Profile

Cover photo
Nick Alcock
Worked at Oracle Corporation
231 followers|54,056 views
AboutPostsPhotosVideos

Stream

Nick Alcock

Shared publicly  - 
 
Oh, the increasing whoopses coming out of #panamapapers are so very much fun to read about. (Though, of course, only the half-decent people with an actual sense of shame will resign upon being found out.)
The president of the Chileanbranch of Transparency International resigned on Monday afterdocuments from a Panamanian law firm showed he was linked to atleast five offshore companies.
4
1
Add a comment...

Nick Alcock

Shared publicly  - 
 
So, could anyone more insane than me take a look at in particular the first of these:

<https://sourceware.org/git/?p=glibc.git;a=blob;f=sysdeps/unix/sysv/linux/i386/pthread_cond_timedwait.S;b=master;hb=HEAD>
<https://sourceware.org/git/?p=glibc.git;a=blob;f=nptl/pthread_mutex_lock.c;b=master;hb=HEAD#l524>
<https://sourceware.org/git/?p=glibc.git;a=blob;f=nptl/pthread_mutex_unlock.c;b=master;hb=HEAD#l36>

(specifically, the calls to __pthread_mutex_unlock_usercnt() and __pthread_mutex_cond_lock_adjust()) and tell me if their caller in that assembly file (and also the other caller in the corresponding pthread_cond_wait.S) is really doing what I think it's doing?

... that is, calling both of those functions after setting up their stack frame but without bothering to adjust the stack pointer, depending on the layout of the stack frame of the things it calls to not overwrite its own variables?

Am I seeing things? (Various experiments would suggest not, but I'm really not very good at x86 asm and could well be wrong.)
1
1
Richard Kettlewell's profile photoNick Alcock's profile photo
4 comments
 
Patch resulting from your suggestion: <https://sourceware.org/ml/libc-alpha/2016-03/msg00226.html>.

Thank you!
Add a comment...

Nick Alcock

Shared publicly  - 
 
Number of times glibc built and tested in the last two days: 26. Number of tests with different configure flags, bitnesses, etc left to do if all goes well: 6.

This is really boring.
6
2
David Cameron Staples's profile photoAlex Bennée's profile photoNick Alcock's profile photo
11 comments
 
... aaand it's under review. Let's hope this will quash all bugs even remotely resembling CVE-2015-7547 forever (yeah right). (The process of cleaning it up was worthwhile: I figured out the cause of, and fixed, the one remaining subtle horrible bug in it -- a timing-dependent stack corruption in C code called directly from x86 assembler with a handcrafted stack frame -- while I was trying to characterize said bug enough to describe it to the glibc hackers.)
Add a comment...

Nick Alcock

Shared publicly  - 
 
So everyone's talking about the black hole merger results but nobody's linking to the publication summary... (warning: rather slashdotted, server may be slow).
4
1
Add a comment...

Nick Alcock

Shared publicly  - 
 
Misleading headlines of our time.

Another way to put this might be 'Apple reports highest ever iPhone sales', but because they're not growing fast enough this is clearly terrible. Because the market for phones is not in any way saturatable and everyone will still want another phone even when they have ten.
Apple has reported the slowest ever iPhone sales growth since the product's 2007 launch.
1
Jasper Janssen's profile photoPete Taylor's profile photoNick Alcock's profile photo
5 comments
 
+Pete Taylor, I tried to run Linux on a lentil but it said I couldn't do that as long as I had a pulse, so I gave up. I guess we know what the zombies are using to coordinate their shambles though.
Add a comment...

Nick Alcock

Shared publicly  - 
 
Friends don't let friends use JavaScript.
Ajax fans are thought to be responsible for a major disturbance in Glasgow on Thursday before their team's Europa League game against Celtic.
2
Pete Taylor's profile photoNick Alcock's profile photoKoos van den Hout's profile photo
3 comments
 
Those fans of running their entire webshite in your browser are a rowdy bunch!
Add a comment...

Nick Alcock

Shared publicly  - 
 
I'm wondering about the structure of the Young Wizards interim novel Lifeboats. My ebook has no less than three empty chapters (chapters 1, 12, and 13) and a total page count that shows as 333 even though the last page in the ebook is numbered 272. I'm reasonably sure it's complete -- it reads like a complete work -- but the combination of multiple empty chapters and a page count that's fifty pages off makes me wonder if everyone else is reading fifty pages more of this than I am...

(typo rate: lowest of all. Still, I have on the order of five hundred to a thousand typos in the New Millennium editions spotted now, including pervasive ebook conversion faults in A Wizard of Mars that make it quite hard to read in larger font sizes. Next step: stick them in a sqlite database so I can make some sense of this giant haul.)
1
Diane Duane's profile photoNick Alcock's profile photo
5 comments
 
Ping? No unexpired link has come my way yet... :)
Add a comment...
Have him in circles
231 people
Lenz Grimmer's profile photo
Eleanor Blair's profile photo
Lauren Clark's profile photo
Toan Le's profile photo
Martin Karlsson's profile photo
Seth Breidbart's profile photo
Benjamin Dara's profile photo
Jan Badenhorst's profile photo
Zhou Augustinus's profile photo

Nick Alcock

Shared publicly  - 
 
It turns out I am a normal human being: to wit, I can be sent into drooling gadget-fanboy mode by a single tweet, as long as it's about the right gadget.

I don't even need, y'know, any details at all. It could be total crap, it could actually be made of manure (ecologically friendly!) but I'm still thinking BUY NOW.

This is clearly irrational and when actual details come out I might change my mind. But, BOOKS.
3
Jasper Janssen's profile photoNick Alcock's profile photo
12 comments
 
+Jasper Janssen, yeah, but this originated back in the K4 era, when they had only 1.2GiB or even less, and for people using 3G Kindles size optimization saves Amazon money too: and it doesn't affect customers other than to avoid bloating up their devices with guaranteed useless information (the things are already stripped down so that only the preferred format is shipped: this is just part of that, really). So why on earth would they stop doing it? All that together saves on the order of 70% of book size, which means they can advertise three times the effective storage capacity for the same manufacturing cost.

It's just common sense.

(And I do maintain my entire collection on-device. So do quite a lot of people who like to reread stuff at whim.)
Add a comment...

Nick Alcock

Shared publicly  - 
 
"Currently git won't set people whose commit is rejected on fire, but you can explain in the company policy that they're obliged to set themselves on fire in this case."
2
Add a comment...

Nick Alcock

Shared publicly  - 
 
Another interesting paper, ASTROPHYSICAL IMPLICATIONS OF THE BINARY BLACK-HOLE MERGER GW150914, by (as usual) Abbott and several pages of et al.
2
1
Add a comment...

Nick Alcock

Shared publicly  - 
 
So at least some cyanobacteria don't use the good old bacterial time-based "wander randomly but turn more often if conditions are getting worse" technique for finding the light: instead, they use their own cell bodies as a lens, focusing incoming light rays onto their own cell membranes, then move away from the side hit by the light (thus towards the light source).

(The lens actually forms a low-resolution image of its surroundings, but I can't imagine that the bacterium uses that imagery for much other than light detection, given the mechanism in this paper. Mind you, I'd have thought the same was true of box jellies, and then some results emerged suggesting that some species could identify trees though the water surface and loiter beneath waiting for prey to fall into the water, and possibly even distinguish between tree species -- and all this with no more than a diffuse nerve net.)
2
Add a comment...

Nick Alcock

Shared publicly  - 
 
It is hard to describe how awesome this is, if true.

This new carbon phase (produced at room temperature and pressure, more or less!) is believed to be denser and harder than diamond, ferromagnetic (!) and appears to have interesting electrical properties as well. To me it looks analogous to liquid crystals: an amorphous supercooled solid, only made out of pure carbon.

The phys.org link is to a simplified publicity wossname. The original papers appear to be:

<http://scitation.aip.org/content/aip/journal/jap/118/21/10.1063/1.4936595>
<http://scitation.aip.org/content/aip/journal/aplmater/3/10/10.1063/1.4932622>

With results like that, that they can also produce diamond in various unusual forms (like nanoneedles) at room temperatures and pressures (the focus of the second paper) seems almost boring.
 
Harder Than Diamond
Just like water appears in different phases sporting distinct properties (solid ice, liquid water, gaseous vapor) so too carbon, that most versatile of all the chemical elements, appears in multiple solid phases including graphite (pencil lead) and diamond.

Recently researchers at the North Carolina State University have reported the discovery of a heretofore never seen solid phase of carbon that they named Q-carbon. Q-carbon expresses extraordinary properties and dispositions: it is ferromagnetic and also harder than diamond, furthermore it can glow when exposed to even a small amount of energy.

For more information, see:

http://phys.org/news/2015-11-phase-carbon-diamond-room-temperature.html

http://www.sci-news.com/physics/q-carbon-new-allotrope-carbon-03476.html

http://www.forbes.com/sites/ericmack/2015/11/30/scientists-create-new-kind-of-diamond-at-room-temperature/

https://en.wikipedia.org/wiki/Q-carbon

http://scitation.aip.org/content/aip/journal/aplmater/3/10/10.1063/1.4932622
Researchers from North Carolina State University have discovered a new phase of solid carbon, called Q-carbon, which is distinct from the known phases of graphite and diamond. They have also developed a technique for using Q-carbon to make diamond-related structures at room temperature and at ambient atmospheric pressure in air.
1 comment on original post
7
2
Richard Kettlewell's profile photoGreg A. Woods's profile photoNick Alcock's profile photo
4 comments
 
Yeah -- there's a reason I hunted down the papers (which phys.org did not link to that I can see, sigh). phys.org is... not the most reliable of secondary sources, shall we say: I much prefer the original papers when at all possible.
Add a comment...
People
Have him in circles
231 people
Lenz Grimmer's profile photo
Eleanor Blair's profile photo
Lauren Clark's profile photo
Toan Le's profile photo
Martin Karlsson's profile photo
Seth Breidbart's profile photo
Benjamin Dara's profile photo
Jan Badenhorst's profile photo
Zhou Augustinus's profile photo
Basic Information
Gender
Male
Other names
Nix
Work
Employment
  • Oracle Corporation
    Software Developer
Links