Remember how parallel ATA was replaced by serial ATA? Despite fewer wires, it can handle more data, because it's easier to push a serial protocol at a very high clock rate than to get a bunch of wires to synchronize perfectly at a high clock rate. And crosstalk between signal wires is a serious issue; check a parallel ATA cable sometime and notice how many ground wires it has. (To use the fastest parallel ATA modes, you must use an 80-wire cable, and over half of those 80 wires are ground wires, just to guard against crosstalk.)
So I found it surprising that HDMI was a parallel cable spec! And I do not find it surprising at all that this new standard will be a very high clock rate serial protocol over standard Ethernet cabling.
Note that this came out of industry, and not out of an ivory-tower standards group.
At work I have a Windows PC, and I was always frustrated by the very poor performance of Flash video. The video would freeze, then unfreeze over a second later with the video frames in between just dropped. (When you are watching a 5 second film this problem makes the movie almost unwatchable!) And it's a quad-core AMD Phenom II system. It should be fast.
So now, I'm trying out 3.6.4 and the difference is stunning. Now the Flash video playback is perfectly smooth.
I still want WebM in HTML5 instead of Flash, but what the heck, this is working now and I'm happy about it.
Sample size of one, YMMV, etc. But I'm happy about it.
Would you believe me if I claimed that an algorithm that has been on the books as "optimal" for 46 years, which has been analyzed in excruciating detail by geniuses like Knuth and taught in all computer science courses in the world, can be optimized to run 10 times faster?
It would of course be unjust and unreasonable to blame Williams for not realizing that Atlas had invalidated one of the tacit assumptions of his algorithm: only hindsight makes that observation possible. The fact is, however, 46 years later most CS-educated professionals still ignore VM as a matter of routine. This is an embarrassment for CS as a discipline and profession, not to mention wasting enormous amounts of hardware and electricity.
I can tell you that they never covered any of this stuff when I was in college. I learned about it by reading articles like this one.
P.S. He never claimed that Knuth was "wrong". In fact, when the data set fits purely in RAM and your server isn't swapping, the algorithm's original assumptions are valid and it is in fact fastest. (And there are numbers in TFA explaining exactly this!) But he is absolutely correct that in real-world applications, swapping matters a lot. His algorithm isn't much slower in the non-swapping case, and can be 10x faster or more in the swapping case.
He did use phrases like "You're doing it wrong", which will drive some readers nuts. But I view it as just his way of making the article more interesting. This material could have been very dry and boring; as written, it's a fun article that still communicates the material well.
The article is interesting, and worth reading, but if you RTFS without RTFA you'll be dumber than you were before. Thanks, kdawson.
Not so, friend! I talked a bunch about Apple's use of "undocumented" as an excuse to deny all non-Apple software the ability to use APIs, and I made it very clear that I was using the word "undocumented" the way Apple is using it. I even put it into those scare-quotes to make sure it was clear I was using a specific meaning.
There is a difference between "undocumented", as in nobody wrote documentation, and "Apple Computer undocumented", which means that if you use the API, they send trained ninja cats with laser claws to assassinate you in your sleep. And they remove your apps from the App Store, which is even worse.
I failed to see instances of paid Apple apps in your answer ?
So, this is your standard then: as long as Apple is not stepping on a large company, it is okay for them to do this? All the small companies and individuals who developed now-rejected apps can take solace in the fact that Apple doesn't charge extra for any of the affected Apple software?
Was it also okay when Microsoft "cut off Netscape's air supply" by offering Internet Explorer for free? IE wasn't a "paid Microsoft app" so according to your standard Microsoft must not have been evil.
I don't actually know whether Apple's paid apps use "undocumented" APIs or not. And I don't really care, because I think this whole "paid apps" standard is disingenuous.
THat's in direct contrast to MS, who WAS using private APIs to SELL more Office.
Did you miss the part in this post where I said I worked at Microsoft, in the early 90's, on Microsoft Word for Windows, and I never saw any secret private APIs?
I'm going to assume you missed that, and didn't mean to imply that I'm lying about my direct personal experience.
So, I've got direct personal experience. What evidence have you got to the contrary?
P.S. As I keep saying, give me Linux rather than Windows or Mac. Microsoft has done some bad things, like buying an ISO standard for their XML file format. I just think it's stupid to bash Microsoft over false rumors, while at the same time giving Apple a free pass to actually do what Microsoft was accused of doing. Bash Microsoft over the things they actually did, please, and don't have one standard for Microsoft and a different standard for Apple.
I'm not sure if you are joking, you didn't understand me, or you are disingenuous. So I'm going to give you three answers.
If you were joking: sorry dude. I do have a sense of humor, honest.
If you didn't understand me: then I will simply point out that Apple controls app distribution for iPod Touch, iPhone and iPad; and if you use any APIs undocumented by Apple, your app will be banned from distribution. So, in this context, "undocumented" carries a special meaning: something that has not been officially documented and blessed for use by Apple, and apps use undocumented APIs at their peril. There is no such thing as a Linux API that we are all forbidden to use because some entity is trying to keep that API secret, and especially there is no entity that can insert APIs into Linux and use them exclusively while simultaneously forbidding us to use them.
If you were disingenuous: then you already knew the explanation above, and just were pretending not to, in which case I am sad. (I have met Apple fanboys who are this zealous in their fanatic defense of Apple.)
To advantage which of it own Apps does Apple use its OS advantage ?
On the iPad, only Apple software can multitask (this article has a list: email client, SMS text client, and other apps). On any of their platforms, only Apple software may use the APIs that let you customize the way the UI widgets display. Only Apple software can use the full functionality of the accelerometer. Here is a blog post discussing some undocumented OS X features that made Safari much faster than Firefox 3. And here is a blog post discussing how several apps were rejected for using undocumented functionality. And here is a whole article discussing undocumented Apple APIs, with examples of cool stuff that only Apple's own software is allowed to do. And here is an article discussing cool things that Safari can do, that Firefox isn't allowed to do. And here is a column that claims that Apple inserts undocumented APIs and uses them in its own code for years, without ever documenting them (but presumably without breaking them because it would break Apple's own code). Even the APIs for the WiFi are undocumented.
I understand the argument that Apple doesn't want to commit to supporting these APIs forever, like Microsoft has had to do with even obscure APIs in Windows. If you use these undocumented APIs to do cool things, and Apple revises the OS, your app may break. And Apple doesn't want the customer to think it's Apple's fault that your app broke.
But I also understand the argument that some of these APIs allow for really cool stuff, which is currently reserved only for Apple. People don't like this.
As for me, give me Linux anyway. No such thing as an "undocumented" API, and there is no entity that has an unfair advantage over everyone else, and I can install any software I want.
Didn't Micro$oft have API's that they used and didn't want anyone else to use? Didn't they get lambasted for that?
Oh yeah. I worked at Microsoft in the early 90s, and I even worked on one of the flagship applications (Microsoft Word for Windows). I never saw any "secret backdoor" APIs, and I firmly believe that those rumors were wildly overblown.
But Apple is actually doing it. They have undocumented APIs that they won't let anyone else use. Even on the Mac they have them, and they have been known to break the undocumented APIs seemingly just to burn apps that dare to use them. Now that they have the "app store" it's even more up front: if Apple figures out that you are using their double-secret APIs, they will reject your app.
Microsoft was widely flamed over rumors of doing this, while few people care that Apple actually does it a lot.
I want you to suffer for writing pretentious, angsty things like "Why do you want me to suffer?"
As for the programming stuff, if ease of grepping is the most important thing to you, then go ahead and stick with C. The global namespace will force you to write globally unique function names, which are perfectly greppable.
I haven't worked on a large C++ project but I have worked on a large Python project, and I can tell you that I didn't "suffer" due to the lack of globally unique function names.
How often do you need to grep for every reference to a function? And why do you need to do it? I work, all by myself, on a rather large code base, and I can't remember the last time I needed to do this.
P.S. I'm a command-line fiend, and my favorite development "environment" is vim plus ctags. Nonetheless, I do have access to things like Eclipse, and if I ever need to "refactor" code (example: rename a function, both the definition of it and all calls) I am willing to fire up Eclipse and use the GUI refactor tool. And then, I rebuild my ctags file and go back to using vim to do my main editing work.
Spend less time yelling at the kids to get off your lawn, and spend more time looking at the available tools.
The GCC guys are not going crazy here. They are discussing what subset of C++ to allow.
If you use all the wild features of C++, the results could be scary. For example, operator overloading is great if used judiciously, but if used badly it can make the code a mess. And if it is used at all, then it means that you can't look at one page from a printout and know for sure what that code does; you need to look at all the class functions to make sure there aren't tricky overloaded operators.
I use plain C all the time at work, and the top C++ feature they should be using is simply the object-oriented class stuff. With a single global namespace you need to make functions like MyClassAddFloatAndInt(), but in C++ you could just call that function add(); it would be part of MyClass, and if you have other "add" functions with other type signatures, they won't collide. They could go from: {
MyClass m;
MyClassInitialize(&m, foo, bar);
MyClassAddFloatAndInt(&m, 3.0f, 2);
MyClassDoSomething(&m);
MyClassCleanup(&m); }
to: {
MyClass m(foo, bar);
m.add(3.0f, 2);
m.do_something(); }
Even better if they allow the use of C++ namespaces to keep a large project organized.
The other major win that comes to mind is simply being able to use powerful C++ libraries like the STL. Not having to cook up some kind of container data structure in plain C, but being able to use std::vector<SomeType> and std::map<SomeType, OtherType> and such is a huge win.
P.S. I read through much of the discussion and here was my favorite post:
As Chad Birch pointed out, the JavaScript thing was an April Fool's joke and I failed to notice. That's what I get for posting in a hurry. Thank you for the correction, Chad Birch.
I agree with the other comments: if he doesn't have the interest, or if he doesn't have the aptitude, then trying to push him into coding is a waste of time.
That said: check out PyGame. PyGame is a set of libraries for Python, specifically intended for creating new games.
Hmmm. I just went there, and it says that PyGame has now been ported to JavaScript. That probably makes sense, given the major efforts to speed up JavaScript in the new-generation web browsers.
At the PyGame web site, there are a bunch of games people have written, with source code available; and some of these games are half-done and half-broken. If he has the inclination to code, he might get interested in a half-done game and start fixing it up. Or even take a game that isn't half-baked, and start adding new features to it.
A coming technology that is still in very early days is a desktop fabber, also called a 3D printer.
There is an open source project to make useful fabbers. Current fabbers are designed to use easily-acquired parts. I look forward to the day when someone makes a fabber that can fabricate all the parts needed to build another fabber, but that day is distant.
You could try inventing science and/or medical sensors that would be very inexpensive to mass-produce, and share the designs with developing countries.
Medical devices are super-expensive, partly because in developed countries you have to get government approval which costs huge money. If you could produce, say, an EEG system that just plugs by USB into a cheap computer/netbook/whatever, and share the specs with the world, you might do some good.
My dad was in the hospital recently, and he had a half-dozen medical sensors connected; his pulse rate, blood pressure, respiration rate, and several other things were monitored by computer. If any reading went seriously out of spec, an alarm would sound. Could you make something like this that would be cheap enough for developing nations to build and use it? Maybe even make it WiFi so you can just set it next to each patient's bed, and one nurse with one computer could monitor all patients for alarms going off.
By the way, the sensor for pulse and blood oxygen level is a Pulse Oximeter.
You might want to research whether there is already a project to make open source/open hardware stuff for developing countries; no sense in duplicating something that someone might already be doing.
I just did a Google search, and USB oscilloscopes exist and are surprisingly cheap. I wonder if you can adapt one for medical uses?
If you are not already a software developer, and you want to write a demo app to show what your equipment can do, I suggest Python for the language on the host PC.
Thank you for the correction. Even as a dilettante, I should have figured that Avast wouldn't be free for a mass rollout at a school.
Also, I meant to say: Windows 7 would be a much better choice than XP, because I think Microsoft really has improved security versus the really old XP code base.
First of all, my credentials: I'm not a professional sysadmin, just a professional software developer who also admins a few systems. You may want to give more weight to opinions posted by actual school sysadmins.
Okay, my recommendation: Ubuntu. And my dis-recommendation: you don't want XP.
With Ubuntu, if the computers are reasonably fast, you can do a full re-install in a short amount of time. I imagine there will be cases where someone manages to monkey around with a computer and mess it up, and ease and speed of re-install will be a win for Ubuntu. With Ubuntu you don't need to install, reboot, install a driver, reboot, install another driver, etc. Note that if you know what you are doing as a Windows sysadmin, I believe you can make a "slipstreamed" CD image with the drivers and such pre-installed, which would mitigate this a lot.
Other Linux distros would also work well, but Ubuntu has the momentum as the free home-user Linux distro of choice. Some of the high-school kids will possibly already be running Ubuntu at home. I would suggest having all students login as "guest" and have the "guest" account set up so that, when the user logs out, the/home/guest directory is removed and then replaced with a copy of a standard/home/guest directory. If some clever little black-hat wannabe edits.profile or something to try to set up a joke on the next user, this would sort that out.
With XP, you can configure the logins to be non-Administrator, and unless you are totally insane you will do so. It's hard enough to keep the computers virus-free in any event, and if you let high school kids have privileges to install software, you are just asking for trouble. But you will absolutely want anti-virus, and that means you will want to keep the anti-virus up to date, and that is a big headache that you can completely avoid with Ubuntu.
Now, as for software. Everything I'm discussing below is available on both Linux and Windows. On Ubuntu, it's trivial to install these; on Windows it would be more work. (But again you could probably work something out with a "ghost" disk image or some such.)
Depending on what kind of engineering we are talking about, many of these students may move on to using Matlab in their working lives. Matlab is expensive, but educational copies are probably available. But there is also the completely free GNU Octave, which is basically a Matlab-alike, and is freely available on Ubuntu. There is even a GUI wrapper for Octave, written for the KDE environment, but I haven't tested that. I have only used Octave for DSP work, but it is adequate for that, and did I mention that it's free. The graphing tools in Octave are not as good as real Matlab but they can get the job done.
I have hopes that someday Matlab will fall by the wayside, and Python will replace it. Specifically, Python with the NumPy and SciPy extensions. With SciPy, Python can do much of what Matlab can do, and it does it with a much better base language; Python is a marvelously clean and tidy language, while the Matlab language is just annoying. Let's face it, it will be many years if ever before Matlab could be replaced by Python, but there are science labs and engineering groups out there using Python and the numbers will only grow. The graphing tools in SciPy are IMHO better than the ones in GNU Octave, and almost as good as the ones in Matlab. You can get Python with all the SciPy stuff pre-installed as the project Python(x,y):
There is also a project to take every bit of free math software and glue it all together into an amazing giant math tool: this is called Sage. Sage, also, is based on Python. You run Sage as a server and you use any standard web browser as the GUI client. You can run Sage on the same computer as the web browser, or you can set up one
They are taking an expensive container unit, and a very expensive solar array, and buying ten thin clients, and the main computer is a Pentium 4?
That's just crazy. All I can think is that they are a computer recycling outfit, and they had a Pentium 4 on hand and just said "Eh, good enough."
I would have spent a few hundred and gotten an AMD dual-core or even quad-core chip and some ECC RAM. And probably a flash boot drive. You want the computer to be as bulletproof as possible, and it would be nice if it was energy-efficient. A Pentium 4 is basically a device for heating a small room... in other words, not power-efficient. It has a single core. I don't believe any Pentium 4 chipset supports ECC RAM. (Hence the AMD suggestion. Yes, I am an AMD fan, but Intel reserves ECC as a feature of only their expensive server chipsets; whereas ECC has been a standard feature with AMD for years now.)
If you are sharing one computer among ten users, extra cores would be a very good thing. Large cache would be a very good thing. Both of those argue for a modern CPU.
The article didn't say what the OS is. However, the article has a link to another article about the first such cyber cafe, and that previous article says they are using NComputing thin client technology. NComputing appears to be a solution for hooking up external thin clients to virtual machine images running on a host. Which begs the question: ten virtual machine images running on a Pentium 4?!? Let's hope they are least are using one of the models of Pentium 4 that supports the Intel virtualization "VT-x" instructions.
Oh well, I'm sure it works, and it's a heck of a lot better than nothing.
There are various independent sources of this temperature data, of which CRU produced one. All the various independent sources are in very close agreement. What are you proposing to "start over" on?
I would like to see data sets with complete paper trails attached, and all source codes to all processing provided. I would like to see independent people like the "Climate Audit" guy be able to completely reproduce all steps in data processing without needing to file any FOIA requests. This shouldn't be controversial; if the AGW proponents want the governments of the world to suppress whole sectors of the economy, with costs in the trillions of dollars, then the AGW proponents need to provide absolutely iron-clad and maximally transparent data.
Luckily for me, it looks like this may be happening:
You say that all the independent data sets agree. Can you point me to an independent data set where there is a full paper trail attached to all the data, and source codes are provided for all processing of the data? Bonus points if this data set was actually used in an IPCC report or some similar official report.
I agree about keeping your mouth shut and your ears open. I've never worked in a truly bad shark tank sort of environment (although it sounds like you have) but it's always good to keep your mouth shut and your ears open.
There is an old business advice book called Bravely, Bravely in Business that contains some all-around good advice. There is one rule in that book that is very very hard to follow, but a good idea: "Never say anything about anyone that you wouldn't say exactly the same way to their face." Think about what would happen if you say something like "Joe is a total idiot who can never get anything done!" or worse, and someone overhears and tells Joe exactly what you said (or mis-remembers it as something even worse). Or if Joe overhears.
Don't hesitate to say something nice when you honestly feel it is deserved. Do hesitate to say something negative, even if you honestly feel it is deserved. "If you can't say something nice, say nothing at all." That's trite but it's good advice.
Like others, I disagree on the "always eat lunch alone" thing. It's good to keep some separation between work and the rest of your life; your boss doesn't need to be your best friend. But you want to learn to get along with your co-workers.
At my second job ever, all the software developers used to go to lunch together, and we would often chat about problems we were working on. It was a fun, informal way to share ideas and information; insisting on eating alone would not have been a good thing in that situation. (Later, management at that company forced some changes in the cafeteria, and as a direct result all the software developers stopped eating together and started going out to lunch. The camaraderie was lost. It was a shame.)
P.S. Looks like you can get Bravely, Bravely in Business pretty cheaply right now. You might want to order a copy and read it; it's good. (Read the customer reviews on Amazon.)
If it is not really about control, then why did Apple say that programs must be written by hand in Objective C?
If a Flash translator can output Objective C code, which is then compiled down to native code as per usual; and if the resulting app conforms in all ways to the Human Interface Guidelines and other requirements; and if that app runs efficiently and provides an excellent user experience... that app is still forbidden. Why, exactly?
Some station data held by the CRU was not made available publicly because it is the intellectual property of some national meteorological services around the world and subject to non disclosure agreements.
It is much worse than that. The CRU guys would not even give a list of exactly which data sets they were using. Some of the guys trying to check the numbers were willing to go around and sign NDAs to get copies of the numbers, and the CRU guys wouldn't even tell them which number sets to ask for. The CRU guys claimed that they were under non-disclosure agreements that forbade them from even identifying which stations' data they were using.
So, the CRU guys want to impact the global economy to the tune of many trillions of dollars, which is an abstract way of saying they want to inflict misery on millions of people. They claim it is a necessary evil, needed to prevent horrible misery down the line. But they also ask us to just trust them, and not review the data at all, which is just plain unacceptable.
The Large Hadron Collider guys aren't handing out copies of their data, but they also aren't asking the world to impose Draconian controls on the economy. The global warming people are demanding extraordinary interventions into the economy, and they must have an extraordinarily high level of disclosure. They haven't met this.
The CRU should have had all the data sitting on a web server, and all the source code they used to massage that data and make graphs. They not only didn't do that, they didn't even have the original data (they claimed it was "lost"). I am stunned at the level of confidence you have in these guys.
If there really are as many data sets available as you are claiming, then some climate scientists should start over and do the whole research project again, this time with full transparency. If they reproduce the results, then maybe we can talk about Cap and Trade.
Also, I think I was remiss in not mentioning that the Tegra 2 is only available in small (prototype) quantities right now. But I believe the first Tegra 2 devices will ship within the next 3 or 4 months.
Slashdot Mirror
Remember how parallel ATA was replaced by serial ATA? Despite fewer wires, it can handle more data, because it's easier to push a serial protocol at a very high clock rate than to get a bunch of wires to synchronize perfectly at a high clock rate. And crosstalk between signal wires is a serious issue; check a parallel ATA cable sometime and notice how many ground wires it has. (To use the fastest parallel ATA modes, you must use an 80-wire cable, and over half of those 80 wires are ground wires, just to guard against crosstalk.)
So I found it surprising that HDMI was a parallel cable spec! And I do not find it surprising at all that this new standard will be a very high clock rate serial protocol over standard Ethernet cabling.
Note that this came out of industry, and not out of an ivory-tower standards group.
steveha
Coral Cache:
http://www.goodgearguide.com.au.nyud.net/article/351651/12_most_interesting_unusual_useful_linux_distros/
List of the distros:
steveha
At work I have a Windows PC, and I was always frustrated by the very poor performance of Flash video. The video would freeze, then unfreeze over a second later with the video frames in between just dropped. (When you are watching a 5 second film this problem makes the movie almost unwatchable!) And it's a quad-core AMD Phenom II system. It should be fast.
So now, I'm trying out 3.6.4 and the difference is stunning. Now the Flash video playback is perfectly smooth.
I still want WebM in HTML5 instead of Flash, but what the heck, this is working now and I'm happy about it.
Sample size of one, YMMV, etc. But I'm happy about it.
steveha
My favorite quotes from the article:
I can tell you that they never covered any of this stuff when I was in college. I learned about it by reading articles like this one.
P.S. He never claimed that Knuth was "wrong". In fact, when the data set fits purely in RAM and your server isn't swapping, the algorithm's original assumptions are valid and it is in fact fastest. (And there are numbers in TFA explaining exactly this!) But he is absolutely correct that in real-world applications, swapping matters a lot. His algorithm isn't much slower in the non-swapping case, and can be 10x faster or more in the swapping case.
He did use phrases like "You're doing it wrong", which will drive some readers nuts. But I view it as just his way of making the article more interesting. This material could have been very dry and boring; as written, it's a fun article that still communicates the material well.
The article is interesting, and worth reading, but if you RTFS without RTFA you'll be dumber than you were before. Thanks, kdawson.
I agree on all points.
steveha
Just pointing out the fallacy of your claim
Not so, friend! I talked a bunch about Apple's use of "undocumented" as an excuse to deny all non-Apple software the ability to use APIs, and I made it very clear that I was using the word "undocumented" the way Apple is using it. I even put it into those scare-quotes to make sure it was clear I was using a specific meaning.
There is a difference between "undocumented", as in nobody wrote documentation, and "Apple Computer undocumented", which means that if you use the API, they send trained ninja cats with laser claws to assassinate you in your sleep. And they remove your apps from the App Store, which is even worse.
steveha
I failed to see instances of paid Apple apps in your answer ?
So, this is your standard then: as long as Apple is not stepping on a large company, it is okay for them to do this? All the small companies and individuals who developed now-rejected apps can take solace in the fact that Apple doesn't charge extra for any of the affected Apple software?
Was it also okay when Microsoft "cut off Netscape's air supply" by offering Internet Explorer for free? IE wasn't a "paid Microsoft app" so according to your standard Microsoft must not have been evil.
I don't actually know whether Apple's paid apps use "undocumented" APIs or not. And I don't really care, because I think this whole "paid apps" standard is disingenuous.
THat's in direct contrast to MS, who WAS using private APIs to SELL more Office.
Did you miss the part in this post where I said I worked at Microsoft, in the early 90's, on Microsoft Word for Windows, and I never saw any secret private APIs?
I'm going to assume you missed that, and didn't mean to imply that I'm lying about my direct personal experience.
So, I've got direct personal experience. What evidence have you got to the contrary?
P.S. As I keep saying, give me Linux rather than Windows or Mac. Microsoft has done some bad things, like buying an ISO standard for their XML file format. I just think it's stupid to bash Microsoft over false rumors, while at the same time giving Apple a free pass to actually do what Microsoft was accused of doing. Bash Microsoft over the things they actually did, please, and don't have one standard for Microsoft and a different standard for Apple.
steveha
Linux has loads of undocumented APIs
I'm not sure if you are joking, you didn't understand me, or you are disingenuous. So I'm going to give you three answers.
If you were joking: sorry dude. I do have a sense of humor, honest.
If you didn't understand me: then I will simply point out that Apple controls app distribution for iPod Touch, iPhone and iPad; and if you use any APIs undocumented by Apple, your app will be banned from distribution. So, in this context, "undocumented" carries a special meaning: something that has not been officially documented and blessed for use by Apple, and apps use undocumented APIs at their peril. There is no such thing as a Linux API that we are all forbidden to use because some entity is trying to keep that API secret, and especially there is no entity that can insert APIs into Linux and use them exclusively while simultaneously forbidding us to use them.
If you were disingenuous: then you already knew the explanation above, and just were pretending not to, in which case I am sad. (I have met Apple fanboys who are this zealous in their fanatic defense of Apple.)
steveha
To advantage which of it own Apps does Apple use its OS advantage ?
On the iPad, only Apple software can multitask (this article has a list: email client, SMS text client, and other apps). On any of their platforms, only Apple software may use the APIs that let you customize the way the UI widgets display. Only Apple software can use the full functionality of the accelerometer. Here is a blog post discussing some undocumented OS X features that made Safari much faster than Firefox 3. And here is a blog post discussing how several apps were rejected for using undocumented functionality. And here is a whole article discussing undocumented Apple APIs, with examples of cool stuff that only Apple's own software is allowed to do. And here is an article discussing cool things that Safari can do, that Firefox isn't allowed to do. And here is a column that claims that Apple inserts undocumented APIs and uses them in its own code for years, without ever documenting them (but presumably without breaking them because it would break Apple's own code). Even the APIs for the WiFi are undocumented.
I understand the argument that Apple doesn't want to commit to supporting these APIs forever, like Microsoft has had to do with even obscure APIs in Windows. If you use these undocumented APIs to do cool things, and Apple revises the OS, your app may break. And Apple doesn't want the customer to think it's Apple's fault that your app broke.
But I also understand the argument that some of these APIs allow for really cool stuff, which is currently reserved only for Apple. People don't like this.
As for me, give me Linux anyway. No such thing as an "undocumented" API, and there is no entity that has an unfair advantage over everyone else, and I can install any software I want.
steveha
Didn't Micro$oft have API's that they used and didn't want anyone else to use? Didn't they get lambasted for that?
Oh yeah. I worked at Microsoft in the early 90s, and I even worked on one of the flagship applications (Microsoft Word for Windows). I never saw any "secret backdoor" APIs, and I firmly believe that those rumors were wildly overblown.
But Apple is actually doing it. They have undocumented APIs that they won't let anyone else use. Even on the Mac they have them, and they have been known to break the undocumented APIs seemingly just to burn apps that dare to use them. Now that they have the "app store" it's even more up front: if Apple figures out that you are using their double-secret APIs, they will reject your app.
Microsoft was widely flamed over rumors of doing this, while few people care that Apple actually does it a lot.
Feh. Give me Linux, please.
steveha
http://www.youtube.com/watch?v=zX09WnGU6ZY
steveha
Why do you want me to suffer?
I want you to suffer for writing pretentious, angsty things like "Why do you want me to suffer?"
As for the programming stuff, if ease of grepping is the most important thing to you, then go ahead and stick with C. The global namespace will force you to write globally unique function names, which are perfectly greppable.
I haven't worked on a large C++ project but I have worked on a large Python project, and I can tell you that I didn't "suffer" due to the lack of globally unique function names.
How often do you need to grep for every reference to a function? And why do you need to do it? I work, all by myself, on a rather large code base, and I can't remember the last time I needed to do this.
P.S. I'm a command-line fiend, and my favorite development "environment" is vim plus ctags. Nonetheless, I do have access to things like Eclipse, and if I ever need to "refactor" code (example: rename a function, both the definition of it and all calls) I am willing to fire up Eclipse and use the GUI refactor tool. And then, I rebuild my ctags file and go back to using vim to do my main editing work.
Spend less time yelling at the kids to get off your lawn, and spend more time looking at the available tools.
steveha
The GCC guys are not going crazy here. They are discussing what subset of C++ to allow.
If you use all the wild features of C++, the results could be scary. For example, operator overloading is great if used judiciously, but if used badly it can make the code a mess. And if it is used at all, then it means that you can't look at one page from a printout and know for sure what that code does; you need to look at all the class functions to make sure there aren't tricky overloaded operators.
I use plain C all the time at work, and the top C++ feature they should be using is simply the object-oriented class stuff. With a single global namespace you need to make functions like MyClassAddFloatAndInt(), but in C++ you could just call that function add(); it would be part of MyClass, and if you have other "add" functions with other type signatures, they won't collide. They could go from:
{
MyClass m;
MyClassInitialize(&m, foo, bar);
MyClassAddFloatAndInt(&m, 3.0f, 2);
MyClassDoSomething(&m);
MyClassCleanup(&m);
}
to:
{
MyClass m(foo, bar);
m.add(3.0f, 2);
m.do_something();
}
Even better if they allow the use of C++ namespaces to keep a large project organized.
The other major win that comes to mind is simply being able to use powerful C++ libraries like the STL. Not having to cook up some kind of container data structure in plain C, but being able to use std::vector<SomeType> and std::map<SomeType, OtherType> and such is a huge win.
P.S. I read through much of the discussion and here was my favorite post:
http://gcc.gnu.org/ml/gcc/2010-05/msg00757.html
steveha
A link would be good. Here's one for starters:
http://en.wikipedia.org/wiki/EPUB
Tim O'Reilly agrees with you.
steveha
As Chad Birch pointed out, the JavaScript thing was an April Fool's joke and I failed to notice. That's what I get for posting in a hurry. Thank you for the correction, Chad Birch.
steveha
I agree with the other comments: if he doesn't have the interest, or if he doesn't have the aptitude, then trying to push him into coding is a waste of time.
That said: check out PyGame. PyGame is a set of libraries for Python, specifically intended for creating new games.
http://pygame.org/
Hmmm. I just went there, and it says that PyGame has now been ported to JavaScript. That probably makes sense, given the major efforts to speed up JavaScript in the new-generation web browsers.
At the PyGame web site, there are a bunch of games people have written, with source code available; and some of these games are half-done and half-broken. If he has the inclination to code, he might get interested in a half-done game and start fixing it up. Or even take a game that isn't half-baked, and start adding new features to it.
steveha
A coming technology that is still in very early days is a desktop fabber, also called a 3D printer.
There is an open source project to make useful fabbers. Current fabbers are designed to use easily-acquired parts. I look forward to the day when someone makes a fabber that can fabricate all the parts needed to build another fabber, but that day is distant.
http://fabathome.org/
P.S. A commercial 3D printer was used to make props for Iron Man 2, including the gloves for the suit.
http://www.ecouterre.com/static/17545_iron_man.php
steveha
You could try inventing science and/or medical sensors that would be very inexpensive to mass-produce, and share the designs with developing countries.
Medical devices are super-expensive, partly because in developed countries you have to get government approval which costs huge money. If you could produce, say, an EEG system that just plugs by USB into a cheap computer/netbook/whatever, and share the specs with the world, you might do some good.
My dad was in the hospital recently, and he had a half-dozen medical sensors connected; his pulse rate, blood pressure, respiration rate, and several other things were monitored by computer. If any reading went seriously out of spec, an alarm would sound. Could you make something like this that would be cheap enough for developing nations to build and use it? Maybe even make it WiFi so you can just set it next to each patient's bed, and one nurse with one computer could monitor all patients for alarms going off.
By the way, the sensor for pulse and blood oxygen level is a Pulse Oximeter.
You might want to research whether there is already a project to make open source/open hardware stuff for developing countries; no sense in duplicating something that someone might already be doing.
I just did a Google search, and USB oscilloscopes exist and are surprisingly cheap. I wonder if you can adapt one for medical uses?
If you are not already a software developer, and you want to write a demo app to show what your equipment can do, I suggest Python for the language on the host PC.
steveha
Thank you for the correction. Even as a dilettante, I should have figured that Avast wouldn't be free for a mass rollout at a school.
Also, I meant to say: Windows 7 would be a much better choice than XP, because I think Microsoft really has improved security versus the really old XP code base.
steveha
First of all, my credentials: I'm not a professional sysadmin, just a professional software developer who also admins a few systems. You may want to give more weight to opinions posted by actual school sysadmins.
Okay, my recommendation: Ubuntu. And my dis-recommendation: you don't want XP.
With Ubuntu, if the computers are reasonably fast, you can do a full re-install in a short amount of time. I imagine there will be cases where someone manages to monkey around with a computer and mess it up, and ease and speed of re-install will be a win for Ubuntu. With Ubuntu you don't need to install, reboot, install a driver, reboot, install another driver, etc. Note that if you know what you are doing as a Windows sysadmin, I believe you can make a "slipstreamed" CD image with the drivers and such pre-installed, which would mitigate this a lot.
Other Linux distros would also work well, but Ubuntu has the momentum as the free home-user Linux distro of choice. Some of the high-school kids will possibly already be running Ubuntu at home. I would suggest having all students login as "guest" and have the "guest" account set up so that, when the user logs out, the /home/guest directory is removed and then replaced with a copy of a standard /home/guest directory. If some clever little black-hat wannabe edits .profile or something to try to set up a joke on the next user, this would sort that out.
With XP, you can configure the logins to be non-Administrator, and unless you are totally insane you will do so. It's hard enough to keep the computers virus-free in any event, and if you let high school kids have privileges to install software, you are just asking for trouble. But you will absolutely want anti-virus, and that means you will want to keep the anti-virus up to date, and that is a big headache that you can completely avoid with Ubuntu.
Now, as for software. Everything I'm discussing below is available on both Linux and Windows. On Ubuntu, it's trivial to install these; on Windows it would be more work. (But again you could probably work something out with a "ghost" disk image or some such.)
Depending on what kind of engineering we are talking about, many of these students may move on to using Matlab in their working lives. Matlab is expensive, but educational copies are probably available. But there is also the completely free GNU Octave, which is basically a Matlab-alike, and is freely available on Ubuntu. There is even a GUI wrapper for Octave, written for the KDE environment, but I haven't tested that. I have only used Octave for DSP work, but it is adequate for that, and did I mention that it's free. The graphing tools in Octave are not as good as real Matlab but they can get the job done.
http://www.gnu.org/software/octave/
I have hopes that someday Matlab will fall by the wayside, and Python will replace it. Specifically, Python with the NumPy and SciPy extensions. With SciPy, Python can do much of what Matlab can do, and it does it with a much better base language; Python is a marvelously clean and tidy language, while the Matlab language is just annoying. Let's face it, it will be many years if ever before Matlab could be replaced by Python, but there are science labs and engineering groups out there using Python and the numbers will only grow. The graphing tools in SciPy are IMHO better than the ones in GNU Octave, and almost as good as the ones in Matlab. You can get Python with all the SciPy stuff pre-installed as the project Python(x,y):
http://www.pythonxy.com/
There is also a project to take every bit of free math software and glue it all together into an amazing giant math tool: this is called Sage. Sage, also, is based on Python. You run Sage as a server and you use any standard web browser as the GUI client. You can run Sage on the same computer as the web browser, or you can set up one
They are taking an expensive container unit, and a very expensive solar array, and buying ten thin clients, and the main computer is a Pentium 4?
That's just crazy. All I can think is that they are a computer recycling outfit, and they had a Pentium 4 on hand and just said "Eh, good enough."
I would have spent a few hundred and gotten an AMD dual-core or even quad-core chip and some ECC RAM. And probably a flash boot drive. You want the computer to be as bulletproof as possible, and it would be nice if it was energy-efficient. A Pentium 4 is basically a device for heating a small room... in other words, not power-efficient. It has a single core. I don't believe any Pentium 4 chipset supports ECC RAM. (Hence the AMD suggestion. Yes, I am an AMD fan, but Intel reserves ECC as a feature of only their expensive server chipsets; whereas ECC has been a standard feature with AMD for years now.)
If you are sharing one computer among ten users, extra cores would be a very good thing. Large cache would be a very good thing. Both of those argue for a modern CPU.
The article didn't say what the OS is. However, the article has a link to another article about the first such cyber cafe, and that previous article says they are using NComputing thin client technology. NComputing appears to be a solution for hooking up external thin clients to virtual machine images running on a host. Which begs the question: ten virtual machine images running on a Pentium 4?!? Let's hope they are least are using one of the models of Pentium 4 that supports the Intel virtualization "VT-x" instructions.
Oh well, I'm sure it works, and it's a heck of a lot better than nothing.
steveha
There are various independent sources of this temperature data, of which CRU produced one. All the various independent sources are in very close agreement. What are you proposing to "start over" on?
I would like to see data sets with complete paper trails attached, and all source codes to all processing provided. I would like to see independent people like the "Climate Audit" guy be able to completely reproduce all steps in data processing without needing to file any FOIA requests. This shouldn't be controversial; if the AGW proponents want the governments of the world to suppress whole sectors of the economy, with costs in the trillions of dollars, then the AGW proponents need to provide absolutely iron-clad and maximally transparent data.
Luckily for me, it looks like this may be happening:
http://www.timesonline.co.uk/tol/news/environment/article7039264.ece
You say that all the independent data sets agree. Can you point me to an independent data set where there is a full paper trail attached to all the data, and source codes are provided for all processing of the data? Bonus points if this data set was actually used in an IPCC report or some similar official report.
steveha
I agree about keeping your mouth shut and your ears open. I've never worked in a truly bad shark tank sort of environment (although it sounds like you have) but it's always good to keep your mouth shut and your ears open.
There is an old business advice book called Bravely, Bravely in Business that contains some all-around good advice. There is one rule in that book that is very very hard to follow, but a good idea: "Never say anything about anyone that you wouldn't say exactly the same way to their face." Think about what would happen if you say something like "Joe is a total idiot who can never get anything done!" or worse, and someone overhears and tells Joe exactly what you said (or mis-remembers it as something even worse). Or if Joe overhears.
Don't hesitate to say something nice when you honestly feel it is deserved. Do hesitate to say something negative, even if you honestly feel it is deserved. "If you can't say something nice, say nothing at all." That's trite but it's good advice.
Like others, I disagree on the "always eat lunch alone" thing. It's good to keep some separation between work and the rest of your life; your boss doesn't need to be your best friend. But you want to learn to get along with your co-workers.
At my second job ever, all the software developers used to go to lunch together, and we would often chat about problems we were working on. It was a fun, informal way to share ideas and information; insisting on eating alone would not have been a good thing in that situation. (Later, management at that company forced some changes in the cafeteria, and as a direct result all the software developers stopped eating together and started going out to lunch. The camaraderie was lost. It was a shame.)
P.S. Looks like you can get Bravely, Bravely in Business pretty cheaply right now. You might want to order a copy and read it; it's good. (Read the customer reviews on Amazon.)
steveha
If it is not really about control, then why did Apple say that programs must be written by hand in Objective C?
If a Flash translator can output Objective C code, which is then compiled down to native code as per usual; and if the resulting app conforms in all ways to the Human Interface Guidelines and other requirements; and if that app runs efficiently and provides an excellent user experience... that app is still forbidden. Why, exactly?
steveha
Some station data held by the CRU was not made available publicly because it is the intellectual property of some national meteorological services around the world and subject to non disclosure agreements.
It is much worse than that. The CRU guys would not even give a list of exactly which data sets they were using. Some of the guys trying to check the numbers were willing to go around and sign NDAs to get copies of the numbers, and the CRU guys wouldn't even tell them which number sets to ask for. The CRU guys claimed that they were under non-disclosure agreements that forbade them from even identifying which stations' data they were using.
http://climateaudit.org/2007/05/31/why-does-cru-have-a-confidentiality-agreement-with-germany/
So, the CRU guys want to impact the global economy to the tune of many trillions of dollars, which is an abstract way of saying they want to inflict misery on millions of people. They claim it is a necessary evil, needed to prevent horrible misery down the line. But they also ask us to just trust them, and not review the data at all, which is just plain unacceptable.
The Large Hadron Collider guys aren't handing out copies of their data, but they also aren't asking the world to impose Draconian controls on the economy. The global warming people are demanding extraordinary interventions into the economy, and they must have an extraordinarily high level of disclosure. They haven't met this.
The CRU should have had all the data sitting on a web server, and all the source code they used to massage that data and make graphs. They not only didn't do that, they didn't even have the original data (they claimed it was "lost"). I am stunned at the level of confidence you have in these guys.
If there really are as many data sets available as you are claiming, then some climate scientists should start over and do the whole research project again, this time with full transparency. If they reproduce the results, then maybe we can talk about Cap and Trade.
steveha
ARM 8 means no out-of-order or dual issue.
Correction: ARM Cortex 8 does mean dual issue. I apologize for the error. (I was correct about the lack of out-of-order execution.)
http://www.arm.com/products/processors/cortex-a/cortex-a8.php
Also, I think I was remiss in not mentioning that the Tegra 2 is only available in small (prototype) quantities right now. But I believe the first Tegra 2 devices will ship within the next 3 or 4 months.
steveha