And mid-2006, XP is five years old. That's close to twice the age of 98 when XP was released (and remember 98SE and Me in-between, even if they didn't tout upgrading to those that much).
Windows 3.0 (not 3.1) was as old when Windows 95 was released that it looks like XP will be when Longhorn is released. It's not exactly a once-a-year rush.
I think their "story" is that System.Windows.Forms has good support, maybe a bit better than plain GDI (assuming a bit more about how the stuff in SWF actually accesses things), but Avalon is still fully on its way. It's bad while Mono is making progress, but I would say that a serious rethinking of the drawing model on Windows is needed to get on par with MacOS X in these aspects. They might accomplish that, if it's done properly.
One point is that when/if 1.3 billion Chinese are up on a close-to-Western standard of living, that will be a huge economic power. The sheer number would make it quite impressive. Japan, on the other hand, is a quite different thing -- to exceed the U.S. in importance the U.S. would have to break.
I just love being modded redundant for being the first one to point out the obvious costs of stopping in orbit. The first answer was purely around the problem of synchronizing funds. Bah, mod points, give them all to me!
If it was embedded in a more general executable, I don't see the strange thing about it. After all, much spyware is just binaries that the user is made to load, by an exploit or by a bit of social engineering.
1. Send up a probe, using a lot of fuel to put it in an ISS-friendly orbit.
2. Add enough fuel to accelerate from that orbit into a completely different trajectory to reach Mars sometime soon.
3. ?????????
4. Profit!
Also, I wouldn't be so sure that the probe wouldn't benefit from extra protection during the heavy acceleration when leaving Earth orbit and possibly during the voyage itself (more of a radiation matter, there).
The most important point -- it is not cheap to bring up a lot of fuel into space. Currently, you should better be on the right track from the start. Going into orbit and then leaving it is not a way to do so.
It may be true for rover activity, but I seriously doubt that orbital surveys, radar and photographic, would be done better by manned missions (in orbit?!). The communication and weather satellites around Earth are not manned and they do their job, giving significant information about our planet, which we wouldn't have by just staying on the surface and which would be quite dull to manage from orbit.
Even with an ambitious manned mission, the coverage of the planet would be "spotty". Automated studies give us the broader view.
OTOH, it's well-known that hearing implants on from-birth deaf children are more successful if put in place early on. The real test of adapting to random data would be putting a microphone in the eye and start sending away Fourier transformed sound data and watch if any patient developed hearing through that thing. I really doubt it.
It is more likely to find organical life on the surface of Sol, than on those planets, radiation-wise. Any self-reproducing structures to be considered life would make carbon-based and for example silicon-based life (if it exists) so much alike that they would be indistinguishable, while this would need to be a totally different type.
So, the problems continue with IIS, and are not simply because there was a buggy version with NT4. Although, I do agree that it's a target because it's in the default install, just like OpenSSH is a huge target because it's running on so many Linux boxes.
Well, even Windows 2000 vanilla is getting 5 years old now, but I of course don't know when this happened. There have been service packs to slipstream for years and Windows 2003 also for a number of years. And, of course, that doesn't excuse the bug itself. But the severity is caused by the huge number of computers, never really intended to be real web servers, which have been turned into zombies. As noted here and in the grandparent, we're not interested in market-share for real servers, just the number of hosts possible to reach in an attack. I bet most Apache and IIS setups don't host any real site, but are just there to toy with or by "accident"/"oh, how cool, a web server" interest.
All so true. Even remembering the need to code for a 64k segment based real mode model gives some good (and some bad) habits regarding saving memory. I think the worst problem these days is when coders don't realize that they do really stupid things, like using random access on a linked list or implementing their own O(N^2) sorting, or what's basically equivalent to a sorting, without realizing it.
The only thing that's worth than that is when the same coders think that you benefit by instantiating a complex data structure just to store five-or-so elements. Or saving memory by a cache implementation with 50-200 % overhead on the objects cached (unless they are really expensive to create).
How do you get a big Word document by just using the feature-equivalent of Mac Write? No embedded pictures, no macros, no fancy stuff. An empty document is 20 K, yeah, that's a level of bloat, but you can add a lot of text just to reach 30 or 40.
You easily run out of space on a floppy quite easily, even by just writing lots of source code, especially properly documented code. (No-tabs will of course add even more, crappy coding "standards"...)
You're forgetting the effects of increasing data density, alleviating the sequential read problem. And if you only use the first few tracks, you get somewhat lower access times, too. The increase in size is not a problem. Increasing fragmentation in disk access patterns is problem, but one fact is that with more CPU cycles/disk access, we can spend more time on smart prediction and pre-fetching, as sequential speed is still acceptable.
RAM has increased tremendously since 1998, although not as fast as CPUs. Look at dual channel DDR with a base frequency of 200 MHz compared to the brand new 100 MHz SDRAM of the time. I also would really think that most 7200 RPMs have made some significant inroads during these seven years.
Exactly -- I could understand concerns against logging of which public domain material I download (a truely legitimate P2P use) so I would want a reputable service that I could trust would not abuse the data I create about me. But, unless I knew the maintainer, that would probably mean that the service would be on the radar, not below it...
Is the compiler/VM smart enough to unwind it properly for an array, compared to other collections? foreach in.NET doesn't bound-check in the same way as other accesses, but it still manages to be slightly slower than a common for loop for an ArrayList or similar. I haven't benchmarked the Java 5 implementation. Have you?
Those sites, IIRC, only add a row in their DB and encodes the primary key of the row into the non-negative numbers and letters. You can't make up the tinyurl for a certain URL on the fly without first contacting their server to register it.
That said, I don't see why this would be patent-worthy. Neater than long floating point numbers, yes, special compared to hex or anything, no.
Wouldn't Gilette be more like you have a "cheap" space telescope that you need to go up and service regularly at a higher price than new, equivalent, hardware. Just attach some moniker like "Mach" or "Turbo" or something to it.
So, no, Gilette should have sponsored the original Hubble telescope, with extra payouts at each service mission.
Ah, ok, that's one place where it would make sense. Of course, maybe both Copy and Paste special would make more sense, as you sometimes know that you want to do something special when copying, sometimes it's more a function of the environment you are pasting into.
I know that there are some keyboard shortcuts for doing strange things in the RTF based rich text control (like accidentally getting Right-to-left text...), but I guess there is none for "paste text", and that it wouldn't work in IE anyway.
On the other hand, I don't think many types of protective casing would have saved CDRs with bad dye, just like not much has saved some of the twenty+ year old 5 1/4s. Have you actually made a CD unplayable by dropping it? Dropping it on what?
Well, if it is only a number describing read-transfer speed from the SRAM cache, it says nothing at all of the underlying media. I would assume that it's some kind of sequential or burst rate that would actually involve the interface. As I said, I've seen no single disk drive giving numbers that would really match those. On the other hand, they should have been matched and exceeded by the time this technology actually gets into real products, if that ever happens.
Lots of spyware likes the Browser Helper Object infrastructure (BHO). It's what the Google toolbar and other stuff uses in IE to get hold of the DOM and get events about what's going on. Anyway, in old IE versions (5.0 or something), you could only ever use ONE BHO. If you added several in the appropriate registry key, only the first one enumerated was actually loaded.
So they've gone to great lengths to improve spyware interoperability! Kudos for that!
Slashdot Mirror
Windows 3.0 (not 3.1) was as old when Windows 95 was released that it looks like XP will be when Longhorn is released. It's not exactly a once-a-year rush.
I think their "story" is that System.Windows.Forms has good support, maybe a bit better than plain GDI (assuming a bit more about how the stuff in SWF actually accesses things), but Avalon is still fully on its way. It's bad while Mono is making progress, but I would say that a serious rethinking of the drawing model on Windows is needed to get on par with MacOS X in these aspects. They might accomplish that, if it's done properly.
One point is that when/if 1.3 billion Chinese are up on a close-to-Western standard of living, that will be a huge economic power. The sheer number would make it quite impressive. Japan, on the other hand, is a quite different thing -- to exceed the U.S. in importance the U.S. would have to break.
I just love being modded redundant for being the first one to point out the obvious costs of stopping in orbit. The first answer was purely around the problem of synchronizing funds. Bah, mod points, give them all to me!
If it was embedded in a more general executable, I don't see the strange thing about it. After all, much spyware is just binaries that the user is made to load, by an exploit or by a bit of social engineering.
1. Send up a probe, using a lot of fuel to put it in an ISS-friendly orbit.
2. Add enough fuel to accelerate from that orbit into a completely different trajectory to reach Mars sometime soon.
3. ?????????
4. Profit!
Also, I wouldn't be so sure that the probe wouldn't benefit from extra protection during the heavy acceleration when leaving Earth orbit and possibly during the voyage itself (more of a radiation matter, there).
The most important point -- it is not cheap to bring up a lot of fuel into space. Currently, you should better be on the right track from the start. Going into orbit and then leaving it is not a way to do so.
Even with an ambitious manned mission, the coverage of the planet would be "spotty". Automated studies give us the broader view.
OTOH, it's well-known that hearing implants on from-birth deaf children are more successful if put in place early on. The real test of adapting to random data would be putting a microphone in the eye and start sending away Fourier transformed sound data and watch if any patient developed hearing through that thing. I really doubt it.
It is more likely to find organical life on the surface of Sol, than on those planets, radiation-wise. Any self-reproducing structures to be considered life would make carbon-based and for example silicon-based life (if it exists) so much alike that they would be indistinguishable, while this would need to be a totally different type.
Did that turn Earth-Sol into a binary system in the 1950s?
It's strange how hard it is to relate to a text of this type, I'm somehow distracted by the almost identical lines above and below and I get lost.
The only thing that's worth than that is when the same coders think that you benefit by instantiating a complex data structure just to store five-or-so elements. Or saving memory by a cache implementation with 50-200 % overhead on the objects cached (unless they are really expensive to create).
You easily run out of space on a floppy quite easily, even by just writing lots of source code, especially properly documented code. (No-tabs will of course add even more, crappy coding "standards"...)
RAM has increased tremendously since 1998, although not as fast as CPUs. Look at dual channel DDR with a base frequency of 200 MHz compared to the brand new 100 MHz SDRAM of the time. I also would really think that most 7200 RPMs have made some significant inroads during these seven years.
Exactly -- I could understand concerns against logging of which public domain material I download (a truely legitimate P2P use) so I would want a reputable service that I could trust would not abuse the data I create about me. But, unless I knew the maintainer, that would probably mean that the service would be on the radar, not below it...
Is the compiler/VM smart enough to unwind it properly for an array, compared to other collections? foreach in .NET doesn't bound-check in the same way as other accesses, but it still manages to be slightly slower than a common for loop for an ArrayList or similar. I haven't benchmarked the Java 5 implementation. Have you?
That said, I don't see why this would be patent-worthy. Neater than long floating point numbers, yes, special compared to hex or anything, no.
Please don't tell me that this new "Hubble A" will go out to seek God, while hunted by klingons and piloted by Spock's (half-)brother.
So, no, Gilette should have sponsored the original Hubble telescope, with extra payouts at each service mission.
I know that there are some keyboard shortcuts for doing strange things in the RTF based rich text control (like accidentally getting Right-to-left text...), but I guess there is none for "paste text", and that it wouldn't work in IE anyway.
On the other hand, I don't think many types of protective casing would have saved CDRs with bad dye, just like not much has saved some of the twenty+ year old 5 1/4s. Have you actually made a CD unplayable by dropping it? Dropping it on what?
Well, if it is only a number describing read-transfer speed from the SRAM cache, it says nothing at all of the underlying media. I would assume that it's some kind of sequential or burst rate that would actually involve the interface. As I said, I've seen no single disk drive giving numbers that would really match those. On the other hand, they should have been matched and exceeded by the time this technology actually gets into real products, if that ever happens.
By forking networks, they create more room for possibly redundant chatter without anyone noticing.
So they've gone to great lengths to improve spyware interoperability! Kudos for that!