Posts by Paul Clark21 posts • joined Thursday 10th January 2008 16:36 GMT
I don't get this at all... The problem with strcpy() which causes so many overflow vulnerabilities is that the amount copied depends on the unpredictable C-string length of 'src', but memcpy() has an explicit length. Allowing that length to be greater than the size of 'dest' is sheer stupidity, not just laziness, and memcpy_s() won't protect you from that.
But who in Microsoft other than kernel/driver folk (who ought to know better, and who will resent the redundant comparison) is working in pure C anyway?
Ignoring the substantive point of the article entirely, the subhead was the biggest laugh (well, aggravated smile) I've had all day. subbie.points++;
First August 29th 1997, then July 25th 2004, now possibly sometime 2011:
http://terminator.wikia.com/wiki/Skynet
Just don't expect to beat it with Cavalry...
Please get your basic physics right: The majority of the ice in the Arctic is floating and if it melts has no effect on sea levels. It's the Antarctic ice you want to worry about in that regard.
Another effect of coal soot that you didn't mention is that the ice may be getting 'dirty' so that its albedo reduces, leading to local warming.
Huh? I spend one or two hours a day chatting on Skype with colleagues in Hong Kong from sunny (sometimes) Cornwall. Ten years ago I would be hopping back and forth on a plane.
We deliver all our products electronically; twenty years ago we would at minimum be burning and shipping CDs and using couriers all the time.
Today I'm logged into our company e-mail and development systems from home which has saved me a 16 mile round trip.
Perhaps the Professor should think about 21st century technology instead of 20th...
Congratulations are due to the Debian team for getting this out; a mammoth effort. First I knew about it was when the automatic updater offered me 800+ new packages (I have my apt sources set to stable). One 'upgrade' and 'dist-upgrade' later I was running Lenny perfectly - just one voluntary reboot to switch to the new kernel and I was off. That's what I call an upgrade mechanism!
Iceweasel = Firefox in all but name. Do some basic research before posting!
As someone who has thrown way too many old PC cases into the metal scrap bin, it's a nice idea, but it would almost certainly have to be lined with tinfoil to get theough electromagnetic emissions requirements.
Hang on a minute... Disk on an iPhone (any phone, netbook...) is Flash, which has limited write capacity, and Phantom is going to continuously snapshot memory?
I'm not convinced this is needed at the OS level, but it would be nice if more languages supported proper persistent objects.
Imagine a network where you could offer multiple independent services on the same subscriber line, each with configurable bandwidth rates and type (constant, variable, available, unspecified), with tiny packets and simple cell switching to keep router costs down. Then you could buy a low-jitter but low-bandwidth service for your VoIP *and* a high-bandwidth but best-effort connection for your P2P apps.
Funny thing is, you're probably already using it: ATM. That's what the VC/VP numbers are in your ADSL router - but there's only one channel down which we're trying shoe-horn everything.
But back to the topic: Akamai and others are already doing this edge-caching trick, available to anyone who can pay. Last time I checked, Tier 1 backbone connectivity and container-loads of servers weren't free, either. What's different about Google wanting to fix a few bottlenecks for itself?
Whenever I read about these huge datacentre builds and their massive energy consumption, I always wonder if anyone is seriously looking at the efficiency of the software platforms and applications.
There are so many layers of abstraction in modern platforms that I doubt anyone knows where all these cycles are going, but most of these services are doing a fairly simple job, just at a huge scale. Do you really need a container-load of servers to run Hotmail? It's a big NAS system, basically. Big disks, yes; fast networks, yes, but this shouldn't be a CPU-bound process.
One had the impression that Google in the early days understood this, and that their search architecture was built from the ground up with Real Programmer technology (Linux, custom C++ daemons, splitting IO from processing) - now they're doing every application under the sun, have they lost that skill? Or is it just there aren't enough Real Programmers to cover all those bases?
(Mine's the one with the 1985 (confidential) ARM1 spec. in the pocket.)
Tim: Like most people in the software business, I know all about GVA through intangibles, thanks. Musicians and artists are the same, as you point out, as are labour-only services (hairdressers, gardeners,...) and you could argue to put financial services in that pot, too. But these are all tertiary additions to our existing and growing resource usage in the underlying primary/secondary economy.
As 'Pete' points out above, the major issue is how we get the rest of the world up to the level where you can even think about growth in terms of intangibles, within the planet's many resource constraints. Even the primary needs of food, shelter and clean water are a mammoth task, let alone the transport systems, chip fabs and concert halls necessary for your examples. So you might be right in some academic economics sense within a limited scope, but it bears no relation to what is happening in the real world.
What possible meaning can now be ascribed to the "program for a computer" exclusion if they allow this? Arguably (and it will be so argued) any program creates a "better computer". The UK-IPO's stance on this was entirely logical and consistent with the intent of the legislation, but the CoA has just stomped all over it, apparently in the interest of "harmonisation" with the EPO.
This argument that this is good for SMEs is completely bogus: most SMEs are far more likely to get beaten into the ground by other people's software patents than have any chance of writing, filing and maintaining one of their own. Even the research to verify whether you might be tripping over something would effectively cripple any actual development.
My MP3 player is thinner than that! Why not provide a micro-B USB lead like every other portable device out there? Or better still, give away a USB-7816 adaptor...
Est-ce un billet SNCF dans votre poche, ou êtes-vous simplement heureux de me voir?
Je vais prendre mon manteau.
Steve,
FWIW, I've estimated the difference in baselines as follows... UAH and RSS are basically the same, and the lowest, so taking them as the datum:
HADCRUT = UAH/RSS + 0.146K
GISS = UAH/RSS + 0.238K
Here's a plot of all four series with baselines adjusted.
http://www.woodfortrees.org/plot/hadcrut3vgl/last:360/offset:-0.146/mean:12/plot/uah/last:360/mean:12/plot/rss/last:360/mean:12/plot/gistemp/last:360/offset:-0.238/mean:12
If you'd like to check the working, here's where I first posted it:
http://wattsupwiththat.wordpress.com/2008/05/05/rss-msu-lt-global-temperature-anomaly-for-april-2008-flat/#comment-13730
Paul Clark
woodfortrees.org
This all has the sense of a lot of greybeards musing over a theoretical problem which went away 10 years ago in the real world. In commercial development all serious Web and client-server development platforms (Web servers; databases etc.) are already multithreading, and application developers are already using multithreading without even being aware of it.
In non-trivial desktop applications, embedded and systems-level development, the need for synchronisation etc. never went away, and although C++ doesn't have synchronisation primitives, most people either use Boost or have their own wrapper around pthreads.
Just about the only high-volume area which I can think of where people might need to think a little harder is in areas which need serious grunt like video codecs - but these are often precisely the easiest things to decompose into parallel execution.
[Paris, because she lives in a parallel universe]
I thought my fellow IT folk might enjoy a programmer's playing in this area:
All four temperature series (5-year running means)
http://www.woodfortrees.org/plot/hadcrut3vgl/mean:60/plot/uah/mean:60/plot/rss/mean:60/plot/gistemp/mean:60
Detail of last ten years
http://www.woodfortrees.org/plot/hadcrut3vgl/last:120/plot/uah/last:120/plot/rss/last:120/plot/gistemp/last:120
Bear in mind that the different sources use different baselines to calculate the anomaly values so it's only the slope that counts.
(Tux because that's what it runs on)
The worst part about all this is that some the fuel is being grown on cleared tropical rainforest and peatlands, which causes hugely more CO2 emissions - not to mention loss of biodiversity, soil erosion, flooding etc. etc. - than whatever infinitesimal gain is supposed to come from not using fossil fuel. At the very least there needs to be an immediate ban on fuels from such places.
Flames, because that's what Indonesia will look like if this isn't stopped soon.
What really incenses me - after a lifetime of never buying tropical hardwoods - is that the biodiesel now running 1/40th of my car is quite likely sourced from palm oil grown on devasted Indonesian rainforest.
I'm all for local, sustainable biomass, but this whole biofuel malarky is an attempted (and royally failed) attempt to look like Government is doing something while assiduously avoiding fixing the real problem, which is the appalling inefficiency of vehicles and our transport systems in general.
Paris because she has about as strong a gasp of market economics and unintended consequences as the Powers That Be.
This group of "former military officers" doesn't include one General Jack Ripper, by any chance?
Folks, this isn't about which language is best for production development; it's about which is best for illustrating the core principles of Computer Science. Java is great for certain kinds of high-level development, largely because of the scope of the libraries and development platforms that come with it.
If you actually want to understand what's under the hood, though - which is what Computer Scientists are supposed to do - you'd be better off trying to implement a JVM or JIT in C or C++. And if you _really_ want to understand what's happening, you need to implement a kernel, which can only be done in C with small bits of assembler.
(CS degree, former assembler & C programmer, now using C++ *and* Java *and* PHP, for different jobs)
I wouldn't write off P2P yet; the execution may be flaky in this case, but the principle is sound.
If the iPlayer streaming version takes off, the server bandwidth required is going to be astronomical. Also, what do you do for people on the edge of the network with limited broadband? Background download may be their only option.
Of course, the real solution to this is to download over multicast...
