So, if there's 3,000,000 people with an operating system, but our members have only sold 2,000,000, that's 1,000,000 pirated copies of our member's operating systems! [... ] Surely that can't be how they work it out. [... ]
Nope, that's exactly how they work it out. Download and read their "study" yourself (the methodology section is toward the back). Their "piracy" estimates are based on nothing more than wild guesses as to how many copies "should" have sold, given the number of computers out there.
Then, just for fun, they turn around and claim to their shareholders that sales exceeded expectations. Well, which is it, Chucko? Either you sold less than you anticipated (inflating the "piracy" figures), or you sold more (inflating your stock price). Either way, your market projections are way off.
Oh please. You're talking about an engineer here, not a soldier. An engineer has no duty. They have a job. [... ]
No duty? Go read this story (all the way through) and then try saying that again with a straight face. (If you can, the true nature of your moral character will be revealed.)
There is no "duty" to go back later and fix those bugs. [... ] The only "duty" anybody has is to do what is in their best economic interest. [... ]
There is always a cost! That's not a "business/accounting/shareholder standpoint", that's economic fact.
So, are you suggesting that engineers everywhere should subordinate their sense of moral, ethical, and social responsibility to the "facts" of economics? That a faulty bridge should not be repaired because of the "opportunity cost" to the engineering firm? That an easily-punctured gas tank shouldn't be relocated because the re-engineering costs exceed the anticipated court-imposed liability payouts? That a medical device that infects or disfigures its users shouldn't be fixed or recalled because those dollars can be, "better leveraged," elsewhere?
The moment you start to affect other people's lives -- and make them pay for the privilege -- you are expected to accept responsibility for the effect you have. The precise degree of responsibility is under constant debate, and highly situation-dependent. But the responsibility exists, and it is subordinate to your economic models. Caveat Emptor is not, nor should it be, the whole of the law.
Dollars are ephemeral; but reputations, good and bad, can last for millennia.
I suggest maybe you spend some time in the trenches of the software industry. You statement is laughable on it's face.
I've been in the trenches of software development for 25 years, and it is you who I'm laughing at. I can't begin to describe how appalling I found your post. But I'll refrain from further pithy insults and get straight to the invalid concepts.
But still, it cost you nothing other than a few bucks in bandwidth and maybe a little goodwill.
Wrong in both cases. In both cases the person doing the fixes lost the opportunity to do other work. [... ]
If you have a known bug in your program that is causing people problems, it is a software engineer's first duty to fix that bug. Working on a bug doesn't mean you've, "lost the opportunity to do other work." The opportunity does not exist because you have no other work until the bug is fixed. Adding new features and animated buttons and gee-gaws always -- ALWAYS -- takes a back seat to creating a reliable, robust product. To do otherwise would result in an unmitigated disaster like -- just to pick a completely random example out of thin air -- Windows.
The time spent on the fixes is lost forever to the engineers.
Invalid concept; presupposes false economy. The time spent on the fixes is time that would have been spent in the first place by the engineers had they done a proper job. Perhaps they were incompetent boobs. Perhaps they were top-flight engineers who, like all humans, made an honest mistake. Perhaps they were top-flight engineers who were being pressured by senior management to, "get something out the door making money now and worry about the details later." Whatever the reason for the bug, it's engineering time that is inextricably part of the product, whether you choose to "spend" it or not.
If it is a really significant bug it could take dozens or a hundred people to prepare the fix - from programmers to testers to QA to legal to webmasters to documentation experts to channel partners to vendors to hardware suppliers to PR. All of which has a significant and non-trivial cost.
Yup; no argument here. That's why you hire top-flight engineers and give them the time and resources they need to do the best possible job the first time. That way you avoid the (much larger) costs of bug fixes later on. What you don't do is use those costs to justify not developing and shipping a fix. That simply shifts those costs to your customers.
I'm sure your point of view makes perfect sense from a business/accounting/shareholder standpoint. But from the perspective of an engineer, or even simple human decency, it stinks.
Schwab
Re:The Real Reason is that Geeks Are Submissive
on
Nerds Make Better Lovers
·
· Score: 2, Insightful
As women of the world quickly progress towards femdom (yes the practice is growing as more of us realize that we can own the world) the geek will become the more desirable mate as they like dicipline and submission. Just remember to dominate your sub and make him well aware of who the boss is.
I'm sorry, but this paragraph just drips with disrespect, and I find I can't let it pass unchallenged.
You've also, I believe, got it wrong: Geeks don't crave discipline and submission (honestly, when a world-class geek faces down a computer, which do you think does the submitting?). What we do like is structure and determinism. We would like the wetware world to be just as orderly and predictable as the software and hardware worlds. So if I, for example, performed activity X, and consistently received from you response Y, I would therefore presume that Y has something to do with X, and will build my mental models accordingly. OTOH, if I perform X, and I receive response mrand48(), then I would presume you're insane, and would be incapable of forming any kind of mental structure.
Geeks are uncomfortable when they don't know what to do because they don't know what the rules are -- because they don't know the structure. In this respect, dear lady, you hold a slight advantage in that such structure is often formalized in D/S (and no, I'm not going to lowercase the 'S' because I'm not that pretentious). But I think you're failing to note an inconsistency in your position. To wit:
I did date a few more "normal" guys, but they were either boring, stupid or both. They were obsessed with themselves instead of me [... ]
When reading your post, one starts to form an opinion of where your obsession lies. You decry people obsessed with themselves, yet seem to grant yourself an exception. This incongruity, not to mention apparent disrespect for your partners (by referring to them as "subs" rather than more human terms), may well come back to bite you unpleasantly on the bottom if you're not cautious.
The only reason I bother to reply at all is because D/S can be a tremendously fun space to play in, but it's enormously frustrating to see D/S abused by people trying to, "prove something," or even to reflect the abuses they themselves endured. I am not accusing you of abuse. I've never met you, and therefore grant you the benefit of the doubt. Indeed, you are probably rather charming company. But working entirely from your post paints an unflattering picture. If you don't wish to be seen that way, you should re-examine how you articulate your position.
Funny that this story should pop up about a calculator that will automatically convert a decimal representation into a fraction. I was trying to address this very issue yesterday, only in slightly different form.
Basically, given a fractional value between 0 and 1, find two integers whose ratio most closely approximates the fractional value, and which will fit in a given bit width. This sort of thing is useful when trying to compute the integer coefficients to stuff into the registers of a PLL clock generator.
For a 10-bit range, you can just do a brute-force search (which is what I did), but for anything wider than 20 bits or so, it'd be nice to have an algorithm that converges on a solution quickly. If ever I was taught this in math class, it fell out of my brain long ago. Now I find out a TI calculator will apparently just do it.
System performance will be below par compared to other Intel-based offerings,
Why would you think that? OS X will blow Windows out of the water.
You'll be able to compare apples to, well, Apples this time. Instead of vague handwaving about MacOS-X/PPC performance versus Windows/Intel, you'll be able to compare directly: Windows/Intel(clone) versus Windows/Intel(Apple) (and, conversely, MacOS/Intel(clone) versus MacOS/Intel(Apple)). You'll be able to eliminate the variables that have confounded meaningful performance comparisons. I contend that Apple's HW offerings won't be at all remarkable, and will place it around the middle of the pack, performance-wise.
It is possible that, once this price/performance gap gets wide airplay, Apple may go back to the drawing board and get themselves into the top 5% or so. But that's not something they've had to do in the past -- their machines have always been comparatively unique, with little else in the market to compare to -- so I don't expect to see truly stellar Intel-based machines from Apple for at least four years.
Microsoft won't let them.
You're living in the past.
I don't think so. Microsoft have been absolutely unrepentant for their misdeeds. They have not altered their behavior in any meaningful way, and they continue to demonstrate that product quality is their first priority only when they've exhausted all other options. Those anti-competitive contracts are still in place and, with the threat of Linux and now MacOS-X looming larger, it's no great leap to assume Microsoft will work to maintain those exclusionary deals however it can.
I've long said, Apple needs to decide what they are: a hardware company or a software company. If they're a hardware company, release the machine specs to Linux. If they're a software company, port to Intel.
It seems Apple has decided they're a software company.
Predictions for Apple's Future under Intel
Apple currently makes most of its money on hardware margins. Therefore, there will be a brief flirtation with binding MacOS-X-Intel to Apple-branded Intel-based systems. Despite fruitless (no pun intended) lawsuits to combat the practice, installer patches will rapidly be developed and widely spread to allow MacOS-X-Intel to be installed on any x86-based system (thereby increasing the popularity and spread of MacOS-X, but Jobs will almost certainly be incapable of seeing that, at least initially). Apart from their industrial design, which is absolutely first rate, there will be nothing to recommend Apple-designed boxes over dirt-cheap clones. System performance will be below par compared to other Intel-based offerings, and not enough people will be interested in paying the $500 premium just to get a pretty box.
The reverse transition will also be true: People will try installing Windows on an Intel-based Apple box. Apple will try (and fail) to prevent this, too.
Moving to the Intel-based platforms places Apple in direct competition with Microsoft. The relationship between Apple and Microsoft has long been one of, shall we say, détente. This state has survived because neither has directly tried to enter the other's playground (there is no version of Windows for PowerPC). There may even be secret agreements between the two companies to maintain this state -- indeed, such agreements may be the driving force behind Apple's initial attempts to keep Windows off Apple-branded Intel machines, and MacOS-X-Intel off clone machines. By supporting the Intel platform, Microsoft may feel itself no longer bound by such "gentlemen's agreements," and start pulling overtly dirty tricks to undermine MacOS-X. Expect to see threats of Microsoft ending support for MacOS-X versions of Office. Expect also to see Microsoft even more shamelessly mimic the MacOS look-and-feel in upcoming Windows releases (Shorthair^WLonghorn is still far enough off that it could be completely re-specced).
Once Apple realizes that it can't bind MacOS-X only to its own machines, they will attempt to form OEM relationships with the major PC manufacturers (Dell, HP, IBM, etc.). They will then run smack-bang into the same wall Jean-Louis Gassée did when he tried to get BeOS bundled with PCs -- Microsoft won't let them. The anti-trust accusations will heat up again, this time with Apple behind it. By the time it reaches the courts, George W. Puppet and his operators will no longer be in office, so it's impossible to predict what the political pressures might be. Microsoft will start at a disadvantage, since it already has a criminal judgement against it, but a lot will depend on political orientation of the Justice Department and the Federal Trade Commission.
In short, Apple has stepped into a very different and very messy business landscape. I sincerely hope they're ready for it; I'd hate to see them go away.
You bought a 22" LaCie CRT? Damn, dude, those things are expensive. I've been thinking about replacing my 19" Hitachi CM751, which is starting to show its age, and was always kinda fuzzy at 1280*1024 and above. LaCie looks like a worthy replacement but, because they're professional color-balanced monitors, they command a premium over ordinary CRTs. I don't need precise gamma or color balancing; all I really want is 1600*1200@75Hz with no ghosting or fuzz, and without the smear of LCDs.
Easy. The driver gets a 'sync' command from the OS. However, the driver writer believes that most other programmers call fsync() when they don't really need to, and decides to "optimize" this case. So he passes the command on to the drive, but returns immediately (allowing the drive command to complete asynchronously). This makes his driver appear faster.
Fortunately, most driver writers have their priorities straight about data integrity, so this kind of thinking isn't very common.
Yes, except there is a 'sync' command packet that is supposed to make the drive commit outstanding buffers to the platters, and not signal completion until those writes are done. It would appear, at first blush, that the drives are mis-handling this command when write-caching is enabled.
There is historical precedent for this. There were recorded incidents of drives corrupting themselves when the OS, during shutdown, tried to flush buffers to the disk just before killing power. The drive said, "I'm done," when it really wasn't, and the OS said Okay, and killed power. This was relatively common on systems with older, slower disks that had been retrofitted with faster CPUs.
However, once these incidents started ocurring, the issue was supposed to have been fixed. Clearly, closer study is needed here to discover what's really going on.
Most of these peripherals are integrated into the North- and/or South-Bridge chips these days. The additional cost of these on-chip features has already been amortized out, so the cost savings of not having them integrated into the chips is effectively zero.
If you are a 31337 g4m3r, integrated graphics is indeed a joke. However, it's good enough for 85% of the users out there, who will almost never run anything more intensive than Word, IE, and the occasional Flash-based game. Same deal with integrated sound -- for Windows event beeps and boops, it's more than plenty.
10/100 Mb/s speeds are now common on integrated Ethernet controllers, and most of them have very little braindamage these days. 1Gb/s on-chip controllers are also already starting to appear.
To put it another way: Parallel, RS-232 serial, and PS/2 mouse/keyboard ports used to require separate expansion cards. Today, they are integrated into the motherboard chipset, and no one thinks the worse for this. For those who need extra ports or special high-performance ports, third-party PCI expansion cards are still available.
So, in short, the way systems are being put together these days, there's no cost savings to be had by breaking out the peripherals you don't need. If you feel a need to put the old parts to good use, donate them to a school, or use them to build a Frankenbox on which to do kernel or driver development:-).
Are these people completely stupid? What on earth possessed them to publish the fscking ISOs? If there's a clearer way of establishing a case of copyright infringement, I can't think of it.
Let me be clear: I am no fan of the current copyright regime, since it's at loggerheads with the fundamental nature of computers. The DMCA is an especially execrable piece of... legislation. But by publishing the ISOs to games, the people behind this reverse-engineering effort are almost immediately discrediting the value of their good work. Sony will calmly stand before a judge and say, "See? There's no academic or social or Fair Use argument here. Upon successfully bypassing our security regime, the very first thing these criminals did was copy and publish the games. We therefore ask for a permanent injunction."
Posting the ISOs was completely unnecessary. Now that the discs are (apparently) readable, anybody interested in reverse-engineering the games themselves could just as easily bought their own copy and worked from that.
Some time back, I was watching a discussion on a mailing list about the total worldwide power consumption of computers. As I recall, there are too many variables to make an accurate assessment of "true" power consumption. The consensus eventually settled out to, yes, CPU power consumption is rising, but is still dwarfed by the power needs of heavy industry. And if you measure energy consumption by amount of oil burned, then computers trail far, far behind passenger cars.
Back in 2000, duing the California power "crisis," Amory Lovins of the Rocky Mountain Institute was asked what things citizens could do to conserve power. His response: "Conserve water. The lagest consumer of power in California is electric water pumps. So if you save water, you'll save power."
Still, every little bit helps. By residents switching over from incandescents to screw-in fluorescents duing the power "crisis," California reversed approximately 8-10 years of power consumption increase (according to some estimates).
Yeah, [patent law] works well if your a member of the Billionare club, Incoporated.
Actually, the statement is mostly correct; the law itself is mostly just fine. What's horribly broken is the patent examination and granting process. The examiners have done a shameful job in maintaining the integrity of patents by allowing patents on trivial "inventions" or re-purposing of existing inventions; by allowing patents that do not fully describe how to re-implement the claimed invention(s); and by allowing patents that are nearly unreadable with legalese and deliberately vague language.
Fix the examination and approval process, and the patent system will almost certainly sort itself out again without any legislative changes.
Hmm. Sounds like Microsoft is offering to get into "partnership" with OSS, thereby leveraging their respective core competencies to achieve market synergy and better computing products and services for everyone.
Yeah, right.
Just ask any company that has "partnered" with Microsoft how well it's worked out for them. The list of mangled corpses is long and grotesque indeed.
Interoperating with OSS is no secret. Read the RFCs, download the sources, and knock yourself out. An ersatz "partnership" is not required.
Um yeah and you've levied a broad stroke yourself. What about niche industry applications? [... ]
Just so; I stand corrected. There are a large variety of narrow-market niche applications that are extremely mature with an equally mature user base, and which are only available for Windows. (OrthoTrac is one example that comes to mind.) F/OSS duplicates of these products would be extremely non-trivial.
If you want an inexpensive machine for Computer Science studies or to learn UNIX networking or even as a SOHO server for the advanced user, Linux is your game. Similarly, for gaming, business applications, enterprise servers or streaming media from your computer to your TV you won't go wrong with Windows.
Ah, the joys of the broad-stroked brush. Let's take this apart, shall we?
Gaming
Which aspect of gaming: the server or the client? If you're talking about the client then, sadly, yes, Windows holds the edge here, since the graphics and sound drivers are more mature and better supported. However, for a server -- especially a public one -- you'd be a complete fool to put anything less secure than a Linux box on the net. Even better to put up one of the BSD variants.
Business Applications and Enterprise Servers
The strokes don't get broader than this. But basically, all Windows is good for here is running Office (Word, Excel, Powerpoint). Everything else is better off running on Linux: Intranet Web servers, email servers, file servers, backup servers, Oracle, and corporate firewall.
Streaming Media
There is nothing magic about streaming media, especially when it's done from a bunch of audio files stored on disk. Windows' proprietary DirectX brings nothing to the table here. So that means selecting for a robust, secure server platform, which makes the choice fairly obvious...
Microsoft End-User Unilateral Non-Negotiable License "Agreement:"
[... ]
92. Definition of Terms
[... ]
xxi. "Working" shall be defined as any software product or service that can reliably and repeatably produce Guaranteed Functionality.
[... ]
clxiv: "Guaranteed Functionality" shall be limited to the Blue Screen of Death or any other diagnostic output or display. All other software functionality shall be deemed Extensions or Enhancements and may or may not operate with any degree of reliability.
Adobe likes to think it has software patents on certain user interface elements of Photoshop. Macromedia was sued not long ago for emulating some of these elements in their Flash authoring tool (no idea how that suit resolved). Adobe are also litigious so-and-sos who are all too willing to harass people in defense of their intellectual "property".
So if Adobe feels even the slightest bit threatened by it, expect the project to receive a nastygram.
...the SANS Internet Storm Center has raised their alert level to Yellow following a rash of active DNS poisonings.
ATTENTION: ALERT LEVEL UPDATE. The authorities at SANS (Sebben-Affilliated Network Security) have issued this network alert update:
The DNS cache poisoning alert has been upgraded from "Yellow" to "Blackwatch Plaid." Repeat: DNS cache poisoning alert level is now at Blackwatch Plaid.
Available information does not yet justify a further upgrade to alert level "Moving Pictures."
And for everyone's safety and security, and to preserve our way of life, SANS is taking a drastic step and installing a network monitor. Just one. For safety, security, and omniscient, unblinking information gathering of everyone's activities.
:-),
Schwab
I Got Yer "Authorization" Right Here...
on
Inside the PSP
·
· Score: 5, Insightful
"While taking apart the device is not technically reverse-engineering, Sony explicitly outlaws any modifications to their PSP. If you don't believe me, check out page 15 of the manual where they state, "No authorization for the analysis or modification of the system, or the analysis and use of circuit configurations, is provided."
Very funny. Fortunately, as we all know, Sony does not have the right to make, much less enforce, this request. "Authorization" for opening and/or modifying the machine is embodied in the receipt from the retailer. Sony can void your warranty, of course, but you knew that when you started unscrewing the thing.
Let us not take such histrionics seriously; it gives people the wrong idea.
Are they going to be x86-compatible? MIPS compatible? What?
Since it's Freescale (née Motorola) that's mentioned in the article, any general-purpose CPU appearing from this effort will probably be ARM-based. However, the most likely application will be specialized processors for multi-gigabit network routers.
[Disclaimer: I work for Tapwave, Inc., makers of the Zodiac.]
So, let's see here. For $250, you get a cool-looking but big and chunky thing that's a closed platform with proprietary media formats and only a handful of games.
On the other hand, for just $20 more, you can get a Tapwave Zodiac with:
32M RAM,
two SD (not MemoryStick) slots,
higher resolution touchscreen display,
built-in MP3 player,
bundled video player from Kinoma,
full PalmOS v5 compatibility,
wider selection of games,
wireless connectivity and multiplayer gaming via Bluetooth,
no copy protection/DRM nonsense imposed on your media files,
metal casework, and,
small enough to comfortably fit in your pocket.
It also supports Web browsing and POP/IMAP email via Bluetooth networking or an 802.11b SDIO card (sold separately).
All new platforms have teething pains, and Sony will undoubtedly be addressing the PSP's shortcomings over the following months (including its rather high price). But if you don't want to wait, or you'd rather have something you can actually comfortably carry around with you, you might want to amble on over to CompUSA or Fry's and take a look at the Zodiac.
And I'm not just saying that because I work for them:-).
Slashdot Mirror
Nope, that's exactly how they work it out. Download and read their "study" yourself (the methodology section is toward the back). Their "piracy" estimates are based on nothing more than wild guesses as to how many copies "should" have sold, given the number of computers out there.
Then, just for fun, they turn around and claim to their shareholders that sales exceeded expectations. Well, which is it, Chucko? Either you sold less than you anticipated (inflating the "piracy" figures), or you sold more (inflating your stock price). Either way, your market projections are way off.
Schwab
No duty? Go read this story (all the way through) and then try saying that again with a straight face. (If you can, the true nature of your moral character will be revealed.)
So, are you suggesting that engineers everywhere should subordinate their sense of moral, ethical, and social responsibility to the "facts" of economics? That a faulty bridge should not be repaired because of the "opportunity cost" to the engineering firm? That an easily-punctured gas tank shouldn't be relocated because the re-engineering costs exceed the anticipated court-imposed liability payouts? That a medical device that infects or disfigures its users shouldn't be fixed or recalled because those dollars can be, "better leveraged," elsewhere?
The moment you start to affect other people's lives -- and make them pay for the privilege -- you are expected to accept responsibility for the effect you have. The precise degree of responsibility is under constant debate, and highly situation-dependent. But the responsibility exists, and it is subordinate to your economic models. Caveat Emptor is not, nor should it be, the whole of the law.
Dollars are ephemeral; but reputations, good and bad, can last for millennia.
Schwab
I've been in the trenches of software development for 25 years, and it is you who I'm laughing at. I can't begin to describe how appalling I found your post. But I'll refrain from further pithy insults and get straight to the invalid concepts.
If you have a known bug in your program that is causing people problems, it is a software engineer's first duty to fix that bug. Working on a bug doesn't mean you've, "lost the opportunity to do other work." The opportunity does not exist because you have no other work until the bug is fixed. Adding new features and animated buttons and gee-gaws always -- ALWAYS -- takes a back seat to creating a reliable, robust product. To do otherwise would result in an unmitigated disaster like -- just to pick a completely random example out of thin air -- Windows.
Invalid concept; presupposes false economy. The time spent on the fixes is time that would have been spent in the first place by the engineers had they done a proper job. Perhaps they were incompetent boobs. Perhaps they were top-flight engineers who, like all humans, made an honest mistake. Perhaps they were top-flight engineers who were being pressured by senior management to, "get something out the door making money now and worry about the details later." Whatever the reason for the bug, it's engineering time that is inextricably part of the product, whether you choose to "spend" it or not.
Yup; no argument here. That's why you hire top-flight engineers and give them the time and resources they need to do the best possible job the first time. That way you avoid the (much larger) costs of bug fixes later on. What you don't do is use those costs to justify not developing and shipping a fix. That simply shifts those costs to your customers.
I'm sure your point of view makes perfect sense from a business/accounting/shareholder standpoint. But from the perspective of an engineer, or even simple human decency, it stinks.
Schwab
I'm sorry, but this paragraph just drips with disrespect, and I find I can't let it pass unchallenged.
You've also, I believe, got it wrong: Geeks don't crave discipline and submission (honestly, when a world-class geek faces down a computer, which do you think does the submitting?). What we do like is structure and determinism. We would like the wetware world to be just as orderly and predictable as the software and hardware worlds. So if I, for example, performed activity X, and consistently received from you response Y, I would therefore presume that Y has something to do with X, and will build my mental models accordingly. OTOH, if I perform X, and I receive response mrand48(), then I would presume you're insane, and would be incapable of forming any kind of mental structure.
Geeks are uncomfortable when they don't know what to do because they don't know what the rules are -- because they don't know the structure. In this respect, dear lady, you hold a slight advantage in that such structure is often formalized in D/S (and no, I'm not going to lowercase the 'S' because I'm not that pretentious). But I think you're failing to note an inconsistency in your position. To wit:
When reading your post, one starts to form an opinion of where your obsession lies. You decry people obsessed with themselves, yet seem to grant yourself an exception. This incongruity, not to mention apparent disrespect for your partners (by referring to them as "subs" rather than more human terms), may well come back to bite you unpleasantly on the bottom if you're not cautious.
The only reason I bother to reply at all is because D/S can be a tremendously fun space to play in, but it's enormously frustrating to see D/S abused by people trying to, "prove something," or even to reflect the abuses they themselves endured. I am not accusing you of abuse. I've never met you, and therefore grant you the benefit of the doubt. Indeed, you are probably rather charming company. But working entirely from your post paints an unflattering picture. If you don't wish to be seen that way, you should re-examine how you articulate your position.
Schwab
Basically, given a fractional value between 0 and 1, find two integers whose ratio most closely approximates the fractional value, and which will fit in a given bit width. This sort of thing is useful when trying to compute the integer coefficients to stuff into the registers of a PLL clock generator.
For a 10-bit range, you can just do a brute-force search (which is what I did), but for anything wider than 20 bits or so, it'd be nice to have an algorithm that converges on a solution quickly. If ever I was taught this in math class, it fell out of my brain long ago. Now I find out a TI calculator will apparently just do it.
So how's it done?
Schwab
You'll be able to compare apples to, well, Apples this time. Instead of vague handwaving about MacOS-X/PPC performance versus Windows/Intel, you'll be able to compare directly: Windows/Intel(clone) versus Windows/Intel(Apple) (and, conversely, MacOS/Intel(clone) versus MacOS/Intel(Apple)). You'll be able to eliminate the variables that have confounded meaningful performance comparisons. I contend that Apple's HW offerings won't be at all remarkable, and will place it around the middle of the pack, performance-wise.
It is possible that, once this price/performance gap gets wide airplay, Apple may go back to the drawing board and get themselves into the top 5% or so. But that's not something they've had to do in the past -- their machines have always been comparatively unique, with little else in the market to compare to -- so I don't expect to see truly stellar Intel-based machines from Apple for at least four years.
I don't think so. Microsoft have been absolutely unrepentant for their misdeeds. They have not altered their behavior in any meaningful way, and they continue to demonstrate that product quality is their first priority only when they've exhausted all other options. Those anti-competitive contracts are still in place and, with the threat of Linux and now MacOS-X looming larger, it's no great leap to assume Microsoft will work to maintain those exclusionary deals however it can.
Schwab
I've long said, Apple needs to decide what they are: a hardware company or a software company. If they're a hardware company, release the machine specs to Linux. If they're a software company, port to Intel.
It seems Apple has decided they're a software company.
Predictions for Apple's Future under Intel
Apple currently makes most of its money on hardware margins. Therefore, there will be a brief flirtation with binding MacOS-X-Intel to Apple-branded Intel-based systems. Despite fruitless (no pun intended) lawsuits to combat the practice, installer patches will rapidly be developed and widely spread to allow MacOS-X-Intel to be installed on any x86-based system (thereby increasing the popularity and spread of MacOS-X, but Jobs will almost certainly be incapable of seeing that, at least initially). Apart from their industrial design, which is absolutely first rate, there will be nothing to recommend Apple-designed boxes over dirt-cheap clones. System performance will be below par compared to other Intel-based offerings, and not enough people will be interested in paying the $500 premium just to get a pretty box.
The reverse transition will also be true: People will try installing Windows on an Intel-based Apple box. Apple will try (and fail) to prevent this, too.
Moving to the Intel-based platforms places Apple in direct competition with Microsoft. The relationship between Apple and Microsoft has long been one of, shall we say, détente. This state has survived because neither has directly tried to enter the other's playground (there is no version of Windows for PowerPC). There may even be secret agreements between the two companies to maintain this state -- indeed, such agreements may be the driving force behind Apple's initial attempts to keep Windows off Apple-branded Intel machines, and MacOS-X-Intel off clone machines. By supporting the Intel platform, Microsoft may feel itself no longer bound by such "gentlemen's agreements," and start pulling overtly dirty tricks to undermine MacOS-X. Expect to see threats of Microsoft ending support for MacOS-X versions of Office. Expect also to see Microsoft even more shamelessly mimic the MacOS look-and-feel in upcoming Windows releases (Shorthair^WLonghorn is still far enough off that it could be completely re-specced).
Once Apple realizes that it can't bind MacOS-X only to its own machines, they will attempt to form OEM relationships with the major PC manufacturers (Dell, HP, IBM, etc.). They will then run smack-bang into the same wall Jean-Louis Gassée did when he tried to get BeOS bundled with PCs -- Microsoft won't let them. The anti-trust accusations will heat up again, this time with Apple behind it. By the time it reaches the courts, George W. Puppet and his operators will no longer be in office, so it's impossible to predict what the political pressures might be. Microsoft will start at a disadvantage, since it already has a criminal judgement against it, but a lot will depend on political orientation of the Justice Department and the Federal Trade Commission.
In short, Apple has stepped into a very different and very messy business landscape. I sincerely hope they're ready for it; I'd hate to see them go away.
Schwab
You bought a 22" LaCie CRT? Damn, dude, those things are expensive. I've been thinking about replacing my 19" Hitachi CM751, which is starting to show its age, and was always kinda fuzzy at 1280*1024 and above. LaCie looks like a worthy replacement but, because they're professional color-balanced monitors, they command a premium over ordinary CRTs. I don't need precise gamma or color balancing; all I really want is 1600*1200@75Hz with no ghosting or fuzz, and without the smear of LCDs.
Schwab
Easy. The driver gets a 'sync' command from the OS. However, the driver writer believes that most other programmers call fsync() when they don't really need to, and decides to "optimize" this case. So he passes the command on to the drive, but returns immediately (allowing the drive command to complete asynchronously). This makes his driver appear faster.
Fortunately, most driver writers have their priorities straight about data integrity, so this kind of thinking isn't very common.
Schwab
There is historical precedent for this. There were recorded incidents of drives corrupting themselves when the OS, during shutdown, tried to flush buffers to the disk just before killing power. The drive said, "I'm done," when it really wasn't, and the OS said Okay, and killed power. This was relatively common on systems with older, slower disks that had been retrofitted with faster CPUs.
However, once these incidents started ocurring, the issue was supposed to have been fixed. Clearly, closer study is needed here to discover what's really going on.
Schwab
If you are a 31337 g4m3r, integrated graphics is indeed a joke. However, it's good enough for 85% of the users out there, who will almost never run anything more intensive than Word, IE, and the occasional Flash-based game. Same deal with integrated sound -- for Windows event beeps and boops, it's more than plenty.
10/100 Mb/s speeds are now common on integrated Ethernet controllers, and most of them have very little braindamage these days. 1Gb/s on-chip controllers are also already starting to appear.
To put it another way: Parallel, RS-232 serial, and PS/2 mouse/keyboard ports used to require separate expansion cards. Today, they are integrated into the motherboard chipset, and no one thinks the worse for this. For those who need extra ports or special high-performance ports, third-party PCI expansion cards are still available.
So, in short, the way systems are being put together these days, there's no cost savings to be had by breaking out the peripherals you don't need. If you feel a need to put the old parts to good use, donate them to a school, or use them to build a Frankenbox on which to do kernel or driver development :-).
Schwab
"...Unless the public policy in question is copyright or anti-trust law. Then we're all over it."
Schwab
Let me be clear: I am no fan of the current copyright regime, since it's at loggerheads with the fundamental nature of computers. The DMCA is an especially execrable piece of... legislation. But by publishing the ISOs to games, the people behind this reverse-engineering effort are almost immediately discrediting the value of their good work. Sony will calmly stand before a judge and say, "See? There's no academic or social or Fair Use argument here. Upon successfully bypassing our security regime, the very first thing these criminals did was copy and publish the games. We therefore ask for a permanent injunction."
Posting the ISOs was completely unnecessary. Now that the discs are (apparently) readable, anybody interested in reverse-engineering the games themselves could just as easily bought their own copy and worked from that.
I expect this to end poorly...
Schwab
Back in 2000, duing the California power "crisis," Amory Lovins of the Rocky Mountain Institute was asked what things citizens could do to conserve power. His response: "Conserve water. The lagest consumer of power in California is electric water pumps. So if you save water, you'll save power."
Still, every little bit helps. By residents switching over from incandescents to screw-in fluorescents duing the power "crisis," California reversed approximately 8-10 years of power consumption increase (according to some estimates).
Schwab
Actually, the statement is mostly correct; the law itself is mostly just fine. What's horribly broken is the patent examination and granting process. The examiners have done a shameful job in maintaining the integrity of patents by allowing patents on trivial "inventions" or re-purposing of existing inventions; by allowing patents that do not fully describe how to re-implement the claimed invention(s); and by allowing patents that are nearly unreadable with legalese and deliberately vague language.
Fix the examination and approval process, and the patent system will almost certainly sort itself out again without any legislative changes.
Schwab
Yeah, right.
Just ask any company that has "partnered" with Microsoft how well it's worked out for them. The list of mangled corpses is long and grotesque indeed.
Interoperating with OSS is no secret. Read the RFCs, download the sources, and knock yourself out. An ersatz "partnership" is not required.
I say ignore them.
Schwab
Just so; I stand corrected. There are a large variety of narrow-market niche applications that are extremely mature with an equally mature user base, and which are only available for Windows. (OrthoTrac is one example that comes to mind.) F/OSS duplicates of these products would be extremely non-trivial.
Schwab
Ah, the joys of the broad-stroked brush. Let's take this apart, shall we?
Which aspect of gaming: the server or the client? If you're talking about the client then, sadly, yes, Windows holds the edge here, since the graphics and sound drivers are more mature and better supported. However, for a server -- especially a public one -- you'd be a complete fool to put anything less secure than a Linux box on the net. Even better to put up one of the BSD variants.
The strokes don't get broader than this. But basically, all Windows is good for here is running Office (Word, Excel, Powerpoint). Everything else is better off running on Linux: Intranet Web servers, email servers, file servers, backup servers, Oracle, and corporate firewall.
There is nothing magic about streaming media, especially when it's done from a bunch of audio files stored on disk. Windows' proprietary DirectX brings nothing to the table here. So that means selecting for a robust, secure server platform, which makes the choice fairly obvious...
Schwab
It Just Works(R)
______________
Microsoft End-User Unilateral Non-Negotiable License "Agreement:"
[ ... ]
92. Definition of Terms
[ ... ]
xxi. "Working" shall be defined as any software product or service that can reliably and repeatably produce Guaranteed Functionality.
[ ... ]
clxiv: "Guaranteed Functionality" shall be limited to the Blue Screen of Death or any other diagnostic output or display. All other software functionality shall be deemed Extensions or Enhancements and may or may not operate with any degree of reliability.
____________
Schwab
So if Adobe feels even the slightest bit threatened by it, expect the project to receive a nastygram.
Schwab
ATTENTION: ALERT LEVEL UPDATE. The authorities at SANS (Sebben-Affilliated Network Security) have issued this network alert update:
The DNS cache poisoning alert has been upgraded from "Yellow" to "Blackwatch Plaid." Repeat: DNS cache poisoning alert level is now at Blackwatch Plaid.
Available information does not yet justify a further upgrade to alert level "Moving Pictures."
And for everyone's safety and security, and to preserve our way of life, SANS is taking a drastic step and installing a network monitor. Just one. For safety, security, and omniscient, unblinking information gathering of everyone's activities.
Schwab
Very funny. Fortunately, as we all know, Sony does not have the right to make, much less enforce, this request. "Authorization" for opening and/or modifying the machine is embodied in the receipt from the retailer. Sony can void your warranty, of course, but you knew that when you started unscrewing the thing.
Let us not take such histrionics seriously; it gives people the wrong idea.
Schwab
Since it's Freescale (née Motorola) that's mentioned in the article, any general-purpose CPU appearing from this effort will probably be ARM-based. However, the most likely application will be specialized processors for multi-gigabit network routers.
Schwab
Schwab
So, let's see here. For $250, you get a cool-looking but big and chunky thing that's a closed platform with proprietary media formats and only a handful of games.
On the other hand, for just $20 more, you can get a Tapwave Zodiac with:
It also supports Web browsing and POP/IMAP email via Bluetooth networking or an 802.11b SDIO card (sold separately).
All new platforms have teething pains, and Sony will undoubtedly be addressing the PSP's shortcomings over the following months (including its rather high price). But if you don't want to wait, or you'd rather have something you can actually comfortably carry around with you, you might want to amble on over to CompUSA or Fry's and take a look at the Zodiac.
And I'm not just saying that because I work for them :-).
Schwab