (Often on the order of $1500 per article.) Moreover, many of said tenure review panels and grant review committees judge you not just on whether you've published, but where. Impact factor matters, and that again tends to steer people towards glammy, paywalled journals like New England Journal of Medicine (which just made a big kerfluffle about research parasites), Nature, Science, etc.)
So, there's a lot going on here. And even the scientists who want to just post preprints and move on are facing tremendous pressures.
Mobile hotspot. ... etc.
For when you don't want to pay the outrageous wifi charges at a hotel conference center, or Starbucks' connection is flaky, or Charter / Comcast / Timewarner crashes for the evening and you have work to do, or
Although some of these SOCs reported heat problems, the heat pipe here might have as much to do with the water-resistant enclosure (can't vent air past a heatsink as readily in w watertight case ...)