Posts by phil dude 1937 publicly visible posts • joined 30 Nov 2011
Everytime I see another IT cockup from Govt that *we* pay for I am reminded of something from a P.J. O'Rourke book about the 4 ways to spend money.
"If you are spending someone else's money on someone else, who cares?"
I know this is not a magic solution, but if Govt contracts had a default clause that all the source code for these failed "spend until you bleed" projects became the public property? Perhaps we would get solutions that were more likely to converge on the correct solution, rather than the random walk these inevitably seem to be....
P.
I browsed the paper. I had to get to it behind a paywall, so it must be hot off the press....
From mammalian biology, all mammals are female unless they respond to androgenic hormones.
Female body type != XX about 1in 100,000 times (hard to collect stats on this).
It would have been more interesting if they had taken measurements of the sex hormones and correlated those with "connection analyses" since the negative control already exists (XY, andogren insensitive, female body type...).
I wonder if the study was done blind, did the researchers just get de-identified data?
Sorry for being the cynical scientist, but studies as complicated as this will have all sorts of misinformation read into them...
P.
P.
yes green have a *nasty* habit of turning off, and being slow to turn back on!
Get RAID drives for RAID setup..!!
I got this HP Micro server box with 4x2TB WD RE4-GP using RAID6/1 in it. A lovely piece of kit and relatively cheap when you are buying 4 drives, especially since it takes normal ECC memory.
P.
I was at a party over TG and got talking to a neutron scientist. Apparently you can make helium-3 to order and they do, since they have to recycle all they use...
This came as a surprise, but when synthesis is possible, it suggests the market may produce a solution...
Other than that, it would seem that recycling would be prudent...
P.
yes I am familiar with that technique. Leaving a protein scaffold for new cells. Quite a nice procedure if I do say, as it nicely sidesteps the massive challenge of inducing self-assembly into cells of unknown state.
The eye is part of the brain, hence , growing one is quite a challenge..!! In mice where genomic plasticity has been engineered it is possible to mutate all sorts of cells, but with eyes we can knock them out (induce blindness ) , change their proteins (for sight problems), but the actually structure is quite subtle and dynamic.
It is one of the overarching mysteries of biology that all cells have the same potential (i.e. DNA) but different actual states. Very difficult to measure, but slowly we are getting there (RNAseq etc...).
And it would be effectively impossible without computers.... (IT angle)
P.
It is undoubtedly very cool stuff, and we are just getting started with. The de novo synthesis of cardiac tissue is very impressive.
My synthetic biology work is at the molecular level i.e. the design of protein biosensors that match a patient with a particular illness, and can then be tested against known safe pharmaceuticals looking for a therapeutic match.
In my travels one of the very best supporting explanations for the utility of this technology (organ printing etc.. ) is one of personalisation. Everyone is unique, but more specifically , all our organs shapes are unique too. Although organ donation availability is by far the top driver for synthetic organs, there is also a strong argument for autoimmune compatibility and the ability to match the organ sizes. As it stands you are luck if they find an organ, luckier still if they find the right HLA match, and even luckier if the organ will fit in your body, all within the time frame both you and the organ are viable....
P.
Used PAYG Three and my own phone (N900,N8,N9..) in Oxford. Jumped from VM who couldn't explain charges on their own bill, so perhaps this article is relevant their...
With PAYG no money to steal, (unless phoning those bloody "national rate" help lines). Three are smart enough to allow you to not store information, so topups are done online or entered everytime. Since all-u-can eat data also...
Seriously, if cash is an issue surely PAYG is the way? Here in the USA T-mobile offers $1/day (first use)...
Just saying, not for everyone but...
P.
my local credit union is much better than any UK bank I have dealt with.
No charge if you don't have a cheque book. I haven't written a cheque in years anyway..
But everything is nicely segmented online. You can download all your records (I have to send a
SA request to get this from Barcalys! )
I don't recall where I read it, but the reason chips have been slow to be adopted in the US was the perceived danger in having a piece of information a criminal might need to use your card. So no chip, bag swiped, no need for owner.
Plus, they print the chipless cards in the bank for me... any design you want which is just ,nice.
P.
In fact an experiment has already been done, as mentioned, by Netflix. As a trans-atlantic commuter I was thrilled my US account worked in the UK, only the content was *radically* different.
I would be interested to see if there is an objective comparison of this phenomenon i..e which titles etc... are missing between the 2 English speaking markets.
Here in the USA there is also RedBox, which partly fills the Netflix gap - usually very near (even by Euro standards) and cheap (sometimes < $1), and 24 hour (outside grocery stores.) for a DVD. They often have first run DVD's quite a bit before they are streaming.
The reality is, this is simply too convenient for those in power - any excuse whatsoever to limit the public access to information , is good enough.
An objective list of all media not available for legal streaming vs pirating vs language, would be an interesting study...
P.
I think we all agree we cannot avoid google, however the above approach combined with privoxy and noscript speeds up many sites!!
Page loading time seems as much bandwidth as relative latency between sites, and then the time for the browser to render all that!!
P.
if I could post a screenshot... , and this was in the USA.
Believe it or not I was not searching for a competitor, I was seeing if any of their associated material mentioned linux, as I cannot believe I am the only person to think "will it run linux".
So I entered "linux" and the only match it could come up with was "line".
Actually , I did this twice. The *first* time took me back to the landing page...
what does it do, set off an alarm so they can change the response....?
I stand by my comment, retarded.
P.
the Brown recluse spider bite here in the USA, can produce necrosis. I know, 'cos I have had it...! Really frightening at the time... I have read that its bite contains not just a toxin but a bacterial milieu that causes the necrosis.
As for this spider, I have search a bit on line and I cannot find the molecular structure for the venom, or even what functional groups it may have. Some vague "its an acid"...
We seriously need to DNA sequence many more creatures because at the very least, we can have an objective way of knowing what an organism is rather than shape... and about its biochemistry...
P.
there was an iplayer program a few years ago about superbrand. Alex someone?
Anyway, showing children (<5?) brands, it was very clear that certain sport goods were firmly ingrained, where as perceived "adult" goods were not.
Marketing is the overhead of being a big company, since for every %1 of market share, it becomes exponentially(polynomial?) more difficult to get the next %1 as we approach total saturation.
I am reminded in the Red Dwarf books about the detonating supernovae advertising to plaster "a certain carbonated beverage across the sky for a decades"....
It is the "refinement of advertising memes" that are so fiercely cultivated to be able to "divide and conquer" the buying public...
P.
All companies work on a "pay the staff what's left over is profit" basis.
Publishing companies have spent decades reaming us for dead tree versions (in the UK the "fixed retail price"), and they just want to stay cosy like all companies. I am NOT saying their costs are not a large part of producing a book, but it is complete FUD to say that e-books have not changed the market.I would suggest this is yet another industry that is struggling to adapt to the internet era and as was pointed out earlier in the comments "B&N have lockin on their device". There's the rub. So does Apple, and Amazon, they just DO IT BETTER.
I have dead tree versions of Terry Pratchett books, lets try an experiment, "Mort".
"The brazilian forest store" - Kindle $5.69 , paperback $8.99.
B&N - Nookbook $5.99 , paperback $9.99.
If you want the nice hardware, B&N are selling the simpletouch and HD for $39 and $79, respectively.
P.
this has been my hope for sailfish on a phone. Android but only when you want it. I have a TF101 and is has been a PITA with bugs that stopped getting patched once manufacturer moved on, and apps that can access ALL of you personal information with NO control (other than not installing).
Sorry google, you are looking very much as if you've put cyanide in the bird seed... (I hope that's not too obscure...)
P.
PS Happy Thanksgiving everyone ;-)
As someone who has had an eye on this as owner of N900,N8 and N9...
It has an SD slot, user replaceable battery, and custom other half. I spent a ton of time tagging all my music for the retarded N8 music player, so having it all on an SD card is *really* nice...
The keyboard is likely to be one of the "other half"s... there was a nice laundry list of possible halfs vetted by the company engineers a few months back... The other half connection has power,i2c and NFC capabilities. So a spare battery is definitely on the cards.
Apparently Sailfish OS will run on the N9 (for a risk free try).
I could buy one via the UK, but I am waiting for it to make it to the US so support is not an issue...
It may not be wanting to get between Android and Apple, but by being able to run Android apps and Linux software, it is definitely interesting...
P.
perhaps the issue of anonymous commenting in general is the issue? Why does a corporation need to know who you are anyway? They may want it, they may even protest they need it to make money, but so does selling my information (make money for them, that is...).
A few months back there was a flurry around some inappropriate comments on science websites, and the idea of a "respect" button or similar concept was punted. In slashdot, you can choose to not see anonymous comments?
If having your realname plastered on comments is the answer, I think double-triple-etc entendres could be on the up....
P.
thanks for the heads up (seem my comment below for a cryptlocker shield)!
The whole disk space thing kinda gets out of control but a tip I have found is read vs read/write many.
Home directory is read/write, but your music is probably not. So paritition that way. I know my media partition is at least 10X my user!
But I believe snapshots only occupy space if files change? That is how it worked under DFS...
P.
from what I have seen (they use SLES on some supercomputers...) it has a number of locked down features so that it stays stable. Older but patched versions of some stuff, and a SLES admin interface that is quite advanced, but I do not recall what is unique as some of it I have seen in 12.3+.
The thing about SuSE if you are a user, is that it typically just works. There is an EXCELLENT build studio they provide which means you can build for any platform if you want to play building without filling your system up with unecessary tools. But it does mean many packages not originally for Suse appear on it, and so via software.opensuse.org
It seems more or less compatible with Redhat not just because it uses RPMs (there is a tool called alien which can make you agnostic), but because it follows the same conventions.
I think all Linux's now have the systemd stuff, which if you are a user should be invisible.
But for all this I am still running 12.2!! Why? Because this machine has a buggy HD and weird network card that the new (12.3) suse did not have functioning.. It took me 6 weeks to workout that the HD was going to sleep and taking the rootfs with it!
If you like Gnome or KDE you can get either flavour.
If you are a windoze refugee wine is pretty good, but crossover is excellent - IF they support your need tools... Also it makes your windoze applications portable within linux. Once they work, chances are they will keep working.
They have taken the plunge with BTRFS too! BTRFS offers copy on write facilities. If boot Windoze on a VM, make linux export a samba FS using BTRFS and copy on write underneath, and you'll have a much saner environment... Cryptolocker would be little problem in this scenarios (so long as the data that mattered was on BTRFS...). Just saying....
P.
Think about the "friction" on your dough. Not just taxes , but the X % off EVERY transaction siphoned off because of a bit of plastic. Even changing bitcoins gets skimmed!! And govt's wonder why it is gaining popularity...
Let's face it credit cards are a cartel, and whatever you may think of the politics of bitcoin, it shows that in the internet era all things can be replaced. I'm just not sure with what....
Beer, because there is at least one pub that takes milli-bitcoins...
P.
I seem to remember that the Pan paniscus is a closer common ancestor than Pan troglodytes, genetically that is. The bonobo is also observed to be a maternal , social grooming and tree living rather than paternal, territorial and ground living as the common chimp is observed.
The upshot is, our ancestors survived living in the trees, apparently somewhere in shallow water, and that lead to current humans (via a very complicated and unknown path...).
The whole sexual favours thing (yes we giggled in class too) is perhaps a trait selected along the lines of "make love not war".....? I also remember it was not just hetero more omni ....
Oh and I think the common chimps can be canibalistic too...?
Beer, because that levels the difference between humans and chimps...:-)
P.
As a foreigner down in TN, from what I understand visa holders (H1B) must be paid the prevailing wage (i.e. cannot be cheaper) and the job must also be posted for 30 days in public.
Could you please explain how visa holders get preferential treatment as I was one, and it was a major PITA...!
I certainly didn't feel very special...;-)
P.
yes, I would agree age bias is widespread. But that I think that is a more prevalent UK problem. Someone earlier mentioned managers feeling threatened be younger grads? Make that threatened in general...!
Here in the US I have seen a much greater acceptance for "can you do the job?" and I have learnt a lot from some senior techs. Some with PhD's, some without a BSc, but generally extremely competent and generally very fleixible. It may not be all roses (and my experience is a lab...), but the USA has a far more practical mentality towards graduate education. Granted, it also has a shed loads of foreigners (including me!). Grads here can intern in real labs and get paid... If you are STEM there are many opportunites. There is fierce competition but these kids get involved from high school and up.
I'm a scientist and I am pretty confident there is nothing I couldn't code if I had to. Ok wouldn't be the best solution, but then my key skill set is application based - molecular biophysics/computational biology etc... I have still learnt to code HTML5/JS as it is a tool.
And there in is the rub. Someone who was building computers , writing games, hacking DVD drives (!) when they were at school, are a fundamentally different sort of person to the "management consultant" IT people that get paid a mint. Which is why the Govt IT is such a failure. Big companies want to deal with nice corporate stuctures. IT is a profession where you CAN be a lone agent and make a difference.
I have mentioned before in these forums that software quality is extermely hard to measure and so the metric for most comanies hiring is "cheap as possible" and when selling their cheap ware "charge as much as possible" . If you are building cars perhaps you will not last long, but selling crap software seems to be a feedback free loop. Witness the currnent healtcare.gov and then look at the "health sherpa" site, that was knocked up in a weekend by some grad students.
With as unbelievable as computing technology is (and is still growing!! SC 13 was cool! ) it is amazing the race to the bottom the IT industry seems to be intent on.
P.
The FDA may have a point here. I remember going to a talk from one of the prof's involved, and left feeling a little bit like it was over simplified.
Now if the company want to sequence my DNA and send me a URL to download the results, and THEN if I want them to analyse them, sure. I can get 30 different analysis quotes...
But far too much of the FUD about DNA testing is the simple fact that we have only sequenced a vanishingly small percent of humans ALIVE, COMPLETELY (both alleles).
The distinction for non-biologists is confirmation bias, since you can only have alive patients , deleterious mutations are hard to predict.... Some of the relatively homogenous populations of the world (say Iceland's) can lead us to think DNA is simple so leading to these "cheap" tests. I believe they just test SNPs, not exomes or whole genomes. These techniques can be VERY powerful but you need a lot of other information to be sure.
What is needed is for proper legal protection for what is found in every human's genes, worldwide. It is the most basic form of discrimination that you cannot choose your parents, and so there should be blanket protection. In this way sequncing DNA and therapeutics can be developed for the benefit of all, not just companies looking to lower medical costs, insurance companies looking not to pay out, and governments finding ways to denying human rights to citizens, or the inevitable scam markets. If you wish to have a proxy handle your data (e.g. your doctor) fine, but don't tell ME what I can do with MY data.
I'll end on a computing analogy for El reg. If you were called to fix my computer and I gave you the hardware details, the bios, the OS version, all the software installed, but did not tell you what work I did on it, you would be hard pressed to categorise any failings without observing the computer directly.
Think of that the next time you see your doctor, and think what you have done with your DNA that you cannot hope to explain... (this is epigenetics..)
P.
Well here in Denver I saw and spoke to the groups using the new Phi. I must say I am quite please there is a GPU for computing alternative in the market.
And to add to this discussion, this "phi" comes in all flavours and boots a normal (i686?) linux kernel, and has 240 cores.
I have bought AMD for years, but this is the first time I am thinking of getting an Intel product that looks generally interesting.
Ultimately computers are all interchangeable, it is the cost of the conversion that is decides what code runs on what.
Intel clearly have some excellent engineers....
P.
Guys calm down!
I flew to Denver a few days back on a United flight and the flight attendant gave the following annoucement.
"Due to a change in FAA regulations if your phone or electronic device has a method of disabling the wireless capability, often called a flight mode, you must enable it now and disable all wireless activity otherwise turn the device completely off.".
So, no phone calls unless cabin door is open has not changed.
Being able to read your kindle or phablet before in the air , yes.
Loads of people taking photos on take off and landing, yes.
There was something about devices must be stowed if not in hand and not in the seat back pocket, but other than that just the transmitting mode.
P.
Science suffers when the truth cannot be heard. By this I mean, we have a media obsessed by trivia, politicians obsessed by achieving the appearance of acting, and companies obsessed by the pursuit of profit at all costs.
The talk by the David Keyes (who i have met a few times and has a fine singing voice...;-) repeats an observation of a number of scientists in the US, although the analysis could be applied equally to any country.
Basically everything you think of a science in the USA comes from this tiny fraction of the federal budget. NIH, NSF, DOE , FDA, NASA, NOA.... EVERYTHING. Computing only benefits the world because there are applications that are written using algorithms developed by scientists and engineers towards the solution of scientific problems.
The problem with politics is the need to be seen to be doing something - i.e. picking winners (the spin is the obvious counterpart picking losers). Companies all want something for nothing. Hence, politicians spend more on entitlements , lobbyists get tax breaks for business, and less money is spent on funding creative activities. Putting a man on the moon? Sequencing the human genome? Anti-cancer therapies? Smart phones as powerful as the supercomputers of old running on a battery?
None of these are isolated activities they all require the blending of many, many iterations of scientific publication and engineering implementation. If you want to reap the benefits of science and technology you need to support the people involved in it. And that is not just in education but in the government agencies that (should) implement policies on behalf of all the people, and build the intellectual capital which permits a varied industry and supports a sustainable future.
The NIH doubled its budget in a decade and has had an enormous effect on creating new industries where there one (e.g. therapeutic genome sequencing, genotyping etc.., personalised medicine ). What could we achieve with a doubling of NSF?
I understand the frustration with the climate change fiasco, but this only goes to underscore my point. The real world is far more complex than the 19th century principles our society has modelled itself on. Don't throw the baby out with the bath water! The vast majority of science is independently verifiable and we all benefit as a result.
Computers are an astonishing technology, but without continued investment it will never reach its full potential.
P.
I'm going to some of the applications talks later, so see what facts there are.
But talking with one of the intel experts i asked the question "will it run linux?" as a sort of joke.
Pause. "Definitely soon."
Apparently the Chi has some sort of larabee type arrangement, and so it is 60 cores with 4 threads. The instructions are i686, which means you could probably boot some flavour of linux... I don't know but it would be interesting to see!
In biophysics (namd, gromacs) it looks as if it is about 3x a xeon. Not being a CPU bod, it would be nice to know what that details. But as scientist, faster is nice.
All in all, I went to the larrabee talk at the previous SC, and this is a nicely timed introduction...
P.
Don't forget the med/tech applications , printing organs..
Seriously, the most amazing thing about 3D printing is not the cheap trinkets that the naysayers keep rattling on about. It is the COMPUTER that can design such a thing and then operate the machine to construct it.
As someone who has dabbled in synthetic biology, imagine a computer that could design a biological system to grow itself...
Tell me that wouldn't be cool....!
Beer, 'cos a 3D printed beer would be just silly..
P.
so while we're on the spider stories , the brown recluse has a bite that is not so much poisonous as a biological weapon.
The spider cultures flesh eating bacteria that when injected into the human, begin necrotising the tissue...
The thing about flesh eating bacteria,is you are a live to watch them dissolve you....
P.
The big problem appears to me , is that forecasting events a week in advance how a very tight feedback loop if their wrong. Politically and scientifically. We may all joke about bogus weather forecasting, but it is a very expensive commodity to do with out.
No matter what side you are entrenched on , predicting things 100 years in the future us fraught with problems, and it simply defies belief that the scientific community would not WANT everyone to look at the data.
Unfortunately, until honesty is restored to this whole enterprise, even rational people will have a hard time interpreting "the evidence".
P.
A shout out to all those in Denver ;-)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
