I had XP corporate image, and Sharepoint would crash if I tried any of the AJAX style operations. I could right-click and save, or click directly, but using the arrows and checking out crashed every time.
IE developer toolbar and Fiddler installed, I uninstalled both and it kept crashing, then the lease came up on my box and I got a shiny new Vista. Installed both and some other stuff and never had a problem. But it keeps worrying me that IE with Microsoft-only additions had problems with Microsoft's website, manipulating Microsoft's document types.
The problem I believe is the WEBDAV type interaction. When you authenticate, IE lets you do stuff. Then you open the file, and normal browsers would download the file and ShellExecute() to open it (or maintain their own list, but whatever). IE sends the URL to the application (if it's Office type), which has to re-authenticate since it doesn't share IE's session. That way the Office app can check in/out instead of just opening a local copy.
To tell the difference, you can obviously see the normal IE download dialog if it's downloading and opening, otherwise you get the Office dialog that has "Opening [filename..." and only has a cancel button.
I just used FireFox, and tried to avoid checking things in/out. It's all intertwined - impossible to fix. Probably lots of code duplication as well.
"POS Withdrawal" and "EFT" say this is using automatic debits directly to the customer's account, most likely over the ATM network. The ATM network is basically a bunch of switches and routers. Think about it like an HTTP request, where you ask for something and the server returns a code. There are 'OK' codes, 'Not ok' codes, and 'retry, I think something happened but not really sure so just let me know kthx.'
Yeah I'm dumbing it down a bit, but it's like that. All it takes is missing a 'good' return code in a list of switch/case tests, or a bad one, or misunderstanding the documentation and misinterpreting a code, and you get the rare response code 'Successful, and by the way you have lovely eyes' which you misinterpret as 'Retry' instead of 'Successful'.
Having worked in the bank support business dealing with these return codes, we see these kinds of mistakes all the time. It happened about once a month, to different institutions. They called us, we told them what to look for, they fixed it, most customers never even saw the mistake. It's likely that the payment processor screwed up and update/upgrade. First the vendor will sort it out and get money back where it belongs, and in the process most banks will either waive the fees or punch the vendor in the throat and then waive the fees, because they don't want each individual to call them separately and tie up the phones.
To repeat that - for most banks, there will be enough affected users that it's cheaper to refund the overdraft fees than to deal with each customer. For the few that don't, the customer's call will be enough to refund. At that point if you can't get a resolution, EA will probably call with the customer on the phone, and now you have business vs. business and the customer sits back and watches. At least one bank will say 'tough cookies' and end up on The Consumerist.
As always, if you have a credit union instead of a bank, you're going to feel safer when these things happen. Funny thing is, they usually purchase outsourced service to keep costs down, so depending on the vendor chosen you might see this kind of thing more with a credit union - but they are more proactive in resolving it without the customer having to pester them.
It's too bad that every time we invent a new language, we have to come up with a new translation of "caveat emptor" because 2,000 years later consumers still don't pay attention to what they are buying.
Sony wants to make money, and if you are costing them money they will find a way to make money or cut you off. With all of the coverage about consoles being a loss leader for the first generation, there is very little excuse for someone to be ignorant that they are an expense to the company. Especially if you are doing something other than what it was designed to do - play games.
Debian's business model is to provide a quality product that other companies can rip off. So that kinda evens things out.
Canonical has found a way to make money to support its work (and maybe profit too, I don't know), fully respecting the license agreement Debian provided. Not that Debian has much choice if they want to distribute Linux, but it's what they provided. As long as you don't step on trademarks, ripoff isn't even an insult in this context.
Sure it has a negative connotation, but that #3 definition is quite neutral in its denotation. I'll assume you're using that meaning, in the given context.
"Speed and uninterrupted use of service not guaranteed." That's what you missed. That's what prevents you from multiplying the bits per second times the number of seconds in a month and coming up with a total number of bits you expect. The disclaimer about which factors may affect your experience should be a dead giveaway. "Actual speeds will vary" means the math is guaranteed not to work out. You bought a speed, not a quantity. If you want a quantity, you're going to have to buy based on quantity. The parts you quoted pretty much defeat your entire argument. You expect, but the contract tells you not to, so that's your problem, not Verizon's.
The plans specifically state that if you are causing network congestion, they have the right to alter your service. You might try to sue them, but if they show you are the statistical outlier and have technical data to back it up, they just take the parts that you quoted and say we sold a speed not a number of bits.
And don't start with that contract of adhesion shit from down below, the adhesion idea is that your are stuck and can't get out. A monthly plan can be terminated, even if you have to repay the hardware subsidy in the form of an early termination fee, which you agreed to in your initial contract and has been upheld in court cases where the termination fee was clearly stated. Even if you have no other provider to choose from you aren't stuck in that contract, you could go without or get a different contract/plan now that you know what to look for.
Leaving your kids at home alone so you can go have sex and drugs is one thing. Ensuring they cannot harm themselves while you're away is thoughtful and deserves some consideration.
Ensuring they can't hurt themselves because their fingers are trapped in a closed, locked door to which you have the only key is... I can't even find the words. Anything I say would not be strong enough, I think it speaks for itself. Just thought I'd share.
I had GTA Vice City for PS2, and an icestorm hit. My car was frozen in place, and I had a long weekend. So I just let it melt. Meanwhile with nothing better to do or no place to go, it was 3 days of GTA. I played in the virtual rain a lot and learned a lot about sliding around. I think it was one of the driving course things where you have to make it cross-island and back under a time limit, and I kept having to re-try.
Thaw came, I got in my car to go to Burger King. Pulled out the driveway, blew through 2 stop signs right in a row (within 20 feet of each other) and got pulled over in less than 50 feet since it was right in front of a cop car.
I only escaped a massive ticket because some fire or shooting required his presence more than little old clearly sober me. "You didn't even try to slow down, for either one of those," he said as he hopped back in his car.
Now every time it snows, I get in my 6-speed and drive around for fun, the more sliding the better. Cool story bro, there I saved you the time.
How does the thread know how much to reserve at startup time? Unless you know exactly how much you'll need, you'll run into the problems I mentioned. The specific problem being solved involves a program that makes heavy use of memory allocation, so you probably can't count on knowing how much you need and you'll have to go back and get more. More specifically, this is intended for single-threaded application running on multi-core processors.
Atomic instructions are basically system-wide critical sections, and heavy allocation will make this a bottleneck for the application (not necessarily the system). For most normal programs with occasional allocation it's not going to matter how you implement it.
Windows already does exactly this, reserving 1MB by default and this is controllable within the executable or when creating threads, so if you do know what you need you can certainly do this at thread level. So yes, it would be trivial to write a lockless malloc where the thread controls its own allocations and free list. There just isn't a benefit to doing that, and that's not the problem being solved here.
I'm not defending the application. The claims leave this open to being implemented manually, and the supporting documentation makes it clear that the claims do not match the described invention. We should be objecting to this patent, but not based on the description in the summary. No one here reads the patents, so I wanted to make sure people noticed that the invention itself is not trivial. The patent application is unfortunately just as terrible as the summary of it.
If we are going to have a discussion about a patent, I'd prefer not to wade through hundreds of comments which are completely off-topic, simply because the summary was awful. These bad summaries really make the patent process and applications as far more evil than they actually are, which is quite a feat since in reality it is quite bad without the additional propaganda. We need to fix things that are broken, not simply fuel the fire with misleading garbage. I actually wanted to moderate everyone's misreading as off-topic, but they probably wouldn't understand since there is no "-1 You are objectively incorrect and here's why" moderation option. But I digress.
IBM is not optimizing by trial and error as the summary states - it is preventing performance impact by temporarily reverting adverse refactoring using objective data, according to the claims. Trial and error implies it's just randomly looking for optimizations, but the purpose of this has nothing to do with optimization. It is temporarily reverting specific refactoring. "IBM Patents Optimization" is not remotely close to the actual claims, even just the subset quoted in the summary. I would have maybe said "IBM appears to want to patent undoing refactoring which negatively affects performance." That's what the actual claims are, and it's clearly invalid unless everyone who refactors code does no performance evaluation ever. With many tools, performance testing isn't even part of the refactoring workflow, so it could be a legitimate addition to refactoring tools if implemented as described (not as claimed).
At the very least, claim 3 "undoing refactoring is performed before the computer code is compiled" should be clarified to say it is automatic and temporary, or another claim added "The method of claim 1 wherein the undoing refactoring is automatic and temporary". That would fit the description and properly narrow the scope of the invention.
I would assume that patent law considers the claims to be the determination of whether something is infringing or not, and if that is the case this patent application should be rejected. Whether IBM did this intentionally or not doesn't matter, the patent examiner should catch this. Since it looks like an application and not a granted patent, I'd hope the examiner will probably ask for an update.
This is the invention in a nutshell, which as you said is unrelated to the claims. Hopefully we can now get input on the poor quality of the patent application, and discuss possible prior are which invalidates the application (or a genuine "good job guys" if deserved):
TECHNICAL EFFECTS
[0007] As a result of the summarized invention, technically we have achieved a solution which solves the conflict between program performance and readability in the refactoring process. By analyzing the refactoring history, refactorings which degrade performance will be temporarily undone before compiling to regain the lost performance, but not affect the program's readability obtained by refactoring.
Forget about optimizing at all, consider all optimizing turned off. This invention basically makes sure that your refactoring doesn't impact performance. The compiler can then optimize the rearranged code instead of the original code you're looking at. Optimization will benefit both inputs, but the un-refactored code can still run a lot faster.
An optimizing compiler chooses its optimization from the tricks it knows about. Automatically inlining frequently called functions would be something that an optimizing compiler could do, but that is not refactoring the code - it is the opposite of classic refactoring. So to answer your particular question no, you can't say that. Auto-inlining would be an example of the simplest implementation of this invention but hardly representative.
The essential difference here is that the user chooses to rearrange the code so it is more understandable. An optimizing compiler would rearrange the code so it is faster. The purpose of rearranging is different. The invention in this case is tracking which refactoring was beneficial to performance and which was a negative impact, and un-doing the refactoring with a negative impact while preserving neutral or advantageous refactoring. Specifically, this happens behind the scenes so you don't have to even know which changes were bad.
To put it another way, a normal compiler has a set number of tricks up its sleeve, and it applies those tricks when it thinks it should be beneficial. This invention adds two additional checks. First, it doesn't apply optimizations, it only reverts code on the fly *before* hitting the compiler. Second, it relies on performance data to choose which reverts to apply. I've never seen an optimizing compiler which builds the executable, tests it, benchmarks it, and goes back and updates its optimization strategy based on the results. That's what IBM are claiming. An existing optimizing compiler would do the best it could to optimize what you give it, not revert entire blocks of code. This invention would decide based on performance which version to use.
The other thing I forgot to mention is how it deals with bug fixes. I assume you would go through the 'Refactoring Wizard' for refactoring work, and just change the code directly for bug fixing. But what if the buggy code ran faster than the refactored but bug-free version? The implementation behind this would have to be very interesting in order to not completely hose your app, and that is why this is impressive. They actually think they can do this.
The obvious application is one of the UML-to-code-to-UML IDE's, including Rational, using Agile type methods. A few clicks to refactor, and when the human makes the code worse the wizard ignores the human, but pretends to obey.
You're funny, but the video does not lead to the conclusion that game skills don't apply. An interesting experiment, but real life is not sitting in a dark cabin looking at an overhead monitor. We would have to compare driving skills of gamers and non-gamers, with similar experiences (accounting for country road drivers vs. city drivers), in order to even come close to such a conclusion.
Flight sims used to be very unrealistic, but they were still used and effective at giving people practice in uncommon situations. They have improved of course. Video game driving has probably taught a lot of people about bad weather conditions, where you would never experiment with your tangible car.
This is not science, it's not a conclusion, it's simply entertainment.
This is not a method of gradually figuring out how to boost performance by trying different things.
This invention makes changes to the source code on the fly. The purpose is to allow programmers to re-factor code for readability, but un-do the process before compiling. Presumably this is built in to one of their automatic refactoring applications or IDE. It is about fixing the problems refactoring causes, specifically in this case performance issues. I assume it could work like automatic inlining of small, commonly used functions, but is open to be a lot more complex.
You click the buttons on the GUI, it refactors some things for you and maintains a history of those actions. It stores the resulting code, which you send to source control or final build kit or whatever. When building, it examines which changes had a negative impact on performance and reverts those before compiling.
To the user, the code is refactored. To the computer, it is not.
My question is, how the hell do you debug this kind of output, or store it in source control? You are clearly going to be locked in to the same IDE, and probably same IDE version, and have to check in all of the project files. Cross-compilation is going to be difficult at best, since compiling the raw code won't have the performance enhancements. Clever for a closed-process shop, but kinda impractical.
What will be more interesting is whatever people come up with to deal with the misery this will create.
I enjoy customizing my apps by sub-classing windows or whatever else is needed.
I fixed a particularly stupid application for work which requires you to click on a calendar and choose a date, and again on another calendar. Clicking OK updates the data source, but clears all the values. So every day when you're going through today's data, you select both calendars and click OK and re-set both calendars. Lots of wasted time. I updated the button click to save the calendar date, call the normal click function, then restore the calendar. Took a 4 hour process down to 20 minutes.
IDA helped confirm a few things, but mostly it was just Spy++ from MSVS 6 that helped me figure out what was needed. I could have gotten all cool and stuff, but I just put an injector in the same folder, and a.dll which gets loaded by the injector. Open the app and open the injector and enjoy your efficiency. The author would not consider making an update so I did it for them.
No hack, no crack, no DRM, no derivative works because I was simply calling functions which the application made available.
For those who use Avaya CentreVu, I screen-scraped the agent display so it could send alerts when threshold settings were hit (beep on yellow, send a page on red). Again just using existing functionality and an injector.
I also remember an assembly "reversing" tutorial which added line numbers to Notepad, but that was invasive because it altered the underlying code. I always wanted to go back and do it using a less invasive method, but I never thought about simply grabbing the pixels and shifting them, adding appropriate line numbers based on the scrollbar window properties.
This application of an ages-old technique is particularly interesting because it provides legal coverage for compliance with the ADA, if you happen to be subject to that. Clever.
I've personally seen a few friends who were good programmers strugle in Corporate America, because they were used to working on things as they wanted to, in their free time, with no expectations as to when the next release or patch would come out. And once again, that is the exact opposite of the Real World.
We did this thing where we tracked estimates heavily, with a target of +/-15% on estimation. Estimate correctly and you can set expectations on the release date of the next patch or update. Our process involved making the estimation, doing design, build, testing, sign-off, and project clean-up. Yeah we borrowed time from other slots and in general retroactively updated the estimations so they were right. When asked for a date, we'd say something reasonable. Then when the top execs started looking closely at a few deadlines, we just multiplied the work estimates by 1.5.
They were honestly more worried about meeting an arbitrarily set deadline than being efficient. If I said 2 months and it took two and a half months, we were in the shitcan. If we multiplied it and said 3 months, we took 2 months and 3 weeks and goofed off for a week essentially - take off early a few days or cancel meetings and just zone out. 3 months, just like we said, we made our goal.
If you're doing development, you're doing something you have done before, in which case you should be able to copy and paste, or you're doing something new. The third option is team-wide incompetence where you keep repeating yourself, but those teams are a lot better at estimating. They know exactly how long it takes because they do it all the time.
I've worked in open source, and since I'm writing it for free you'll get what I give you when i give it to you. More businesses need to be like that, but they can't because clients won't agree to pay for anything unless they can get it on a schedule.
Why do we have cost overruns on IT projects which become legendary, like the FBI information sharing software and the child support system that requires operators to fake data so children don't get forgotten? For that matter, why do we have tangible projects like the southern US gate which cost a billion dollars and was so poorly mismanaged that the project just got halted?
Poor estimation is pretty common unless you're doing assembly line type or repeatable work, but when you break new ground it's harder to know what to expect. Open source mindset is more honest about that. Because it can. Walk in their shoes for a week and become a better manager and those open source guys will blossom. That or they are part of the emo gimme generation and you might as well take them out back and show them which grave they will fill if they don't quit whining. That's just fair.
Have every thread allocate its memory from... what? At some point either the operating system or the runtime has to lock something so it doesn't get interrupted, or turn off all interrupts and switching for a few hundred cycles so it can do the allocation. Usually the runtime requests reserved pages far in excess of what it needs, and then doles out pieces, committing them as needed. You need 2k, so the runtime gets 4MB and commits 32k page(s). Next time you need more, then the runtime just returns more of the 32k block.
The operating system has to lock its list for the runtime, and/or the runtime does the same for the program. Someone's locking something somewhere.
He's taking pictures for fun and sharing them. Not for science, but for posterity or public interest, or "cool, look what I got to see". He's posting them for free. From the FUCKING SPACE STATION. 200 miles straight up in the air, traveling at 27 THOUSAND miles per hour, and is able to take pictures and post them on the internet. "Could you please move your hat sir, you're obscuring my view," is a reasonable request. "Could you please move your 300 tonne vehicle out of the way so I can see a few blurry colored pixels that a small part of your SPACE STATION is blocking, without which we wouldn't be able to see the picture at all," this is ridiculous.
Thank you Soichi, thank you from 6.5 billion people on planet earth, except for "Ungrounded Lightning". For some reason he thinks he's an art critic, or the Space Lord Mothermother, or that you can lift an X-wing out of a swamp with the wave of your hand.
Hopefully this puts it in perspective (choose one)
The fact that I cannot kick Microsoft in the balls for Vista (Required by my employer until 7 is tested completely) suggests that companies are not people. We cannot figure out which of the many employees deserve a ball kicking and which were simply following orders, so we can't just kick them all and be fair about it. So we have to fall back on holding the decision makers responsible, or this nonsense will never end.
You're misreading this, I think. Or taking it out of context. Ignoring the second part we have:
Criminal Infringement.--Any person who infringes a copyright willfully [either] for purposes of commercial advantage or private financial gain
That singles out people who willfully infringe copyright for financial gain. You are assuming that downloading a copyrighted work gets you the financial gain mentioned in the definitions. If you take a look just at 506(a), it really reads this way, expanding to clarify:
Criminal Infringement.--Any person who infringes a copyright willfully for purposes of (commercial advantage) or purposes of (private financial gain)
Replacing the definition for the defined word, it becomes:
Criminal Infringement.--Any person who infringes a copyright willfully for purposes of (commercial advantage) or purposes of (private receipt, or expectation of receipt, of anything of value, including the receipt of other copyrighted works.)
In this reading, simple receipt of a copyrighted work is NOT A CRIME. You would have to infringe copyright for purposes of receipt of other copyrighted work to qualify as a criminal. This specifically targets people who trade "warez", and people who upload and download at the same time.
Under your reading, if someone gives me a copy of a movie on DVD or CD I have received financial gain. Ihe problem is, I did not infringe anything for the purposes of financial gain. I did not receive stolen property. How is this a crime?
If I give the copy away (without making one for myself) I still have not infringed. If I trade it with someone for some other movie, I still have not infringed, even though I did receive other copyrighted works, because I did not infringe initially. I did not make the copy.
If I download from Rapidshare, I'm not the one making the copy. I would have to make another copy somehow, and intend to benefit financially (where financial gain is defined).
sure you can argue a different reading, lawyers do it all the time. But unless you have court cases citing how a judge or jury read the law, your post was the most significant, densely packed pile of horseshit I've seen this week. you chose your own reading of the law to support your position.
Effectively, your position is that downloading something for the purposes of financial gain includes the financial gain you acquired by not paying for the download. But they put that word "other" in there. Other copyrighted works. Not trading pieces of something for other pieces of it, but trading something in whole for something else in whole.
I could infringe, and then turn around and trade an empty disk for a full disk, and argue I'm within the law, since the net gain to myself was zero (disk for disk) and I actually took a loss due to the internet cost.
On one hand, the analogy was flawed and had to be corrected. On the other hand, the explanation was poorly done. A better explanation would be that people need to learn things about their browser in order to use it effectively. Like "too good to be true" probably means it's not true. Or there ain't no such thing as a free lunch. Common sense that says don't take unknown things from unknown people. That's what people forget - no other application has opened people up to identity theft just by operating it. Since a browser uses so many external files, it's the exception that no one thinks about.
What's the penalty for not knowing? In the car analogy, safety of the people you might otherwise drive over, or yourself, should be the motivator. People drink and drive, they text and drive, they don't pay attention, or they aren't familiar with their vehicles. Most people do not fall in this category, a few do. The penalty is in a few cases someone successfully installs a spambot the user will never notice. In fewer cases the user's personal files get transferred, and some of those get used and credit card companies block cards due to suspicious activity and a few people lose money. If it were a big problem, we'd hear stories every night on the news, but it only comes up a few times a year.
There's no incentive to learn because 1) it's rare and 2) learning is not a requirement, as it is in a driver's license test. This is where the "Internet driver's license" idea makes sense, until we realize how impractical it would be. Then we're back to the situation where people should learn, but don't, and it's only a problem for a few people a year.
You kinda have a point right up to where you fail to see that bullying behavior is typically learned. Someone has probably already shown them how small they are in the universe and now they are doing the same, once they find someone smaller to pick on.
It doesn't matter how it ends, the bully has to learn that he can't get away with that behavior. Either by having the crap kicked out of him or by being not just punished but actively monitored to ensure he doesn't violate "probation". He just needs to know that if he tries something, something negative will happen.
An equal fighting back will not convince the bully, since some fights will be won and some lost. It's the psychology of slot machines - win a few and you think you'll win the next one as well. The more ingrained the behavior it is, the longer it takes to extinguish. A severe beating can accomplish this, or several attempts at restricting privilages (which involves parents, who may not even care). It's faster to beat the kid, but it might actually feed the need to feel superior and actually work against extinguishing behavior in some cases. It's difficult at best to decide which case will work unless you learna s much as you can about the particular bully, and put aside generalizations.
A patent is an intellectual property right granted by the Government of the United States of America to an inventor “to exclude others from making, using, offering for sale, or selling the invention throughout the United States or importing the invention into the United States” for a limited time in exchange for public disclosure of the invention when the patent is granted.
If you're going to refute something that is over your own head, a citation of some sort helps. "Making" is the first word in the list of forbidden things. That's the patent office website. Go find the actual law and see if it supports your comment.
35 U.S.C. 154Contents and term of patent; provisional rights.
(d)PROVISIONAL RIGHTS.— (1)IN GENERAL.— In addition to other rights provided by this section, a patent shall include the right to obtain a reasonable royalty from any person who, during the period beginning on the date of publication of the application for such patent under section 122(b), or in the case of an international application filed under the treaty defined in section 351(a)designating the United States under Article 21(2)(a) of such treaty, the date of publication of the application, and ending on the date the patent is issued— (A)(i) makes, uses, offers for sale, or sells in the United States the invention as claimed in the published patent application or imports such an invention into the United States;
35 U.S.C. 163Grant. In the case of a plant patent, the grant shall include the right to exclude others from asexually reproducing the plant, and from using, offering for sale, or selling the plant so reproduced, or any of its parts, throughout the United States, or from importing the plant so reproduced, or any parts thereof, into the United States.
The context of my reply was rebutting the idea that repeating "software is math" is not the best argument. The quote given to support that applied equally to software and hardware, and I attempted to draw a distinction between the two.
I could not possibly care any less about the core "patent" idea worldwide, but this is what I base my argument on, it is biased towards a particular country. "Promote progress" does not necessarily mean competition, but it also does not necessarily mean ROI. You seem to have focused on a single sentence: "So GIF encouraged invention, the "legitimate purpose" of patents."
Patents which affect me are based on Clause 8 of the US Constitution, to promote progress. Whether it accomplishes this the intended way (documenting progress and protecting its use for a limited time) or through the law of unintended consequences is a principled distinction, not a practical one. In the example of GIF, the result was wide adoption of PNG. In the example of PKZIP the result was a less efficient and IIRC more CPU-intensive compression step, so nothing was gained and it's a counter-example to PNG.
Clause 8. The Congress shall have Power... To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.
This is the funny thing. In the "How do I get an entry level programming job" the university you go to seems to make a difference. But when we look at individual cases, we can see clear exceptions where the school does nothing to help the individual. It is only in aggregate that the school's name is valuable. This is partly due to applicant self-selection, and partly do to the entrance process selecting people based on prior performance.
I had the option to go to a school with a minimum requirement of 12 on the ACT, increased to 14 "to stress the importance of academics" or something, coincidentally the same year that they added 2 points to ACT scores pretty much across the board, so my 34 was equivalent to my older brother's 32.
People who don't meet the minimum requirements for Ivy League don't get in... unless your life is exceptional, meaning made up of exceptions.
So did they bend the rules to let him in? Did the top-notch universities help him? Did he tarnish the name of those institutions? There is no obvious answer, but it is obvious that universities only hold prestige in aggregate, not in individual cases. And anyone who makes decisions based on the institution attended is a fool.
If you can get in, networking and cronyism are the benefits of Ivy League education, not the value of the education. Secondary is hanging out with people who are as smart as you are, which is easier when you attend a university which suits you.
A new study led by European scientists presents the most comprehensive analysis of data from the most ambitious survey ever undertaken by the NASA/ESA Hubble Space Telescope. These researchers have, for the first time ever, used Hubble data to probe the effects of the natural gravitational "weak lenses" in space and characterise the expansion of the Universe.
The data was old, the analysis and imaging is new.
The 'Links' at the bottom include the new paper, and the old study. The old press release dated "7-errNoSuchMonth-2007": "News Release heic0701 - First 3D map of the Universe's Dark Matter scaffolding" http://www.spacetelescope.org/news/html/heic0701.html
Slashdot Mirror
I had XP corporate image, and Sharepoint would crash if I tried any of the AJAX style operations. I could right-click and save, or click directly, but using the arrows and checking out crashed every time.
IE developer toolbar and Fiddler installed, I uninstalled both and it kept crashing, then the lease came up on my box and I got a shiny new Vista. Installed both and some other stuff and never had a problem. But it keeps worrying me that IE with Microsoft-only additions had problems with Microsoft's website, manipulating Microsoft's document types.
The problem I believe is the WEBDAV type interaction. When you authenticate, IE lets you do stuff. Then you open the file, and normal browsers would download the file and ShellExecute() to open it (or maintain their own list, but whatever). IE sends the URL to the application (if it's Office type), which has to re-authenticate since it doesn't share IE's session. That way the Office app can check in/out instead of just opening a local copy.
To tell the difference, you can obviously see the normal IE download dialog if it's downloading and opening, otherwise you get the Office dialog that has "Opening [filename..." and only has a cancel button.
I just used FireFox, and tried to avoid checking things in/out. It's all intertwined - impossible to fix. Probably lots of code duplication as well.
"POS Withdrawal" and "EFT" say this is using automatic debits directly to the customer's account, most likely over the ATM network. The ATM network is basically a bunch of switches and routers. Think about it like an HTTP request, where you ask for something and the server returns a code. There are 'OK' codes, 'Not ok' codes, and 'retry, I think something happened but not really sure so just let me know kthx.'
Yeah I'm dumbing it down a bit, but it's like that. All it takes is missing a 'good' return code in a list of switch/case tests, or a bad one, or misunderstanding the documentation and misinterpreting a code, and you get the rare response code 'Successful, and by the way you have lovely eyes' which you misinterpret as 'Retry' instead of 'Successful'.
Having worked in the bank support business dealing with these return codes, we see these kinds of mistakes all the time. It happened about once a month, to different institutions. They called us, we told them what to look for, they fixed it, most customers never even saw the mistake. It's likely that the payment processor screwed up and update/upgrade. First the vendor will sort it out and get money back where it belongs, and in the process most banks will either waive the fees or punch the vendor in the throat and then waive the fees, because they don't want each individual to call them separately and tie up the phones.
To repeat that - for most banks, there will be enough affected users that it's cheaper to refund the overdraft fees than to deal with each customer. For the few that don't, the customer's call will be enough to refund. At that point if you can't get a resolution, EA will probably call with the customer on the phone, and now you have business vs. business and the customer sits back and watches. At least one bank will say 'tough cookies' and end up on The Consumerist.
As always, if you have a credit union instead of a bank, you're going to feel safer when these things happen. Funny thing is, they usually purchase outsourced service to keep costs down, so depending on the vendor chosen you might see this kind of thing more with a credit union - but they are more proactive in resolving it without the customer having to pester them.
It's too bad that every time we invent a new language, we have to come up with a new translation of "caveat emptor" because 2,000 years later consumers still don't pay attention to what they are buying.
Sony wants to make money, and if you are costing them money they will find a way to make money or cut you off. With all of the coverage about consoles being a loss leader for the first generation, there is very little excuse for someone to be ignorant that they are an expense to the company. Especially if you are doing something other than what it was designed to do - play games.
Debian's business model is to provide a quality product that other companies can rip off. So that kinda evens things out.
Canonical has found a way to make money to support its work (and maybe profit too, I don't know), fully respecting the license agreement Debian provided. Not that Debian has much choice if they want to distribute Linux, but it's what they provided. As long as you don't step on trademarks, ripoff isn't even an insult in this context.
http://dictionary.reference.com/browse/ripoff
"3. a copy or imitation."
Sure it has a negative connotation, but that #3 definition is quite neutral in its denotation. I'll assume you're using that meaning, in the given context.
"Speed and uninterrupted use of service not guaranteed." That's what you missed. That's what prevents you from multiplying the bits per second times the number of seconds in a month and coming up with a total number of bits you expect. The disclaimer about which factors may affect your experience should be a dead giveaway. "Actual speeds will vary" means the math is guaranteed not to work out. You bought a speed, not a quantity. If you want a quantity, you're going to have to buy based on quantity. The parts you quoted pretty much defeat your entire argument. You expect, but the contract tells you not to, so that's your problem, not Verizon's.
The plans specifically state that if you are causing network congestion, they have the right to alter your service. You might try to sue them, but if they show you are the statistical outlier and have technical data to back it up, they just take the parts that you quoted and say we sold a speed not a number of bits.
And don't start with that contract of adhesion shit from down below, the adhesion idea is that your are stuck and can't get out. A monthly plan can be terminated, even if you have to repay the hardware subsidy in the form of an early termination fee, which you agreed to in your initial contract and has been upheld in court cases where the termination fee was clearly stated. Even if you have no other provider to choose from you aren't stuck in that contract, you could go without or get a different contract/plan now that you know what to look for.
Leaving your kids at home alone so you can go have sex and drugs is one thing. Ensuring they cannot harm themselves while you're away is thoughtful and deserves some consideration.
Ensuring they can't hurt themselves because their fingers are trapped in a closed, locked door to which you have the only key is... I can't even find the words. Anything I say would not be strong enough, I think it speaks for itself. Just thought I'd share.
I had GTA Vice City for PS2, and an icestorm hit. My car was frozen in place, and I had a long weekend. So I just let it melt. Meanwhile with nothing better to do or no place to go, it was 3 days of GTA. I played in the virtual rain a lot and learned a lot about sliding around. I think it was one of the driving course things where you have to make it cross-island and back under a time limit, and I kept having to re-try.
Thaw came, I got in my car to go to Burger King. Pulled out the driveway, blew through 2 stop signs right in a row (within 20 feet of each other) and got pulled over in less than 50 feet since it was right in front of a cop car.
I only escaped a massive ticket because some fire or shooting required his presence more than little old clearly sober me. "You didn't even try to slow down, for either one of those," he said as he hopped back in his car.
Now every time it snows, I get in my 6-speed and drive around for fun, the more sliding the better. Cool story bro, there I saved you the time.
How does the thread know how much to reserve at startup time? Unless you know exactly how much you'll need, you'll run into the problems I mentioned. The specific problem being solved involves a program that makes heavy use of memory allocation, so you probably can't count on knowing how much you need and you'll have to go back and get more. More specifically, this is intended for single-threaded application running on multi-core processors.
Atomic instructions are basically system-wide critical sections, and heavy allocation will make this a bottleneck for the application (not necessarily the system). For most normal programs with occasional allocation it's not going to matter how you implement it.
Windows already does exactly this, reserving 1MB by default and this is controllable within the executable or when creating threads, so if you do know what you need you can certainly do this at thread level. So yes, it would be trivial to write a lockless malloc where the thread controls its own allocations and free list. There just isn't a benefit to doing that, and that's not the problem being solved here.
http://msdn.microsoft.com/en-us/library/ms686774(VS.85).aspx
I'm not defending the application. The claims leave this open to being implemented manually, and the supporting documentation makes it clear that the claims do not match the described invention. We should be objecting to this patent, but not based on the description in the summary. No one here reads the patents, so I wanted to make sure people noticed that the invention itself is not trivial. The patent application is unfortunately just as terrible as the summary of it.
If we are going to have a discussion about a patent, I'd prefer not to wade through hundreds of comments which are completely off-topic, simply because the summary was awful. These bad summaries really make the patent process and applications as far more evil than they actually are, which is quite a feat since in reality it is quite bad without the additional propaganda. We need to fix things that are broken, not simply fuel the fire with misleading garbage. I actually wanted to moderate everyone's misreading as off-topic, but they probably wouldn't understand since there is no "-1 You are objectively incorrect and here's why" moderation option. But I digress.
IBM is not optimizing by trial and error as the summary states - it is preventing performance impact by temporarily reverting adverse refactoring using objective data, according to the claims. Trial and error implies it's just randomly looking for optimizations, but the purpose of this has nothing to do with optimization. It is temporarily reverting specific refactoring. "IBM Patents Optimization" is not remotely close to the actual claims, even just the subset quoted in the summary. I would have maybe said "IBM appears to want to patent undoing refactoring which negatively affects performance." That's what the actual claims are, and it's clearly invalid unless everyone who refactors code does no performance evaluation ever. With many tools, performance testing isn't even part of the refactoring workflow, so it could be a legitimate addition to refactoring tools if implemented as described (not as claimed).
At the very least, claim 3 "undoing refactoring is performed before the computer code is compiled" should be clarified to say it is automatic and temporary, or another claim added "The method of claim 1 wherein the undoing refactoring is automatic and temporary". That would fit the description and properly narrow the scope of the invention.
I would assume that patent law considers the claims to be the determination of whether something is infringing or not, and if that is the case this patent application should be rejected. Whether IBM did this intentionally or not doesn't matter, the patent examiner should catch this. Since it looks like an application and not a granted patent, I'd hope the examiner will probably ask for an update.
This is the invention in a nutshell, which as you said is unrelated to the claims. Hopefully we can now get input on the poor quality of the patent application, and discuss possible prior are which invalidates the application (or a genuine "good job guys" if deserved) :
Forget about optimizing at all, consider all optimizing turned off. This invention basically makes sure that your refactoring doesn't impact performance. The compiler can then optimize the rearranged code instead of the original code you're looking at. Optimization will benefit both inputs, but the un-refactored code can still run a lot faster.
An optimizing compiler chooses its optimization from the tricks it knows about. Automatically inlining frequently called functions would be something that an optimizing compiler could do, but that is not refactoring the code - it is the opposite of classic refactoring. So to answer your particular question no, you can't say that. Auto-inlining would be an example of the simplest implementation of this invention but hardly representative.
The essential difference here is that the user chooses to rearrange the code so it is more understandable. An optimizing compiler would rearrange the code so it is faster. The purpose of rearranging is different. The invention in this case is tracking which refactoring was beneficial to performance and which was a negative impact, and un-doing the refactoring with a negative impact while preserving neutral or advantageous refactoring. Specifically, this happens behind the scenes so you don't have to even know which changes were bad.
To put it another way, a normal compiler has a set number of tricks up its sleeve, and it applies those tricks when it thinks it should be beneficial. This invention adds two additional checks. First, it doesn't apply optimizations, it only reverts code on the fly *before* hitting the compiler. Second, it relies on performance data to choose which reverts to apply. I've never seen an optimizing compiler which builds the executable, tests it, benchmarks it, and goes back and updates its optimization strategy based on the results. That's what IBM are claiming. An existing optimizing compiler would do the best it could to optimize what you give it, not revert entire blocks of code. This invention would decide based on performance which version to use.
The other thing I forgot to mention is how it deals with bug fixes. I assume you would go through the 'Refactoring Wizard' for refactoring work, and just change the code directly for bug fixing. But what if the buggy code ran faster than the refactored but bug-free version? The implementation behind this would have to be very interesting in order to not completely hose your app, and that is why this is impressive. They actually think they can do this.
The obvious application is one of the UML-to-code-to-UML IDE's, including Rational, using Agile type methods. A few clicks to refactor, and when the human makes the code worse the wizard ignores the human, but pretends to obey.
You're funny, but the video does not lead to the conclusion that game skills don't apply. An interesting experiment, but real life is not sitting in a dark cabin looking at an overhead monitor. We would have to compare driving skills of gamers and non-gamers, with similar experiences (accounting for country road drivers vs. city drivers), in order to even come close to such a conclusion.
Flight sims used to be very unrealistic, but they were still used and effective at giving people practice in uncommon situations. They have improved of course. Video game driving has probably taught a lot of people about bad weather conditions, where you would never experiment with your tangible car.
This is not science, it's not a conclusion, it's simply entertainment.
This is not a method of gradually figuring out how to boost performance by trying different things.
This invention makes changes to the source code on the fly. The purpose is to allow programmers to re-factor code for readability, but un-do the process before compiling. Presumably this is built in to one of their automatic refactoring applications or IDE. It is about fixing the problems refactoring causes, specifically in this case performance issues. I assume it could work like automatic inlining of small, commonly used functions, but is open to be a lot more complex.
You click the buttons on the GUI, it refactors some things for you and maintains a history of those actions. It stores the resulting code, which you send to source control or final build kit or whatever. When building, it examines which changes had a negative impact on performance and reverts those before compiling.
To the user, the code is refactored. To the computer, it is not.
My question is, how the hell do you debug this kind of output, or store it in source control? You are clearly going to be locked in to the same IDE, and probably same IDE version, and have to check in all of the project files. Cross-compilation is going to be difficult at best, since compiling the raw code won't have the performance enhancements. Clever for a closed-process shop, but kinda impractical.
What will be more interesting is whatever people come up with to deal with the misery this will create.
Your post references java 4 times, giving it a further boost, so maybe we should re-check the stats now?
Wait wait, you also reference C 10 times so you're probably the one bringing the java average down. Everyone quit typing!
I enjoy customizing my apps by sub-classing windows or whatever else is needed.
I fixed a particularly stupid application for work which requires you to click on a calendar and choose a date, and again on another calendar. Clicking OK updates the data source, but clears all the values. So every day when you're going through today's data, you select both calendars and click OK and re-set both calendars. Lots of wasted time. I updated the button click to save the calendar date, call the normal click function, then restore the calendar. Took a 4 hour process down to 20 minutes.
IDA helped confirm a few things, but mostly it was just Spy++ from MSVS 6 that helped me figure out what was needed. I could have gotten all cool and stuff, but I just put an injector in the same folder, and a .dll which gets loaded by the injector. Open the app and open the injector and enjoy your efficiency. The author would not consider making an update so I did it for them.
No hack, no crack, no DRM, no derivative works because I was simply calling functions which the application made available.
For those who use Avaya CentreVu, I screen-scraped the agent display so it could send alerts when threshold settings were hit (beep on yellow, send a page on red). Again just using existing functionality and an injector.
I also remember an assembly "reversing" tutorial which added line numbers to Notepad, but that was invasive because it altered the underlying code. I always wanted to go back and do it using a less invasive method, but I never thought about simply grabbing the pixels and shifting them, adding appropriate line numbers based on the scrollbar window properties.
This application of an ages-old technique is particularly interesting because it provides legal coverage for compliance with the ADA, if you happen to be subject to that. Clever.
We did this thing where we tracked estimates heavily, with a target of +/-15% on estimation. Estimate correctly and you can set expectations on the release date of the next patch or update. Our process involved making the estimation, doing design, build, testing, sign-off, and project clean-up. Yeah we borrowed time from other slots and in general retroactively updated the estimations so they were right. When asked for a date, we'd say something reasonable. Then when the top execs started looking closely at a few deadlines, we just multiplied the work estimates by 1.5.
They were honestly more worried about meeting an arbitrarily set deadline than being efficient. If I said 2 months and it took two and a half months, we were in the shitcan. If we multiplied it and said 3 months, we took 2 months and 3 weeks and goofed off for a week essentially - take off early a few days or cancel meetings and just zone out. 3 months, just like we said, we made our goal.
If you're doing development, you're doing something you have done before, in which case you should be able to copy and paste, or you're doing something new. The third option is team-wide incompetence where you keep repeating yourself, but those teams are a lot better at estimating. They know exactly how long it takes because they do it all the time.
I've worked in open source, and since I'm writing it for free you'll get what I give you when i give it to you. More businesses need to be like that, but they can't because clients won't agree to pay for anything unless they can get it on a schedule.
Why do we have cost overruns on IT projects which become legendary, like the FBI information sharing software and the child support system that requires operators to fake data so children don't get forgotten? For that matter, why do we have tangible projects like the southern US gate which cost a billion dollars and was so poorly mismanaged that the project just got halted?
Poor estimation is pretty common unless you're doing assembly line type or repeatable work, but when you break new ground it's harder to know what to expect. Open source mindset is more honest about that. Because it can. Walk in their shoes for a week and become a better manager and those open source guys will blossom. That or they are part of the emo gimme generation and you might as well take them out back and show them which grave they will fill if they don't quit whining. That's just fair.
Have every thread allocate its memory from... what? At some point either the operating system or the runtime has to lock something so it doesn't get interrupted, or turn off all interrupts and switching for a few hundred cycles so it can do the allocation. Usually the runtime requests reserved pages far in excess of what it needs, and then doles out pieces, committing them as needed. You need 2k, so the runtime gets 4MB and commits 32k page(s). Next time you need more, then the runtime just returns more of the 32k block.
The operating system has to lock its list for the runtime, and/or the runtime does the same for the program. Someone's locking something somewhere.
He's taking pictures for fun and sharing them. Not for science, but for posterity or public interest, or "cool, look what I got to see". He's posting them for free. From the FUCKING SPACE STATION. 200 miles straight up in the air, traveling at 27 THOUSAND miles per hour, and is able to take pictures and post them on the internet. "Could you please move your hat sir, you're obscuring my view," is a reasonable request. "Could you please move your 300 tonne vehicle out of the way so I can see a few blurry colored pixels that a small part of your SPACE STATION is blocking, without which we wouldn't be able to see the picture at all," this is ridiculous.
Thank you Soichi, thank you from 6.5 billion people on planet earth, except for "Ungrounded Lightning". For some reason he thinks he's an art critic, or the Space Lord Mothermother, or that you can lift an X-wing out of a swamp with the wave of your hand.
Hopefully this puts it in perspective (choose one)
http://www.youtube.com/watch?v=RhWZ4gNqPP0
http://www.blogjam.com/neil_armstrong/
The fact that I cannot kick Microsoft in the balls for Vista (Required by my employer until 7 is tested completely) suggests that companies are not people. We cannot figure out which of the many employees deserve a ball kicking and which were simply following orders, so we can't just kick them all and be fair about it. So we have to fall back on holding the decision makers responsible, or this nonsense will never end.
You're misreading this, I think. Or taking it out of context. Ignoring the second part we have:
That singles out people who willfully infringe copyright for financial gain. You are assuming that downloading a copyrighted work gets you the financial gain mentioned in the definitions. If you take a look just at 506(a), it really reads this way, expanding to clarify:
Replacing the definition for the defined word, it becomes:
In this reading, simple receipt of a copyrighted work is NOT A CRIME. You would have to infringe copyright for purposes of receipt of other copyrighted work to qualify as a criminal. This specifically targets people who trade "warez", and people who upload and download at the same time.
Under your reading, if someone gives me a copy of a movie on DVD or CD I have received financial gain. Ihe problem is, I did not infringe anything for the purposes of financial gain. I did not receive stolen property. How is this a crime?
If I give the copy away (without making one for myself) I still have not infringed. If I trade it with someone for some other movie, I still have not infringed, even though I did receive other copyrighted works, because I did not infringe initially. I did not make the copy.
If I download from Rapidshare, I'm not the one making the copy. I would have to make another copy somehow, and intend to benefit financially (where financial gain is defined).
sure you can argue a different reading, lawyers do it all the time. But unless you have court cases citing how a judge or jury read the law, your post was the most significant, densely packed pile of horseshit I've seen this week. you chose your own reading of the law to support your position.
Effectively, your position is that downloading something for the purposes of financial gain includes the financial gain you acquired by not paying for the download. But they put that word "other" in there. Other copyrighted works. Not trading pieces of something for other pieces of it, but trading something in whole for something else in whole.
I could infringe, and then turn around and trade an empty disk for a full disk, and argue I'm within the law, since the net gain to myself was zero (disk for disk) and I actually took a loss due to the internet cost.
On one hand, the analogy was flawed and had to be corrected. On the other hand, the explanation was poorly done. A better explanation would be that people need to learn things about their browser in order to use it effectively. Like "too good to be true" probably means it's not true. Or there ain't no such thing as a free lunch. Common sense that says don't take unknown things from unknown people. That's what people forget - no other application has opened people up to identity theft just by operating it. Since a browser uses so many external files, it's the exception that no one thinks about.
What's the penalty for not knowing? In the car analogy, safety of the people you might otherwise drive over, or yourself, should be the motivator. People drink and drive, they text and drive, they don't pay attention, or they aren't familiar with their vehicles. Most people do not fall in this category, a few do. The penalty is in a few cases someone successfully installs a spambot the user will never notice. In fewer cases the user's personal files get transferred, and some of those get used and credit card companies block cards due to suspicious activity and a few people lose money. If it were a big problem, we'd hear stories every night on the news, but it only comes up a few times a year.
There's no incentive to learn because 1) it's rare and 2) learning is not a requirement, as it is in a driver's license test. This is where the "Internet driver's license" idea makes sense, until we realize how impractical it would be. Then we're back to the situation where people should learn, but don't, and it's only a problem for a few people a year.
You kinda have a point right up to where you fail to see that bullying behavior is typically learned. Someone has probably already shown them how small they are in the universe and now they are doing the same, once they find someone smaller to pick on.
It doesn't matter how it ends, the bully has to learn that he can't get away with that behavior. Either by having the crap kicked out of him or by being not just punished but actively monitored to ensure he doesn't violate "probation". He just needs to know that if he tries something, something negative will happen.
An equal fighting back will not convince the bully, since some fights will be won and some lost. It's the psychology of slot machines - win a few and you think you'll win the next one as well. The more ingrained the behavior it is, the longer it takes to extinguish. A severe beating can accomplish this, or several attempts at restricting privilages (which involves parents, who may not even care). It's faster to beat the kid, but it might actually feed the need to feel superior and actually work against extinguishing behavior in some cases. It's difficult at best to decide which case will work unless you learna s much as you can about the particular bully, and put aside generalizations.
http://www.uspto.gov/patents/index.jsp
If you're going to refute something that is over your own head, a citation of some sort helps. "Making" is the first word in the list of forbidden things. That's the patent office website. Go find the actual law and see if it supports your comment.
http://www.uspto.gov/web/offices/pac/mpep/consolidated_laws.pdf
35 U.S.C. 154Contents and term of patent; provisional
rights.
(d)PROVISIONAL RIGHTS.—
(1)IN GENERAL.— In addition to other rights provided by this section, a patent shall include the right to obtain a reasonable royalty from any person who, during the period beginning on the date of publication of the application for such patent under section 122(b), or in the case of an international application filed under the treaty defined in section 351(a)designating the United States under Article 21(2)(a) of such treaty, the date of publication of the application, and ending on the date the patent is issued—
(A)(i) makes, uses, offers for sale, or sells in the United States the invention as claimed in the published patent application or imports such an invention into the United States;
35 U.S.C. 163Grant.
In the case of a plant patent, the grant shall include the right to exclude others from asexually reproducing the plant, and from using, offering for sale, or selling the plant so reproduced, or any of its parts, throughout the United States, or from importing the plant so reproduced, or any parts thereof, into the United States.
The context of my reply was rebutting the idea that repeating "software is math" is not the best argument. The quote given to support that applied equally to software and hardware, and I attempted to draw a distinction between the two.
I could not possibly care any less about the core "patent" idea worldwide, but this is what I base my argument on, it is biased towards a particular country. "Promote progress" does not necessarily mean competition, but it also does not necessarily mean ROI. You seem to have focused on a single sentence: "So GIF encouraged invention, the "legitimate purpose" of patents."
Patents which affect me are based on Clause 8 of the US Constitution, to promote progress. Whether it accomplishes this the intended way (documenting progress and protecting its use for a limited time) or through the law of unintended consequences is a principled distinction, not a practical one. In the example of GIF, the result was wide adoption of PNG. In the example of PKZIP the result was a less efficient and IIRC more CPU-intensive compression step, so nothing was gained and it's a counter-example to PNG.
This is the funny thing. In the "How do I get an entry level programming job" the university you go to seems to make a difference. But when we look at individual cases, we can see clear exceptions where the school does nothing to help the individual. It is only in aggregate that the school's name is valuable. This is partly due to applicant self-selection, and partly do to the entrance process selecting people based on prior performance.
I had the option to go to a school with a minimum requirement of 12 on the ACT, increased to 14 "to stress the importance of academics" or something, coincidentally the same year that they added 2 points to ACT scores pretty much across the board, so my 34 was equivalent to my older brother's 32.
People who don't meet the minimum requirements for Ivy League don't get in... unless your life is exceptional, meaning made up of exceptions.
http://www.monkeydyne.com/bushresume/early.html
So did they bend the rules to let him in? Did the top-notch universities help him? Did he tarnish the name of those institutions? There is no obvious answer, but it is obvious that universities only hold prestige in aggregate, not in individual cases. And anyone who makes decisions based on the institution attended is a fool.
If you can get in, networking and cronyism are the benefits of Ivy League education, not the value of the education. Secondary is hanging out with people who are as smart as you are, which is easier when you attend a university which suits you.
The article dated "March 26, 2010":
http://news.discovery.com/space/hubble-3d-map-universe-dark-matter.html
has a source dated 25-Mar-2010::
http://www.spacetelescope.org/news/html/heic1005.html
with this quote which explains everything:
The data was old, the analysis and imaging is new.
The 'Links' at the bottom include the new paper, and the old study. The old press release dated "7-errNoSuchMonth-2007":
"News Release heic0701 - First 3D map of the Universe's Dark Matter scaffolding"
http://www.spacetelescope.org/news/html/heic0701.html
Is the one described your original article:
http://news.bbc.co.uk/2/hi/science/nature/6235751.stm