Yeah, I understand that, but my point was that it is too early to blame Windows. For one thing most OSes provide the reasonable number of required services like failover, decent CPU scheduler and the other most used features. For second, the previous poster and yourself both assume that the trading application suffered from the OS not providing some feature. That is not proved yet. There is no reasonable doubt to assume that would be the case. You are talking about special purpose applications (real time for e.g.) that require special OS features that are not present in commodity OSes and I don't know that the trading application in such and application. If it was then whoever chose general purpose OS like Windows was not thinking and the fact that it ran so long is an indicator that it wasn't the case. So without having any knowledge that the trading application lacked the OS support for some special feature it required from Windows - it is too early and unreasonable to blame Windows.

What? Are you serious? You actually made the same point I did - except that you made it by _NOT_ mentioning the OS! The rest of what you wrote is complete BS - sorry. BS not on its own but BS to the current context. You don't want unhappy traders - sure, but how the fuck is the OS directly responsible to unhappy traders? At least make some points to clarify may be? Or would that be too much to ask for?

Mod parent up. There is a refreshing amount of reality in the parent's post.

The OS is irrelevant - every modern server OS performs well enough to support sanely written software and sanely designed infrastructure. Only the people living in the past and the ones having no clue will argue otherwise.

That is not to say any OS is without its quirks and differences. Part of writing software for a platform is also to understand and work around the platform's limitations, quirks or "ways" - and this obviously applies to every platform. (You don't go over committing memory on Linux and expect it will work as long as malloc() doesn't return NULL - you can do that on Solaris. Similarly I am sure there are things you can do on Linux but not on Solaris etc.) Unfortunately many programmers only know the programming language and its libraries - not the platform or even general OS concepts or scalability for that matter.

The suits are trigger happy - if some thing doesn't work for a time, they will just ask to get rid of it and use other product, redesign the whole thing or do something equally idiotic. I am sure the TradElect system can be fixed to run on Windows 2003 well enough - but the people who make decisions will not make an attempt to locate competent Architects and Programmers that can actually fix it.

[ This reminds me of a situation where we were asked to throw away a complete system because it wasn't able to handle high volumes and caused downtime - as it turned out, adding network timeouts and retries to the right places along with horizontal scaling resolved the issue completely satisfactorily and we are still running the same system 3 years later]

Did not make a noticeable difference on a non-SSD 15" MacBook Pro in my very unscientific but very typical usage test - I guess the hardware designers were sane - if you are not using the speed you don't pay for it. I have seen Intel controller documents that sort of state that 0.4w more power consumed by 3Gbps capable controllers (compared to 1.5Gbps ones) but that is Intel controller specific and the controller in the new MBP is nVidia's.
I am seeing some increased warmth near the speakers after the update under regular use but I classify that under "General Post Update Weirdness" not caused by anything specific :)

I always thought the US of A was the most rebellion if not a 100% accepting of all when it comes to choosing what one wants to do their hands. I still think it is easier to live a happy life in the US opening a bike repair shop than it is back in my home country.
In fact where I come from I always hated parents dictating children what to do and society belittling young people who chose to do something different. Heck even today most parents won't marry their daughter to a person who has got a bike repair shop - they would much rather marry her to a dim with with an IT job ;)
I guess the problem is with pandemic illogical comparison - that guy is in a call center job and earns lots of money just sitting and you dirty your hands every day and struggle to make a living. The flaw of course being that it totally ignores the possibility that I can be happier, healthier and more productive doing bike repair and be able to sustain a good income as opposed to the possibility that the call center worker might have fissures sitting on desk all day and may never "achieve" anything in life - notice I said these are possibilities.
But I guess the so called "society" is same every where you go - children need to be raised to not only do what makes them happy and the most productive but also develop the right skills and mindset to do it.

Intel managed to sell me the "Extreme" Graphics i740 back then and when I look back it just me sad again.

Of course today I would not buy an Intel graphics card even if they named it Larrabee - oh wait they already did?

I discovered DOSBox only a year of so ago when my brother brought back the memories of one of the first DOS games we used to play on our first computer - Zed for DOS by Bitmap Brothers. We had a lot of fun playing that game back then. ( For me it simply is the best game ever made.)

I thought it would be revive the good old times and see if the kids enjoy the game as much as we did - but a) I did not had the game installer any more and b) I had read on forums that the game won't run on modern hardware/os without going thru a lot of hoops.

I solved the first problem by going to ebay and buying a Zed CDROM for $1.99 + SH. Second problem was solved by DOSBOX.

I was expecting to do a lot of stuff (like setting up sound card, giving up on multiplayer, living with instability etc.) but amazingly enough - it was a breeze to get the game working in its full glory - including multiplayer. And it worked on everything - on XP initially and then on Linux on my kids HP Mini and recently on Windows 7 when I upgraded! And the best part is it is so reliable.

Kudos to the guys working on this.

(Heads to dosbox.com to donate)

Pay attention - If you really do end up putting 3G chips in the Mac, please let people choose which cell phone company they want to sell their soul to. I have a nagging feeling Apple will again do the tie-up thing and force people to use the carrier they can make money from. And god forbid if that is AT&T the whole thing is just useless for many people like me who when on ATT 3G get speeds equal to T-Mo Edge and then a move a few miles out of town means not even getting a reliable Edge connection on ATT network. It would be most useful if they went with something like the Gobi.

Just finished quick testing ODF Text conversion of some Architecture/Deign documents of medium to high complexity (Word 2003) and I have to say the results are good. All of the documents converted to ODF in no time, Word 07 SP2 opened the converted documents quickly and to my shock the ODF document looked exactly like the original (that is no immediately visible glitches - haven't spend reviewing line by line as these are huge documents). More joy came when OO3.0.1 opened the converted documents just fine - no compatibility issues there, the documents even looked great in OO. So seems to me like Microsoft has done a great job with ODF conversion - this might take us one step closer to having a truly interoperable document format (alas we still need Word to author complex documents in the enterprise for its collaboration features and other integration stuff).

I find it odd that no one asked any questions about the future of Solaris - although there was a round-about question on x86 which resulted in an somewhat positive answer for SPARC. Oracle seems to be keeping SPARC and thus Solaris alive. (There isn't another OS running SPARC that is in widespread use after all.) This also makes me wonder if Oracle product support for Solaris x86 is going to improve now. This also seems to suggest that Oracle may not be selling Sun's hardware business to HP per the original plan. The idea sounded very interesting - HP would then become the most diversified hardware company selling x86, Itanium and SPARC hardware.

For less than the price of your self built machine, I bought a HP xw6600 workstation with Xeon QC CPU and 10Gb ECC RAM - refurbished, but you can't tell the difference and I've had no problems for 6 months. And this is a very nicely designed workstation with fair amount of upgrade potential.

You can't do that sort of thing with the Apples. Need to do SLI on two PC compatible nVidia cards - you are out of luck with the Mac Pro - you can only buy compatible cards Apple sells for a premium. Want to use particular hardware and OSX offers no drivers for it? Out of luck. (No - third party drivers are either unreliable or simply non-existent for OSX in most cases. And then Apple keeps breaking things. Say what you will about Microsoft but they have improved the driver situation - even for Vista x64 - quite a lot - plug in any damn thing and it will get the right driver automagically.)

On the other hand, I like Mac Laptops. I have 2 MacBook Pros but both of them needed $349 Apple Care - they had to replace the logic board twice and top case once. But I stick with them as a matter of liking.

So yes, I think Apple is overpriced, offers the least options/compatibility and the quality/reliability isn't drastically better than the PC counterparts. Thus it just becomes a matter of liking and ability to spend the extra bucks which determines if a person is going with a Mac or PC.

Of course nothing is 100% secure - it's all relative - how much hard you make it, how big of a deterrent you put up. You are missing the difference that Application bugs are very common compared to JVM bugs. Tell me how many browser bugs have been there and how many JVM bugs that allow arbitrary code execution. The JVM infrastructure provides the mechanism to prevent arbitrary code execution and does it right most of the times - Native code is way too unsafe for Application development. Another Safari bug and it will be exploited in most dangerous ways - another Java application bug and the JVM will restrict the consequences to at the most a DoS. You are continually denying that a VM based sandbox provides much stricter mechanism and much better detterrent against security flaws - that is commonly proven knowledge. You are arguing that JVM has bugs, native code has bugs so it's no different - which is naive. You do not seem to understand the difference between native code executing on CPU and bytecode executing inside a VM. You also seem to completely ignore the security mechanism Java provides - All Java code is subjected to a bytecode verification process and conformance to security model is enforced. Native machine code is unverifiable the application will happily try to execute garbage and crash. Through Java's SecurityManager interface, user agents can selectively impose fine-grained controls on downloaded code, specifying which files it can read or write, which hosts it can connect to, and so on. Such control is impossible when execuing native code. You can read and check the Java VM specification, the source of the SecurityManager, and so on, and satisfy yourself that they don't leave any holes. In short a quality JVM implementation will provide a much better resistance against security problems than standard UNIX or Windows security can ever do. There can hardly be any debate about it if you make an attempt to understand.