Hmmm. I would think it would be a lot easier to do the opposite. That is, only use CISC instructions that are roughly the same as the RISC ones. You know, moves, loads, stores, etc. Then, I think this debate is a few years out of date. Both architectures have lots of complicated instructions, despite the nammes. The register difference may be tough.
Looking at one application like PearPC doesn't prove anything at all. Maybe the application is poorly coded.
Good point, I thought that was stupid criteria too. That said, I don't necessarily think that email isn't an addiction. How about this for a symptom - do you read your email as soon as you get a new one, regardless of your current task? I generally do, even though I think it negatively affects my productivity.
Slightly off topic - did you know there is an 'eieio' PowerPC instruction? Enforce In-order Execution of I/O instructions, I think. I found that amusing.
I think the vast majority of comments are are missing the mark. There seems to be a lack of knowledge about FPGAs. It's not self-aware, and doesn't really configure itself per se. FPGAs are just easily-customizable hardware instead of spending $300k on an ASIC. You "program" the hardware with a Hardware Description Language like VHDL or Verilog. So if you can have custom processing blocks of HDL and dynamically load them onto the FPGAs, I guess it's like "wiring itself". But conceptually it isn't a lot more advanced than say.... dynamically loadable modules. Just that it controls hardware instead of software.
I'm curious about something. If you just put in a dual-core chip is stuff going to run faster? I assume there needs to be a lot of OS support. Is it the same as SMP support or different? Do Linux/Windows/MacOS X already have dual-core support? I assume to take advantage of dual-cores at an application level it is necessary to be elegantly multi-threaded?
Furthermore, what's the difference at the chip level? Are there hardware cache coherency controllers? Are they fast? How does the effective decrease of memory bandwidth affect performance?
OK, I guess I'm really looking for a good technical overview of dual-core technology. I just don't know enough about it. The Wikipedia page isn't bad but it's pretty high-level.
I could be wrong here, but I don't think that is a great example. I think the Chicago Crime hack is trying to use the Standalone Gmail Maps hack which is known to not work on Safari yet. Google Maps itself works fine, this is just someone trying to reverse engineer Google Maps and not having it figured out all the way. It's definitely not right to pin that on AJAX in general.
I'd also like to add that "convergence"-like devices (I hate that usage) seem to be only popular with Slashdot-like geeks. People who would buy the Treo. The same kind of people who would prefer an iRiver with more features to the iPod. The same kind of people who care about Ogg. The same kind of people that don't have much effect on the market. Apple has a good idea what the mass market wants. Hardcore tech geeks don't. It sounds like BillG is being a hardcore tech geek.
Good point. People have been predicing "convergence" will be the next thing for a good 5 or 10 years now, but it hasn't happened. Things like the Treo just never took off. I'm intruiged to see BillG pushing it. I think a lot of people prefer the Unix model where each thing does one thing and one thing well. How do you make a phone that plays MP3 with a good interface? And long battery life? And good form factor? The only company I give a chance of designing one properly is Apple, and I still think it would be hard.
When asked for 4 words to describe it, she chose "Bloody Hell - it's good!". Not perfect granted, but good. Even Ebert's 3.5/4 star review is 80% criticisms.
It wouldn't be too hard to make RSA encryption take the same amount of time no matter what code path is used, and to make its memory access patterns uncorrelated with the keys (perhaps by using randomization during allocation)
Allegedly the OpenSSL code has over 1000 data-dependent if-statements in the RSA algorithm. That would certainly not be very easy to make data-independent.
That said, at least you actually understand the problem unlike many of the rabid Linus defenders here.
There are serious problems with your endearing allegory.
1. Microsoft is just as potentially vulnerable as Linux. Their dirty laundry just doesn't get aired in public.
2. The fix is non-trivial and non-obvious. It's not a simple buffer overflow. Any patches are likely to have serious repercussions on kernel performance. e.g. disabled HT, ensure only two threads of the same U.I.D. are scheduled on the same processor, flush cache at every context switch, etc. It looks that Linus is unwilling to accept them unless this vulnerability moves more from the theoretical to proven.
You are calling the icons the graphics? The guy single-handedly wrote the first ever GUI routines for a consumer product from scratch in assembly, and you say Susan Care did the "graphics"? Wow.
Slashdot Mirror
Good point. What does Intel have in the 64-bit range? Isn't most of the 64-bit stuff AMD right now?
I think (s)he was referring to the 68k to PPC transition, not the OS9 to OS X transition.
http://www.macworld.com/news/2005/06/06/liveupdate /index.php
Looking at one application like PearPC doesn't prove anything at all. Maybe the application is poorly coded.
Good point, I thought that was stupid criteria too. That said, I don't necessarily think that email isn't an addiction. How about this for a symptom - do you read your email as soon as you get a new one, regardless of your current task? I generally do, even though I think it negatively affects my productivity.
Slightly off topic - did you know there is an 'eieio' PowerPC instruction? Enforce In-order Execution of I/O instructions, I think. I found that amusing.
I think the vast majority of comments are are missing the mark. There seems to be a lack of knowledge about FPGAs. It's not self-aware, and doesn't really configure itself per se. FPGAs are just easily-customizable hardware instead of spending $300k on an ASIC. You "program" the hardware with a Hardware Description Language like VHDL or Verilog. So if you can have custom processing blocks of HDL and dynamically load them onto the FPGAs, I guess it's like "wiring itself". But conceptually it isn't a lot more advanced than say.... dynamically loadable modules. Just that it controls hardware instead of software.
Furthermore, what's the difference at the chip level? Are there hardware cache coherency controllers? Are they fast? How does the effective decrease of memory bandwidth affect performance?
OK, I guess I'm really looking for a good technical overview of dual-core technology. I just don't know enough about it. The Wikipedia page isn't bad but it's pretty high-level.
I could be wrong here, but I don't think that is a great example. I think the Chicago Crime hack is trying to use the Standalone Gmail Maps hack which is known to not work on Safari yet. Google Maps itself works fine, this is just someone trying to reverse engineer Google Maps and not having it figured out all the way. It's definitely not right to pin that on AJAX in general.
Yeah, with one key difference. Those other companies are actually delivering things that work, today. Microsoft is only promising them Real Soon Now.
What is this Open Sores thing I keep hearing about? And why would I want to develop it? Wouldn't it be better to let it heal?
I'd also like to add that "convergence"-like devices (I hate that usage) seem to be only popular with Slashdot-like geeks. People who would buy the Treo. The same kind of people who would prefer an iRiver with more features to the iPod. The same kind of people who care about Ogg. The same kind of people that don't have much effect on the market. Apple has a good idea what the mass market wants. Hardcore tech geeks don't. It sounds like BillG is being a hardcore tech geek.
Good point. People have been predicing "convergence" will be the next thing for a good 5 or 10 years now, but it hasn't happened. Things like the Treo just never took off. I'm intruiged to see BillG pushing it. I think a lot of people prefer the Unix model where each thing does one thing and one thing well. How do you make a phone that plays MP3 with a good interface? And long battery life? And good form factor? The only company I give a chance of designing one properly is Apple, and I still think it would be hard.
I don't think he was slamming Linus. Read this comment for an explanation.
Or if they are SSHing into your server to.... check their email, let's say?
You feel that 1000 times is a lot? Space them out every 3 seconds so the processor load is nice and low and it wouldn't even take an hour.
When asked for 4 words to describe it, she chose "Bloody Hell - it's good!". Not perfect granted, but good. Even Ebert's 3.5/4 star review is 80% criticisms.
Another nice ad hominem. The guy is completing his frickin DPhil at Oxford, and is 23 years old. Maybe he's a bit too busy to get a job right now.
Allegedly the OpenSSL code has over 1000 data-dependent if-statements in the RSA algorithm. That would certainly not be very easy to make data-independent.
That said, at least you actually understand the problem unlike many of the rabid Linus defenders here.
1. Microsoft is just as potentially vulnerable as Linux. Their dirty laundry just doesn't get aired in public.
2. The fix is non-trivial and non-obvious. It's not a simple buffer overflow. Any patches are likely to have serious repercussions on kernel performance. e.g. disabled HT, ensure only two threads of the same U.I.D. are scheduled on the same processor, flush cache at every context switch, etc. It looks that Linus is unwilling to accept them unless this vulnerability moves more from the theoretical to proven.
Alexandra DuPont
It's a girl, not a guy.
Right, because porn stars teleconference ALL THE TIME.
You are calling the icons the graphics? The guy single-handedly wrote the first ever GUI routines for a consumer product from scratch in assembly, and you say Susan Care did the "graphics"? Wow.
Good point, thanks.
Correction: I meant private key.