A Fender? Not likely, it looks more like a B.C. Rich, if anything. Didn't check out the Flying V.
I was disappointed the article didn't mention this IBM effort. Another IBM logo, but the whole thing is 3.3 x 8.8 nm. Still not as small as Hitachi's, but smaller than the other IBM logo they cited. They used an atomic force microscope, which is one of the cooler-sounding lab instruments...
It's a bit chatty in places I for one was kind of creeped out by the first page. It read like some kind of flashback scene from a soap opera — very third-person, and the grammar was pretty casual. I had to go back and re-read some parts to figure out what was really going on. Maybe I was expecting something more like the internal e-mail messages we've seen posted recently, which are pretty blunt but professional. I felt that first page should have been set in italics, as kind of a "mood setter" before the actual article began...
Maybe it's just me. Then again, what's with that blue text? Creepy, I'm tellin' ya...
They need to put out a Request For Papers, so they may fairly consider all possible bidders or vendors on a project, Are you sure that's what it's called? Every RFP I've ever worked on was a Request for Proposal; basically, an invitation to send in a bid on a project. The government RFPs I've worked on were often easier than the corporate ones, since they tended to spell out in detail what the project requirements were. I've seen RFPs from large companies that literally just said "Proposal shall be for a replacement to the existing XYZ system", with no explanation of what the XYZ system is or does (e.g., the "existing InfoMax system", which was a home-grown inventory/order-entry system with no known surviving spec)...
Also , fork() can be quite expensive. Not a problem if the process you're firing off will be running for hours It's generally only expensive if you try to use the resulting process like a thread. If you use it for the purposes it was originally designed for (create a subprocess and execute a new program within it), then it's quite reasonable. Especially with today's machines that can use hardware copy-on-write support to avoid having to duplicate the parent's address space.
but it is an issue if the process simply (lets use Apache as an example) reads from a socket , spits some data back down then quits. Exactly.
if you can find an inkjet without a banding problem, it often has output as good as or even superior to a high-end laser printer. The best computer photo prints I've seen have all come from inkjet printers, not laser printers. Depends on the paper you use. If you're printing photos on the proper photo-quality paper, then the inkjet wins hands down. OTOH, if you're like me and typically buy whatever paper's cheapest at the grocery store because I've run out at 11:30 at night, you tend to get varying results. Laser printers (my Lexmark, anyway) tend to give more consistent results across varying paper grades. I have actually chosen to submit a diagram-laden document I printed on the laser over a copy I printed on the inkjet, because the color diagrams saturated the paper to the point that it got "wavy" (and there was some discernable bleeding).
No alternative currently exists for gcc which is free software. I find it hard to believe there are no other OSS/free C/C++ compilers. Yes, I know GCC does more than C/C++, but what more would you need to build the kernel and userland? If the hammer came down, I'm sure it wouldn't be too much trouble to pick up some other compiler and put the work into it to get it to fit in the spot GCC left behind. I mean, GCC's an impressive piece of software, sure, but it's not like you couldn't get another compiler if you had to.
Actually, now that I think about it, why even worry? What improvements are they going to put into GCC anyway? Additional back-ends for new architectures, some bug fixes, but it's not like C or C++ are going to undergo any radical evolution that'll require massive changes.
i'd look into beginning asp.net 2.0 with c# by wrox press Since the question was about C#, why not mention that there's a corresponding Beginning Visual C# 2005? I found it to be very helpful, unlike the other C# book I bought (Core C# and.NET from Prentice-Hall). It starts off with C# itself, and only after covering that thoroughly does it get into Windows-specific topics (ignoring the fact that C# itself is largely Windows-specific).
On a meta-note, be sure to check out what employers in your area (or the area where you'd like to live) are actually looking for. Here in the Midwest, it's largely C#/.NET, but on the coasts there was a lot more Java work. Java also seems to be more prevalent in companies that have large systems, whereas.NET dominates in the small-to-medium size business market. Think about where you want to live and what kind of company you want to work for. I'm sure you can find work no matter where you go, but speaking as a Java developer in the heartland — well, there's a reason I have these "Learning C#" books on my shelf...:/
Having assembly-line "programmers" as IT advances seems counterintuitive to me. Nobody programs in assembly anymore — oh wait, assembly line, n/m...
Seriously, I can easily imagine groups of programmers whose job it is to write system adapters so that various systems can share data. They may not even write the code, they'll see blobs on their screens that represent "objects" (web services, messages in a message queue, EJBs, etc). Control-click on two blobs to define a translation conduit, the up pops a couple of windows showing the attributes of each object and maybe a tools frame. Click on an attribute in one object then on another attribute in the other object defines a mapping between them. You may need to map through a tool (convert double<—>float, Oracle NUMBER(9,2)<—>DB2 FIXED(9,2), whatever). Not something that can be done automatically, but not something that requires a lot of training or skill either.
I actually wonder if it won't go the other way, though. Imagine something like UML, but that can actually describe 80% or more of a system's requirements. You'll have a systems (or "business process") analyst work with the users to define the requirements, and then the UML is translated directly into the final system. Hell, maybe they'll come up with a system that can execute UML directly, who knows? But in those cases all you'll need for most systems are one or two high-end people. Eventually we'll have standards in place to define how data is stored (everything gets defined using ISO domain types), communicated (XML will never die) and processed (no more questions like "do we post revenue in the quarter the sale is made, or when we actually receive payment?"). If you think this won't happen, just stay on the line and I'll connect you to Mssrs. Sarbanes and Oxley...
(the depth and temperature affect the propagation of sound waves) Not just depth and temperature, but salinity and other factors as well. I skeptical as to just how accurate these systems can be, when you can only determine the distance and direction to the marker with limited accuracy. If you're charting your course with a piece of chalk, then I'm sure they're great, but I wouldn't trust one for re-calibrating any internal instruments. And what's the point of using GPS if you're only going to be able to give accuracy to +/- fifty feet?
No, that will be a problem. Once you have the database, what exactly am I supposed to input for searching? Will I need to learn how to create a 3D model in order to search for similar objects? Depends. Did you have to learn how to spell in order to use a text search engine?
The people who are going to be using this sort of database are going to already have tools available to create their models. People have been creating MOL and PDB files for quite awhile now, and if there isn't a file converter/importer then I'm sure there will be soon. Plus, researchers often want to just search for things that are similar to something they're already looking at. So what they'll do is take whatever model they're currently playing with, lop off chunks of it, and submit the remaining bit to the search engine to see if they've got anything similar on file. So it's not like anyone's going to have to sit down and drag-n-drop individual atoms until they have their model built up...
And the reliability studies show that Wikipedia is just about as reliable as Britannica: http://news.com.com/2100-1038_3-5997332.html Yeah, but how reliable is that reliability study? I mean, c'mon, news.com.com? Their domain name doesn't even make sense, and we're supposed to trust their data?
That's fine if most of your users spend time on the command line. However, it's a little difficult to add a quick link to the group web portal to execute your command-line app.
I've written a good half-dozen simple web scripts in the past few months, for specific little tasks. My tool of choice is JavaScript. I write a static HTML page with a form and some documentation on it, then just execute a js function on submit and use DOM to stick the output on the same page.
Example: For hysterical raisins, we have to pass parameters for one our of main web apps in the URL. Even though all connections to the server go through SSL, we still encrypt the query string. I created a simple page that allows one of our support techs to paste an encrypted URL into a text field and hit submit and the page decrypts the URL and presents the contents. This permits them to easily check the parameters and verify that the information was passed correctly.
I'm creating a slightly more complicated function this week, and I'm going to write it as a CGI script. Not because of this article, but because the task at hand is a perfect task for Perl. And I estimate I can write the CGI script in an hour or so, versus the four-plus hours it would take in Java. Or the three-p;us days it would take me in Ruby (since I don't know Ruby...;)
I don't see why hardware vendors don't release their source code. They can patent the hardware if it truly contains innovations and with software patents they could patent all or part of the driver if it's anything special They're afraid that their competitors will learn what they're doing. If they've (let's say) implemented their vector lists using red-black tries or some other common data structure that isn't patentable, then that might give their competitors insight into some aspect of driver writing that they wouldn't have considered. Or take the example of a "unified driver" that appears to be designed to support some unannounced functions ("Hmmm, I wonder why they're storing the GPU context in an eight-entry table, it's not like you can put more than one GPU on a board....") And really experienced driver writers or GPU engineers can recognize hardware features for things like OpenGL acceleration or occluding shaders. So even if the hardware is patented, the competition can see what technologies or features they're targeting. And there's almost always more than one way to implement a feature....
I'd rather have something breaking than a security hole I doubt you'll find many sysadmins agreeing with you there. As someone else mentioned, most sysadmins will already have disabled telnetd. So to install a patch and reboot their systems without warning (possibly during the work day) seems like a little harsh treatment for somebody who's already mitigated the threat.
But it's only reachable via ports 80 and 443. And I installed patch #120069-02 a couple of weeks ago. In fact, I already installed the -03 version of that patch. If you keep up with your security patches, it's really not a problem. Of course, this is easy for me to say, I have one workstation; I'm sure that for sites with dozens (or hundreds) of servers, it's more problematic. I also STR that patch 120069 used to require a reboot after installation, which makes it a bit more of a hassle to install (I usually save those for Fridays, when I can install them and then walk away while the box reboots).
If you want power, just burn the food you eat, that can be done far more efficiently. More efficiently than what? Animals are notoriously inefficient energy converters. I remember an ad for the newly-introduced diesel VW Rabbit titled "It can run a mile cheaper than you can". It explained that when a human body ran a mile, it burned some 200 (?) calories. The Rabbit burned some fraction of a gallon of diesel. They showed that the diesel the car used (at late 70's prices) was cheaper than the 1/4 cheeseburger they used to represent 200 calories.
I'd be interested in seeing the amount of electricity generated plotted against the amount of waste heat given off by the exerciser. You might capture more energy by strapping a Stirling-cycle engine to the back of the seat (and keep the user cooler, too!).
Cars are different--why is this analogy always used? Tradition.
I think you might have missed the point. What's the use of manufacturers installing Linux and offering support if they're just going to pre-install a spyware filled custom version? Well, you might get some "N standard time units of AOL F*R*E*E!!!!" shortcuts on the desktop, but I doubt there's much actual Linux spyware available for them to install. And a dedicated Linux user group would probably turn out a shell script to remove or neutralize any free-as-in-unwelcome software within a day or so after the first boxes shipped.
it doesn't matter that's it's Linux and it's supported, if they find out you're running Fedora or Ubuntu or whatever they say "Piss Off." Wait, you're suggesting they support *every* distro? From CentOS to my hand-rolled LFS installation? If so, then dream on, there's no way.
If you want to argue that Linux is just the kernel, then sure, they could do that, but I doubt that many first-time Linux users will be happy to hear that Dell won't help them configure their new box, or help them transfer their data over, because those aren't kernel issues.
Seriously, the only way for this to work for Dell is to pick one major distro and support that. Hell, maybe their support will be a free one-year contract with a third-party support company. Maybe it'll just be guaranteeing that they'll provide Linux drivers for all the hardware on their machines. But whatever it is, it won't be some kind of broad-based support for whatever configuration you can get to boot. They've got to have a fairly specific known configuration that they can log problems and solutions against. What is a tech that knows all the ins and outs of KDE and Gnome going to do when somebody calls in and they're running AfterStep? Or twm? Have them bring up a terminal window? Great -- the user's default shell is VXDCL, a Linux port of VMS DCL. Having the user enter "/bin/bash" results in "%DCL-E-NOCMD, no command on line". Would you know enough to tell them to enter "RUN SYS$SYSROOT:[BIN]BASH.EXE"? And would that even work? (Don't ask me, I've never used VXDCL, although I have used VMS...)
Sure, that's an extreme example (another tradition), but it could happen. And lots of lesser examples would be just as confounding (scsh? Hell, just csh!). So unless you can guarantee that there's some known configuration to base your diagnoses on, there's really no way to provide any reasonable level of support. The alternative is to have every support call start with "Place your system recovery disk in the slot. When it asks you if you want to repair or reinstall, click 'reinstall'..."
The whole thing is we want something BETTER than the current "Install Linux and you get no support" situation. If you could be a little more specific about what you mean when you say "Linux", they could be a little more specific about what they mean when they say "support"...
The downside is that the "Dells" probably won't offer support if you swap the OS and have access to the source Well, duh. Would you expect a car manufacturer to honor their warranty if you plugged in an aftermarket engine-management chip and destroyed your transmission? At least with a computer, you can install the original OS and prove that the hardware is still bad. Current car chips can note the mileage every time you start up and shut down, and log an event if there is a gap...
OP has shown that he can drop into a new environment and make an impact quickly. This makes him an ideal fit for a contracting position This is an excellent point, and probably the best spin the OP could put on his history. Of course, in contracting it's a given that you're good, and what clients tend to look for is someone with domain knowledge specific to their business. I doubt that someone fresh out of college is really going to pick up any significant amount of business insight in three or four months. That is, assuming that all of his jobs have been in different industries; if he's just jumped around among businesses in the same industry, then he's in better shape.
There's only a problem if he can't get a job. Speaking as someone who's been unemployed, that's a serious problem. And the OP is clearly a job-hopper, and may have hard times ahead if the bottom falls out of the.NET market in his area [0]. If you're looking for work and you've got five jobs in eighteen months on your resume, then you'll be round-filed by some resume-filtering software before anyone even knows you've applied.
Seriously, though, is this guy moving around or what? I can see taking an entry-level position just out of school and getting offered real-world money soon after, but 30-40% increases every time? Where in the same area are there entry-level jobs with that kind of wage disparity?
At the end of the day, if a company needs your skills for something, they'll pay for it. Yes, but will they pay you, or someone who has a more stable work history?
[0] When defense companies don't get contracts, they shed engineers like a boxer shakes off sweat. So you've suddenly got a couple hundred highly-trained professionals competing for the same jobs you are. Standards will increase to match the available resources, so all of a sudden entry-level positions require five years of experience. I've seen it happen, and it's not pretty — like pirahanas in a goldfish tank...
Slashdot Mirror
Arrgh! SI pwns me!
A Fender? Not likely, it looks more like a B.C. Rich, if anything. Didn't check out the Flying V.
I was disappointed the article didn't mention this IBM effort. Another IBM logo, but the whole thing is 3.3 x 8.8 nm. Still not as small as Hitachi's, but smaller than the other IBM logo they cited. They used an atomic force microscope, which is one of the cooler-sounding lab instruments...
Maybe it's just me. Then again, what's with that blue text? Creepy, I'm tellin' ya...
...booth babe.
!!!!! C++ doesn't have thread-local storage — JFC! I knew I was right to steer clear of that miserable hodgepodge.
I swear to $DIETY, reading about C++ is like hearing your best friend's kid is in jail. What went wrong? And the parent is so nice, too...
Actually, now that I think about it, why even worry? What improvements are they going to put into GCC anyway? Additional back-ends for new architectures, some bug fixes, but it's not like C or C++ are going to undergo any radical evolution that'll require massive changes.
Uh, that should be C0FdEE (according to my LED hex keypad...)
On a meta-note, be sure to check out what employers in your area (or the area where you'd like to live) are actually looking for. Here in the Midwest, it's largely C#/.NET, but on the coasts there was a lot more Java work. Java also seems to be more prevalent in companies that have large systems, whereas
Seriously, I can easily imagine groups of programmers whose job it is to write system adapters so that various systems can share data. They may not even write the code, they'll see blobs on their screens that represent "objects" (web services, messages in a message queue, EJBs, etc). Control-click on two blobs to define a translation conduit, the up pops a couple of windows showing the attributes of each object and maybe a tools frame. Click on an attribute in one object then on another attribute in the other object defines a mapping between them. You may need to map through a tool (convert double<—>float, Oracle NUMBER(9,2)<—>DB2 FIXED(9,2), whatever). Not something that can be done automatically, but not something that requires a lot of training or skill either.
I actually wonder if it won't go the other way, though. Imagine something like UML, but that can actually describe 80% or more of a system's requirements. You'll have a systems (or "business process") analyst work with the users to define the requirements, and then the UML is translated directly into the final system. Hell, maybe they'll come up with a system that can execute UML directly, who knows? But in those cases all you'll need for most systems are one or two high-end people. Eventually we'll have standards in place to define how data is stored (everything gets defined using ISO domain types), communicated (XML will never die) and processed (no more questions like "do we post revenue in the quarter the sale is made, or when we actually receive payment?"). If you think this won't happen, just stay on the line and I'll connect you to Mssrs. Sarbanes and Oxley...
The people who are going to be using this sort of database are going to already have tools available to create their models. People have been creating MOL and PDB files for quite awhile now, and if there isn't a file converter/importer then I'm sure there will be soon. Plus, researchers often want to just search for things that are similar to something they're already looking at. So what they'll do is take whatever model they're currently playing with, lop off chunks of it, and submit the remaining bit to the search engine to see if they've got anything similar on file. So it's not like anyone's going to have to sit down and drag-n-drop individual atoms until they have their model built up...
That's fine if most of your users spend time on the command line. However, it's a little difficult to add a quick link to the group web portal to execute your command-line app.
;)
I've written a good half-dozen simple web scripts in the past few months, for specific little tasks. My tool of choice is JavaScript. I write a static HTML page with a form and some documentation on it, then just execute a js function on submit and use DOM to stick the output on the same page.
Example: For hysterical raisins, we have to pass parameters for one our of main web apps in the URL. Even though all connections to the server go through SSL, we still encrypt the query string. I created a simple page that allows one of our support techs to paste an encrypted URL into a text field and hit submit and the page decrypts the URL and presents the contents. This permits them to easily check the parameters and verify that the information was passed correctly.
I'm creating a slightly more complicated function this week, and I'm going to write it as a CGI script. Not because of this article, but because the task at hand is a perfect task for Perl. And I estimate I can write the CGI script in an hour or so, versus the four-plus hours it would take in Java. Or the three-p;us days it would take me in Ruby (since I don't know Ruby...
But it's only reachable via ports 80 and 443. And I installed patch #120069-02 a couple of weeks ago. In fact, I already installed the -03 version of that patch. If you keep up with your security patches, it's really not a problem. Of course, this is easy for me to say, I have one workstation; I'm sure that for sites with dozens (or hundreds) of servers, it's more problematic. I also STR that patch 120069 used to require a reboot after installation, which makes it a bit more of a hassle to install (I usually save those for Fridays, when I can install them and then walk away while the box reboots).
I'd be interested in seeing the amount of electricity generated plotted against the amount of waste heat given off by the exerciser. You might capture more energy by strapping a Stirling-cycle engine to the back of the seat (and keep the user cooler, too!).
If you want to argue that Linux is just the kernel, then sure, they could do that, but I doubt that many first-time Linux users will be happy to hear that Dell won't help them configure their new box, or help them transfer their data over, because those aren't kernel issues.
Seriously, the only way for this to work for Dell is to pick one major distro and support that. Hell, maybe their support will be a free one-year contract with a third-party support company. Maybe it'll just be guaranteeing that they'll provide Linux drivers for all the hardware on their machines. But whatever it is, it won't be some kind of broad-based support for whatever configuration you can get to boot. They've got to have a fairly specific known configuration that they can log problems and solutions against. What is a tech that knows all the ins and outs of KDE and Gnome going to do when somebody calls in and they're running AfterStep? Or twm? Have them bring up a terminal window? Great -- the user's default shell is VXDCL, a Linux port of VMS DCL. Having the user enter "/bin/bash" results in "%DCL-E-NOCMD, no command on line". Would you know enough to tell them to enter "RUN SYS$SYSROOT:[BIN]BASH.EXE"? And would that even work? (Don't ask me, I've never used VXDCL, although I have used VMS...)
Sure, that's an extreme example (another tradition), but it could happen. And lots of lesser examples would be just as confounding (scsh? Hell, just csh!). So unless you can guarantee that there's some known configuration to base your diagnoses on, there's really no way to provide any reasonable level of support. The alternative is to have every support call start with "Place your system recovery disk in the slot. When it asks you if you want to repair or reinstall, click 'reinstall'..." The whole thing is we want something BETTER than the current "Install Linux and you get no support" situation. If you could be a little more specific about what you mean when you say "Linux", they could be a little more specific about what they mean when they say "support"...
Seriously, though, is this guy moving around or what? I can see taking an entry-level position just out of school and getting offered real-world money soon after, but 30-40% increases every time? Where in the same area are there entry-level jobs with that kind of wage disparity? At the end of the day, if a company needs your skills for something, they'll pay for it. Yes, but will they pay you, or someone who has a more stable work history?
[0] When defense companies don't get contracts, they shed engineers like a boxer shakes off sweat. So you've suddenly got a couple hundred highly-trained professionals competing for the same jobs you are. Standards will increase to match the available resources, so all of a sudden entry-level positions require five years of experience. I've seen it happen, and it's not pretty — like pirahanas in a goldfish tank...