the product name is singular, not plural (despite the last word ending in s) Then how do you explain the sentence "Intel® Threading Tools consist of the following:"? "Tools" is plural, and it doesn't matter how many adjectives you throw in front of it. The usual convention to make it singular is to suffix the name with a singular qualifier, like "suite" or "collection". Substituting form for content, as you suggest, isn't going to fix it.
Look at it as as function: sufficient( task scheduler and generic parallel patterns ). It's not a function, it's a poorly-written sentence. And if you knew half as much about threading and synchronization as you pretend to, you'd see that the writer actually meant to use the word "specific". So the sentence should start with "Along with specific task scheduler and generic parallel patterns..." (note the distinction between generic patterns and those specific to task scheduling). Make that fix and...you've still got a five-line run-on sentence.
If you understand what they are this sentence makes perfect sense. Um, I do know what they are (although the phrase "race condition" is generally preferred to "data-race", and "dead-lock" is almost exclusively expressed as "deadlock"), and if that sentence makes perfect sense to you, you should contact your ESL teacher and get a refund. The problem is that there is a random "of" masquerading as content and obscuring the meaning of the sentence.
I don't mind having to double-scan sentences in Slashdot posts, and I usually don't care if I run across something similar on other tech sites. But this is an Intel white paper, and I expect that things like that have been gone over by an editor, not just C&P from some engineer's e-mail. I'm also not faulting whoever wrote it; the content was fine, it just needed to be cleaned up (which is not the engineer's job). The fact that it wasn't is what is annoying me.
Geekoid, I apologize. Your comment does make sense, once I pull my head our far enough that I can tell what part of the thread you're on. I owe you a $BEVERAGE. Feel free to flame me back, I deserve it.
And thanks, stuktongue, I obviously shouldn't be counted on to figure this kind of stuff out on my own...
develops and release an profession compiler Too easy, moving on....
To say snide comments does not help, and shows that you have no real argument. Um, I'm not arguing. I'm making an observation. If you diagree with me, then you're making the argument. Which is fine, just so we know where we stand. Nice non-sequitur, though.
Grow up. So expressing dismay that a respected corporation is showing less-than-professional work is a sign of immaturity? Buy a vowel and solve the puzzle, honey, Real World moves and all...
The Intel Compiler Lab is based in two Russian cities - Moscow and Novosibirsk. Probably the source of the less than optimal text. The point is, whatever tortured, twisted prose was submitted should have been edited and polished before going out with an Intel logo on it. This was a white paper on the corporate web site, not a post on some random Intel engineer's blog — different standards apply.
Seriously, check out this opening paragraph from the Intel® 64 and IA-32 Architectures Application Note: TLBs, Paging-Structure Caches, and Their Invalidation
The Intel® 64 and IA-32 architectures may accelerate the address-translation process by caching on the processor data from the structures in memory that control that process. Because the processor does not ensure that the data that it caches are always consistent with the structures in memory, it is important for software developers to understand how and when the processor may cache such data. They should also understand what actions software can take to remove cached data that may be inconsistent and when it should do so. The purpose of this application note is to provide software developers information about the relevant processor operation. This application note does not comprehend task switches and VMX transitions. Notice how they even get the fact that "data" is plural right? That's the kind of documentation I'm talking about.
I was looking at the Thread Building Blocks paper, and it reads like it was somebody's hastily-scribbled draft:
"The Intel Threading Tools automatically finds correctness and performance issues" (The tools finds?) "Along with sufficient task scheduler and generic parallel patterns" (Who has insufficient task scheduler?) "automatic debugger of threaded programs which detects many of thread-correctness issues such as data-races, dead-locks, threads stalls" (Sarcasm fails me...)
And that's just in the first few paragraphs, I haven't even gotten to the real meat of the article!
I'm used to informative, well-written and reasonably complete technical documentation from Intel — WTF is this?
in about an hour I can create a functioning front end, with or without web functionality and with ODBC interface into SQL, Oracle, or just about any other backend known to man, and i can do it in at least 3 common well known languages. What other development environment will let me do that? Well, if you drop the restriction to ODBC, I think either Eclipse or NetBeans will fill the bill. Both support C/C++ and Java, and Eclipse also supports Tcl and NetBeans supports Ruby. Sun Studio supports C/C++, FORTRAN and Fortress, although I doubt you could create a simple database-backed app with a GUI in an hour. Not with Fortress, anyway, which is all I've used SS for.
In about an hour I can teach a novice programmer how to create a fully functional windowed application that can actually do something, again in multiple languages, and using a familiar interface. Great, now try changing the interface. What if you have to deploy your app with a couple of ancient Motif-based apps (or a couple of new GTK+-based ones), and the client wants them to look the same? Is there any way to drop in an interface library and use it? Not everyone lives in a monoculture (and there are fewer every day).
Not sure if you're really making a point by juxtaposing "novice programmer" and "multiple languages", I think you'd wind up just wiping asploded head off the walls if you wrote your forms in VB and your back-end classes in C# and expected a novice to make sense of it all.
Seriously, VS isn't bad (although the Express versions only support one language at a time), but it's hardly the only IDE you can be productive in. Check out a 4GL sometime if you want to see some serious RAD...
a typewriter really is all one needs to write a novel Hence the reason novels and typewriters appeared simultaneously...
I do not think a computer helps one write a novel thousands of times more quickly Nope, although many people still find it quicker to edit with a red pen on a double-spaced proof page than with the word processor. Of course, the actual revision still has to be done on the computer, but the real work is done on paper.
Personally, I can (and do) work either way, but if you're doing major revisions you can't beat the ability to lay out an entire chapter on a table (or bed, if your table's like mine) and have it all available without having to scroll around. Contrariwise, there's no way I'd ever voluntarily go back to re-typing an entire chapter just because I made some significant revisions. Nowadays I can just write, and when the time comes to submit something, I can just push a button and let the computer print (or more likely send) the words.
Serious enterprises who care about their hot dogs use Oracle, since you can roast over 10,000 dogs at once and optionally impart the taste of filet mignon; Yeah, but only if your locale is set to US-ASCII. If you try to use UTF-8, taste defaults to pate de fois gras. However, under US-ASCII "SELECT beer FROM fridge" will return only 'Budweiser' or 'Miller Lite'...
Exactly what I was going to say. I have separate folders for every project, and rules that automatically route 90%+ of my messages appropriately. Then it's simply a matter of going down the priority list and taking one project at a time. I find that this really helps me to focus on one project at a time, since I'm not just reacting to the latest message that pops up in my inbox. (I also have my new email notification turned off.)
I also have a "reaper" that goes through every weekend and archives everything I have tagged. Anything that's not tagged gets auto-archived to the bit bucket after a month.
Huh. Then it must have popped up with the "Reboot and screw me over" button in the same position as the "Save the document I've been working on all morning" button, since that dialog had just popped up and my pointer had warped to it.
I'd still be much happier with a taskbar bubble (which, now that I mention it, I think I've seen).
the darn thing pops up right when I'm about to press enter or click That's why I finally gave up on "warping" (automatically setting the mouse pointer over the default button whenever a dialog box pops up). Trying to get some work done, and the "Windows has just installed a Very Important Update" dialog pops up just after I got a "Save changes?" dialog. So not only does the machine reboot, but it sends SIGHUP (or whatever the Windows equivalent is) to all the open apps. Which of course my app interprets as "cancel save and then exit". So now not only and I stuck waiting for my machine to boot, but I've lost my work.
Now, not only do I not use pointer warping, but I have automatic updates disabled too. If my company wasn't so firmly in Microsoft's harem, I'd switch to something else...
I do wonder how many people will go randomly clicking around to see what it does, click through the UAC dialogue, and end up doing something like removing permission to access the C: drive for everyone but their pet dog... Oh, c'mon, nobody's that dumb. Well, except maybe this guy...
Seriously, the only things speech recognition is good for are bulk text entry and simple navigation. I imagine trying to use voice commands to operate modern software would be similar to letting my four-year-old help make pancakes — yes, it gets done, but it's so much easier and faster to just do it yourself. Imagine trying to edit a document using just voice commands. Is your WP going to be smart enough you can tell it "find all occurrences of 'scum-sucking bottom feeders' and replace it with 'esteemed colleagues'". Or are you going to have to say "Find. Scum hyphen sucking bottom feeders. Tab. Esteemed colleagues. Replace all." Face it, GUIs have rendered speech recognition for command and navigation moot. Most operations you perform don't have a verbal description, or at least not one that is quicker to say than to do.
I also can't imagine it'd be that useful for actually writing things. I don't think I'm the only one who revises as they write. I think I actually write better when I write things out by hand, because it's slower so I tend to think my phrasing and sentence structure through more before I commit anything to paper. If I could suddenly type two or three times faster, I think it'd probably make my text even more incomprehensible than it usually is...
It's remotely exploitable, if the programmer is dumb enough. Then again, so is Apache + PHP. Doesn't PHP stand for Pretty Hopeless Privacy? I remember it used to be pretty trivial to do SQL injection attacks against a pretty wide spectrum of PHP sites back in the dot-bomb days. Hopefully it's gotten better as security has gotten more press, but even if it's gotten twice as good as it was, that's still pretty bad...
Of course, that would eliminate the use of a UNIVERSAL RESOURCE LOCATION, since it would no longer be centralized. Nope, it just means that your torrent tracker would have to have a way to resolve the reference. Whether something like DNS where you have specific "go-to" hosts, or whether you just ask every host you're connected with, or something else (maybe a kind of dynamic mesh with ad-hoc gateways), the choice is up to you.
Maybe something like NTP, where you have the strata-1 time servers, and then the designated strata-2 servers, and everyone is encouraged to set up a strata-3 server for their own subnet. This way nobody's really dependent on anyone else if they don't want to be. Once this gets set up, maybe you could even have a dtd: protocol that specifies how to find a server, and how to cache DTDs once you get them (and how to expire or occlude them when a new version comes out).
Every 30 years, or so, everything old is new again And vice versa. The first time I heard about commercially-available SSDs (solid-state disks), they were being used to replace the aging HDs on industrial PDP-11 systems. The SSDs were more reliable than the RK05/RP07s they replaced and used less power, but it just seemed so wrong to have these SOTA[0] drives hooked up to these ancient machines. Nothing against PDP-11s, they're great, but why does a machine with a cycle time measured in milliseconds need a disk with a couple hundred MB/sec of bandwidth?
He's probably writing from the US, where some people actually consider a car like a Ford Mustang to be a sports car. Y'know, I actually hadn't thought about the Mustang. Then I checked out this link and was shocked, shocked I tell you, to find out that this sort of travesty would still be played out today!
They even put leaf springs (you know, like on horse-drawn carts) on Corvettes I understand this is required if they want to continue to use a pushrod engine. Some antique auto solidarity thing...
If unsprung weight really mattered, we'd mount the brakes inboard. The Hummer H1 and HMMWV are about the only vehicles to bother with this. Actually, Jaguar was doing this years ago, and Jaguar differentials with the brakes were highly sought-after components for people building kit cars (and others just improving their existing cars). The problem is getting enough air up there for cooling. It's bad enough when the brakes are sitting inside the wheel, now imagine they're tucked up out of the way next to a hot differential. It's not a problem for Hummers, as they don't tend to venture out on to the race track very often, but it's a big deal if you regularly turn in a few hot laps. Unsprung weight is also a big issue with those types of people, and I don't think there's been a sports car with an unsprung differential since the mid-70s.
Electric motors make the most torque at zero RPMs for much better load/towing. But can they make an electric motor with (say) 350-400 lb/ft of torque that you could reasonably expect to use in a truck? I'm sure there are motors that produce that much torque, but are they small/light enough to be used in a vehicle? And are they variable-speed, or synchronous? CVT's aren't currently durable enough to be used behind high-torque motors, and variable-frequency motor controllers are complex and noisy.
once they make the switch to independently powered wheels (an electric motor built into the wheel) you could have much more interesting steering suspension options since there would be no drive shafts getting in the way. Dunno, I'd much rather have a driveshaft "getting in the way" than to have a hundred pounds or so of unsprung weight to try to control. A lot of the space you seem to think you're going to save is going to get eaten up by the shock absorbers.
The other main disadvantage was that having procedures within procedures also made for monolithic programming, requiring that code has to be completely rewritten in order for basic concepts to be used with a new programming language such as C or C++ (lists, trees, hash tables, pointers, etc...) Hmmm, I don't think I've ever worked with code that didn't need to be completely rewritten to be used with a new language...;)
Seriously, though, nested scoping is one of the things I missed most about Pascal when I moved to C. It encouraged (me, anyway) to write more "helper" routines, rather than monolithic code. You could actually use it to provide a sort of polymorphism by allowing a routine to call named functions to obtain its arguments or manipulate variables, and then ensuring that the procedures that called said routine provided those functions. So the functions you actually called were determined dynamically by searching the call stack, instead of being statically linked.
Of course, after I discovered function pointers, there was no going back to Pascal, but I did enjoy some of its higher-end features (part of its ALGOL heritage).
cool, dense air has more oxygen, and so the computer has to increase the fuel usage to maintain the correct mixture. But then aren't you making more power? So wouldn't you be driving at a slightly lower throttle setting, thereby offsetting the increase in mixture? The whole point of forced induction is to get more air into the cylinder so you can mix in more fuel and make more power, how is this really different, apart from the scale? I'd be more inclined to think it's the fuel formulation (more ethanol, which has less energy per unit than gasoline).
You wouldn't even have to do that. Just replace the drive controller. I can't imagine they're going to get custom-built drives, so they're going to be a standard X gig drive with a custom controller. Then all you need to do is go out and get an OEM controller (just buy a standard version of the same drive) and replace the TiVo controller. Granted, it's probably not this simple (the data on the drive is likely encrypted too), but the patent doesn't say anything about encrypting data other than the challenge/response, so maybe it will be.
What useful content is going to be rendered while the applet loads? Um, the rest of the page? You know, kind of like when Flash ads or graphics from other sites are loaded now? The browser (optionally) sticks in a placeholder of the appropriate size for the applet, then renders the rest of the page around it. Unless the applet is the only thing on the page, you'll be able to at least read the text while the applet loads and starts.
It's easier to pick up meaning that you'd miss in written correspondence, or even phone conversation. You're more likely to ask a good question and the back and forth nature of conversation makes it more likely that ambiguities will be cleared up. Amen! It's very handy to be able to note when someone in sales/management/marketing's eyes start to glaze over when you're explaining some technical detail. Then you can backtrack and re-explain so everyone understands, w/o having to respond to somebody's "I didn't understand your comment about X" message two or three days (or weeks!) later. It's also handy when your bogometer goes off and you can glance over at someone else on your team and see if they're likewise wary. I also find they're much quicker than e-mail if you just need to clear up some points on one or two topics.
But if the machine hosting the VM fails, how do you fail over all the virtual machines? Assuming you're working with some kind of distributed app, anyway. Kind of hard to send a "retry" message to a machine that just evaporated. Hmmm, makes me wonder if we'll see some kind of duplex hardware designed just for this kind of environment (Tandem used to make that sort of stuff, I wonder if HP will roll some out).
Interesting idea, though. Your intra-machine bandwidth would be (theoretically) limited only by your memory cycle times, and you'd have a nearly-infinite supply of "hot spares" in case one of your virtual machines crashed.
Slashdot Mirror
I don't mind having to double-scan sentences in Slashdot posts, and I usually don't care if I run across something similar on other tech sites. But this is an Intel white paper, and I expect that things like that have been gone over by an editor, not just C&P from some engineer's e-mail. I'm also not faulting whoever wrote it; the content was fine, it just needed to be cleaned up (which is not the engineer's job). The fact that it wasn't is what is annoying me.
But, uh....ummmmm.....
*sigh* I am such a dickhead...
Geekoid, I apologize. Your comment does make sense, once I pull my head our far enough that I can tell what part of the thread you're on. I owe you a $BEVERAGE. Feel free to flame me back, I deserve it.
And thanks, stuktongue, I obviously shouldn't be counted on to figure this kind of stuff out on my own...
(Damnit! I hate being a fucktard!)
Probably the source of the less than optimal text. The point is, whatever tortured, twisted prose was submitted should have been edited and polished before going out with an Intel logo on it. This was a white paper on the corporate web site, not a post on some random Intel engineer's blog — different standards apply.
Seriously, check out this opening paragraph from the Intel® 64 and IA-32 Architectures Application Note:
TLBs, Paging-Structure Caches, and Their Invalidation The Intel® 64 and IA-32 architectures may accelerate the address-translation process by
caching on the processor data from the structures in memory that control that process.
Because the processor does not ensure that the data that it caches are always consistent
with the structures in memory, it is important for software developers to understand how
and when the processor may cache such data. They should also understand what actions
software can take to remove cached data that may be inconsistent and when it should do
so. The purpose of this application note is to provide software developers information about
the relevant processor operation. This application note does not comprehend task switches
and VMX transitions. Notice how they even get the fact that "data" is plural right? That's the kind of documentation I'm talking about.
I was looking at the Thread Building Blocks paper, and it reads like it was somebody's hastily-scribbled draft:
"The Intel Threading Tools automatically finds correctness and performance issues" (The tools finds?)
"Along with sufficient task scheduler and generic parallel patterns" (Who has insufficient task scheduler?)
"automatic debugger of threaded programs which detects many of thread-correctness issues such as data-races, dead-locks, threads stalls" (Sarcasm fails me...)
And that's just in the first few paragraphs, I haven't even gotten to the real meat of the article!
I'm used to informative, well-written and reasonably complete technical documentation from Intel — WTF is this?
Not sure if you're really making a point by juxtaposing "novice programmer" and "multiple languages", I think you'd wind up just wiping asploded head off the walls if you wrote your forms in VB and your back-end classes in C# and expected a novice to make sense of it all.
Seriously, VS isn't bad (although the Express versions only support one language at a time), but it's hardly the only IDE you can be productive in. Check out a 4GL sometime if you want to see some serious RAD...
Personally, I can (and do) work either way, but if you're doing major revisions you can't beat the ability to lay out an entire chapter on a table (or bed, if your table's like mine) and have it all available without having to scroll around. Contrariwise, there's no way I'd ever voluntarily go back to re-typing an entire chapter just because I made some significant revisions. Nowadays I can just write, and when the time comes to submit something, I can just push a button and let the computer print (or more likely send) the words.
Exactly what I was going to say. I have separate folders for every project, and rules that automatically route 90%+ of my messages appropriately. Then it's simply a matter of going down the priority list and taking one project at a time. I find that this really helps me to focus on one project at a time, since I'm not just reacting to the latest message that pops up in my inbox. (I also have my new email notification turned off.)
I also have a "reaper" that goes through every weekend and archives everything I have tagged. Anything that's not tagged gets auto-archived to the bit bucket after a month.
Huh. Then it must have popped up with the "Reboot and screw me over" button in the same position as the "Save the document I've been working on all morning" button, since that dialog had just popped up and my pointer had warped to it.
I'd still be much happier with a taskbar bubble (which, now that I mention it, I think I've seen).
Now, not only do I not use pointer warping, but I have automatic updates disabled too. If my company wasn't so firmly in Microsoft's harem, I'd switch to something else...
Seriously, the only things speech recognition is good for are bulk text entry and simple navigation. I imagine trying to use voice commands to operate modern software would be similar to letting my four-year-old help make pancakes — yes, it gets done, but it's so much easier and faster to just do it yourself. Imagine trying to edit a document using just voice commands. Is your WP going to be smart enough you can tell it "find all occurrences of 'scum-sucking bottom feeders' and replace it with 'esteemed colleagues'". Or are you going to have to say "Find. Scum hyphen sucking bottom feeders. Tab. Esteemed colleagues. Replace all." Face it, GUIs have rendered speech recognition for command and navigation moot. Most operations you perform don't have a verbal description, or at least not one that is quicker to say than to do.
I also can't imagine it'd be that useful for actually writing things. I don't think I'm the only one who revises as they write. I think I actually write better when I write things out by hand, because it's slower so I tend to think my phrasing and sentence structure through more before I commit anything to paper. If I could suddenly type two or three times faster, I think it'd probably make my text even more incomprehensible than it usually is...
Maybe something like NTP, where you have the strata-1 time servers, and then the designated strata-2 servers, and everyone is encouraged to set up a strata-3 server for their own subnet. This way nobody's really dependent on anyone else if they don't want to be. Once this gets set up, maybe you could even have a dtd: protocol that specifies how to find a server, and how to cache DTDs once you get them (and how to expire or occlude them when a new version comes out).
Seriously, though, nested scoping is one of the things I missed most about Pascal when I moved to C. It encouraged (me, anyway) to write more "helper" routines, rather than monolithic code. You could actually use it to provide a sort of polymorphism by allowing a routine to call named functions to obtain its arguments or manipulate variables, and then ensuring that the procedures that called said routine provided those functions. So the functions you actually called were determined dynamically by searching the call stack, instead of being statically linked.
Of course, after I discovered function pointers, there was no going back to Pascal, but I did enjoy some of its higher-end features (part of its ALGOL heritage).
You wouldn't even have to do that. Just replace the drive controller. I can't imagine they're going to get custom-built drives, so they're going to be a standard X gig drive with a custom controller. Then all you need to do is go out and get an OEM controller (just buy a standard version of the same drive) and replace the TiVo controller. Granted, it's probably not this simple (the data on the drive is likely encrypted too), but the patent doesn't say anything about encrypting data other than the challenge/response, so maybe it will be.
But if the machine hosting the VM fails, how do you fail over all the virtual machines? Assuming you're working with some kind of distributed app, anyway. Kind of hard to send a "retry" message to a machine that just evaporated. Hmmm, makes me wonder if we'll see some kind of duplex hardware designed just for this kind of environment (Tandem used to make that sort of stuff, I wonder if HP will roll some out).
Interesting idea, though. Your intra-machine bandwidth would be (theoretically) limited only by your memory cycle times, and you'd have a nearly-infinite supply of "hot spares" in case one of your virtual machines crashed.