I dunno, I'm still questioning how much better it truly is. It's like all the audiophiles who need 2000Kbps MP3s to be happy [yes I know it doesn't go that high].
Some people just like the stats more than the product. I'm all for digital mediums [movies, cable, etc]. But we're already seeing corners cut in HD cable arenas. Sure your resolution is higher, but the bandwidth isn't increased accordingly. etc, etc....
Then there is the whole DRM crap. Frankly CSS was already too much for me. HDCP, SDMI, etc... just shows that the studios don't give two shits about the work and only care about the money.
So what? You have to record the film in high quality because you lose data in the editing/post production phase.
Doesn't mean the final production has to be 1900x1600 with 100KHz 96-bit audio... Fundamentally the experience isn't better when you move from 720x480 to HD. It's just more expensive, vendor locked in and incompatible.
I'm still waiting for the day that Hollywood stops selling out and actually collectively gets the sense to not make shit movies to fill the season. Either you have a good movie/film/etc to show or you don't. Don't make a shit film just because you're lacking in creativity.
Um, hello, "Pulse"... that's a really shitty movie and I want my 8 bucks back.
Agreed. However, releasing a new cpu more often also carries costs and risks. Ramping up production of a processor costs a lot of money, if the processor doesn't live long [core 1] or is a flop [IA] then you lose money. Worse, because you're spending less time on testing you have the increased probability of fatal non-workaroundable errata. Not that AMD processors are perfect but I don't really any bugs in AMD processors that can't [or aren't] fixed with appropriate BIOS updates.
Well clearly I can't comment about future plans from AMD. Just that the next revision beyond F will address some of the Core 2 specific optimizations.
But also keep in mind benchmarks can say anything. You think Core 2 is the better processor? Ok, drop 8 of them in a HPC system and run 50 independent tasks on them from researchers all over the world. You think your 4MB cache helps when you have so much pollution? There are applications where K8 is STILL the better choice by far.
Most benchmarks Intel picks are for very local applications on a single processor [die] setup... I wonder why...
Typical. This is because most netops are anti-social losers who can't get it up. Remember, your job is to support the employees. You work FOR them not against them.
Where I work they have a habit of accidentally turning of ethernet ports even though I filled out the damn forms, etc, etc, etc. When I contact them it's all a big song and dance about "no problem here, you're imagining it, go into your control panel and fix it" even though it's a SLES10 box...
Caught off guard? Um, try again. Just because a company doesn't releae a new core every 8 days doesn't mean they're not cooking something up. A big part of my job at AMD is performance analysis. We do a lot of studies which get fed through various channels to better optimize the final design.
I've known about Core 2 for at least the better part of this year now. I suspect [*] that AMD mgmt knew about it as well, just a hunch.
I have to object to the "no product to counter it". Aside from the shared L2, Core2 actually SHARES a lot of the design with the K7/K8 design. The shared L2 is a mixed blessing. There are applications for which it is good and applications for which it is totally bad. That and when you [ever] get into multi-processing HT is the way to go. Front side busses are a bit lame nowadays.
Your right to privacy stops at the their colocated servers. Get with reality man.
To expect privacy you have to be conducting yourself in a PRIVATE manner. Broadcasting your search requests in PLAINTEXT over the internet to a server which does not [nor must] guarantee privacy is not conducting yourself in a private manner.
As someone who paid a $50 tech fee and STILL HAD TO bring my OWN laptop cuz the labs were full of non comp.sci students.... bite me.
That and when you're in a group of 8 working in the cafe (again, only open tables) and there is no wifi and only 3 usable ethernet ports... again "bite me".
Many schools treat their students poorly because the halfwit techs they hire couldn't manage a lab let alone a standalone 8 port switch...
Don't forget the gold plated wall jacks [only 25% of which are allowed to work].
I was the type of person to bring an AP with me. Screw the lousy 802.11b. So I brought a bg with me. Plugged it into the ethernet jacks and gave all the people around me decent network access. Loads of fun.
Well you wouldn't want 8 cores in a workstation until the power issues are resolved. I hope you don't think Intel is magically making their 4 core processors run under 90W. You're cute:-) Just plug in your 800W power supply and put the ear plugs in!
Hint: Their dual cores aren't under 90W and there 4 core is just two dual cores strapped together. Do the math.
WTF is that? I was at Linux World. It was nothing more than a bunch of suits running Windows for the presentations talking about the great virtues of Linux. Fuck, they even had an Apple booth at LW!!!
If Apple was so concerned over leaks they would tighten the reigns. The fact that pretty much all of their OSes get leaked in advance [same happens to MSFT] speaks volumes of their security or motives [or both].
Yeah I'm sure. It was "leaked." As in, Apple wants more press so they do what they do with every other release and accidentally get a beta out the door.
Can we stop pretending to be gullible and just call it what it is?
That's the dumbest form of retort ever. I do [LibTomCrypt] my bit to contribute to OSS. Just because I haven't ported all of MSFT software to other OSes doesn't mean I can't point out that they're continuing a trend that is harmful to the public.
If your company will publish more than one game, develop a network layer and reuse it. If you're big enough just license one. If you're small base it on more portable primitives.
A more sound solution for smaller companies
1. Write game with a flexible framework [e.g. C++ class driven, can even emulate them in C] 2. Write first game [edition/release] for Win32 3. As you get money from game spend time on writing Linux [e.g. X11/ALSA] plugins
This is nicely "self-pacing" as you don't use time on porting unless you're popular enough to support it anyways. Also as you grow you can later re-use existing code. So your next games come cheaper to code [leaving you more time for content, art and music].
There is a difference when you add something but the core value isn't tied to it. OS X is a decent OS with or without any given random widget. How useful is Windows without DX or Office or Visual Studio? Not very. What this lets Microsoft do is grow their market and lock them into situations they shouldn't be locked into.
Generally, in terms of software, I'd draw the line at software which is platform specific [for no technical reason] which uses proprietary standards [when published ones exist, or more open methodologies are possible] and is provably released to prop up the sales of another, technically unrelated product.
For instance,
Apple doing a custom commercial port of OpenOffice that uses the OpenOffice file formats, not monopoly.
Apple hijacking OO and re-writing the file save/open formats to lock in customers to OS X, monopoly.
Apple building their own custom collection of ISO C and C++ compiler tools, not monopoly.
Apple building their own compiler, supposedly based on C++ with their own non-portable hacks and use it as the platform language of choice, monopoly.
Slashdot Mirror
Well it was more here too [Toronto] but it's a saying right [South Park].
... um 8 euro.
iirc it was like 13$ or something CDN which is about
Vive la France libre!
Tom
The problem with 10MP videos is that it costs storage and bandwidth. So they just cheat you on accuracy. Oh sure, it's 1900x1600 ... at 4Mbps ...
:-)
And frankly at 30fps I don't care about the small details. That's what they invented a zoom for
If you want to view 10MP stills go get yourself a set of projectors, X11 and map that out.
Tom
I dunno, I'm still questioning how much better it truly is. It's like all the audiophiles who need 2000Kbps MP3s to be happy [yes I know it doesn't go that high].
... just shows that the studios don't give two shits about the work and only care about the money.
Some people just like the stats more than the product. I'm all for digital mediums [movies, cable, etc]. But we're already seeing corners cut in HD cable arenas. Sure your resolution is higher, but the bandwidth isn't increased accordingly. etc, etc....
Then there is the whole DRM crap. Frankly CSS was already too much for me. HDCP, SDMI, etc
Tom
So what? You have to record the film in high quality because you lose data in the editing/post production phase.
... Fundamentally the experience isn't better when you move from 720x480 to HD. It's just more expensive, vendor locked in and incompatible.
Doesn't mean the final production has to be 1900x1600 with 100KHz 96-bit audio
I'm still waiting for the day that Hollywood stops selling out and actually collectively gets the sense to not make shit movies to fill the season. Either you have a good movie/film/etc to show or you don't. Don't make a shit film just because you're lacking in creativity.
Um, hello, "Pulse"... that's a really shitty movie and I want my 8 bucks back.
Tom
First off, HD is a scam. It's a waste of time and effort to make people think they're having a better experience.
How about this, make movies with REAL PLOTS AND HUMAN ELEMENTS?
Second, don't use Vista. I don't see them dictating what GNU/Linux can do.
Third, 720x480 is just fine for movies. Grow up.
Tom
Agreed. However, releasing a new cpu more often also carries costs and risks. Ramping up production of a processor costs a lot of money, if the processor doesn't live long [core 1] or is a flop [IA] then you lose money. Worse, because you're spending less time on testing you have the increased probability of fatal non-workaroundable errata. Not that AMD processors are perfect but I don't really any bugs in AMD processors that can't [or aren't] fixed with appropriate BIOS updates.
Tom
Well clearly I can't comment about future plans from AMD. Just that the next revision beyond F will address some of the Core 2 specific optimizations.
But also keep in mind benchmarks can say anything. You think Core 2 is the better processor? Ok, drop 8 of them in a HPC system and run 50 independent tasks on them from researchers all over the world. You think your 4MB cache helps when you have so much pollution? There are applications where K8 is STILL the better choice by far.
Most benchmarks Intel picks are for very local applications on a single processor [die] setup... I wonder why...
Tom
Typical. This is because most netops are anti-social losers who can't get it up. Remember, your job is to support the employees. You work FOR them not against them.
Where I work they have a habit of accidentally turning of ethernet ports even though I filled out the damn forms, etc, etc, etc. When I contact them it's all a big song and dance about "no problem here, you're imagining it, go into your control panel and fix it" even though it's a SLES10 box...
This is why most netops get zero respect.
Tom
Caught off guard? Um, try again. Just because a company doesn't releae a new core every 8 days doesn't mean they're not cooking something up. A big part of my job at AMD is performance analysis. We do a lot of studies which get fed through various channels to better optimize the final design.
I've known about Core 2 for at least the better part of this year now. I suspect [*] that AMD mgmt knew about it as well, just a hunch.
I have to object to the "no product to counter it". Aside from the shared L2, Core2 actually SHARES a lot of the design with the K7/K8 design. The shared L2 is a mixed blessing. There are applications for which it is good and applications for which it is totally bad. That and when you [ever] get into multi-processing HT is the way to go. Front side busses are a bit lame nowadays.
Tom
[*] My opinions are my own, yada yada.
Your right to privacy stops at the their colocated servers. Get with reality man.
To expect privacy you have to be conducting yourself in a PRIVATE manner. Broadcasting your search requests in PLAINTEXT over the internet to a server which does not [nor must] guarantee privacy is not conducting yourself in a private manner.
Tom
As someone who paid a $50 tech fee and STILL HAD TO bring my OWN laptop cuz the labs were full of non comp.sci students.... bite me.
That and when you're in a group of 8 working in the cafe (again, only open tables) and there is no wifi and only 3 usable ethernet ports... again "bite me".
Many schools treat their students poorly because the halfwit techs they hire couldn't manage a lab let alone a standalone 8 port switch...
Tom
My god, you went to my school!!!
Don't forget the gold plated wall jacks [only 25% of which are allowed to work].
I was the type of person to bring an AP with me. Screw the lousy 802.11b. So I brought a bg with me. Plugged it into the ethernet jacks and gave all the people around me decent network access. Loads of fun.
Tom
Not if you have more than one user. Hint: think about wifi deployed at a school or airport...
... wait for n.
As for the general question, the answer is: Upgrade if you have to. If your users are bitching that the net is too slow, upgrade.
If you just want to be hip and spout the latest and greatest
Tom
Well you wouldn't want 8 cores in a workstation until the power issues are resolved. I hope you don't think Intel is magically making their 4 core processors run under 90W. You're cute :-) Just plug in your 800W power supply and put the ear plugs in!
Hint: Their dual cores aren't under 90W and there 4 core is just two dual cores strapped together. Do the math.
Tom
Yeah.... never would see that before... 4 Opteron Dual Cores ... 8 cores, 4 memory busses. Welcome to... um ... NOW!
Until Intel drops the FSB you can't take their MP setups seriously compared to AMD.
Tom
WTF is that? I was at Linux World. It was nothing more than a bunch of suits running Windows for the presentations talking about the great virtues of Linux. Fuck, they even had an Apple booth at LW!!!
:-(
me
Tom
Phones don't do all of the codec in software. Imagine a hardware assited CELP codec :-)
Tom
If Apple was so concerned over leaks they would tighten the reigns. The fact that pretty much all of their OSes get leaked in advance [same happens to MSFT] speaks volumes of their security or motives [or both].
Tom
Yeah I'm sure. It was "leaked." As in, Apple wants more press so they do what they do with every other release and accidentally get a beta out the door.
Can we stop pretending to be gullible and just call it what it is?
Tom
That's the dumbest form of retort ever. I do [LibTomCrypt] my bit to contribute to OSS. Just because I haven't ported all of MSFT software to other OSes doesn't mean I can't point out that they're continuing a trend that is harmful to the public.
Tom
If your company will publish more than one game, develop a network layer and reuse it. If you're big enough just license one. If you're small base it on more portable primitives.
A more sound solution for smaller companies
1. Write game with a flexible framework [e.g. C++ class driven, can even emulate them in C]
2. Write first game [edition/release] for Win32
3. As you get money from game spend time on writing Linux [e.g. X11/ALSA] plugins
This is nicely "self-pacing" as you don't use time on porting unless you're popular enough to support it anyways. Also as you grow you can later re-use existing code. So your next games come cheaper to code [leaving you more time for content, art and music].
Tom
There is a difference when you add something but the core value isn't tied to it. OS X is a decent OS with or without any given random widget. How useful is Windows without DX or Office or Visual Studio? Not very. What this lets Microsoft do is grow their market and lock them into situations they shouldn't be locked into.
Generally, in terms of software, I'd draw the line at software which is platform specific [for no technical reason] which uses proprietary standards [when published ones exist, or more open methodologies are possible] and is provably released to prop up the sales of another, technically unrelated product.
For instance,
Apple doing a custom commercial port of OpenOffice that uses the OpenOffice file formats, not monopoly.
Apple hijacking OO and re-writing the file save/open formats to lock in customers to OS X, monopoly.
Apple building their own custom collection of ISO C and C++ compiler tools, not monopoly.
Apple building their own compiler, supposedly based on C++ with their own non-portable hacks and use it as the platform language of choice, monopoly.
Tom
My parents use this defense often and it's really annoying. I don't think Apple is of any higher moral standing than Microsoft or Intel [etc].
Microsoft is just the punching bag here since this is a thread about DX.
Tom
My college taught "numerical analysis" in the software comp.eng side. You learn the format of IEEE types, the range, accuracy, precision issues, etc.
We had assignments to not only perform matrix ops but also give the expected error, etc.
Maybe the author of the article should either go to a better school or pay more attention to the classes.
Tom
I don't get your reply. If I reverse engineered it, it would be legal [may violate patents but that's a civil matter not criminal].
My point is people should shun not celebrate DX.
Tom