How about for overreacting to the nth degreee. Would a parent/teacher sitdown and talk meeting have accomplished much more. Panic is the downfall of sentient species. The teacher, principal and judge, all panicked (sp).
Ah, but why?
See, everyone knew it was wrong. Instead of crying, it was wrong, it was obvious, lets find out why they believed a sit-down wouldn't be enough, and deal with that problem.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Re:Why Linux Supporters Should Be Excited About X-
on
More on the MS "X-Box"
·
· Score: 2
No, it'd simply take Microsoft to add a few hardware specific catches in there, and not publish them.
Hardware specific catches break compatibility, lead to lots of headaches down the road, and don't even stand up to a decent amount of reverse engineering on a (wide) open platform like Windows, where every call is just a SoftICE away.
Offer an SDK, and do the same thing 3dfx does with Glide.. FORCE them to use it. If another company truies to write something compatible, sue their pants off.
Not their style. They're using standard hardware for a reason.
This isn't the first time MS has entered a market with overly restrictive incumbents. They're counting on the fact that they'll be much more open than their competitors to make them money.
And this has NOTHING TO DO with OSS. Personally, I see more promise in OpenGL then DirectX. I see better hardware support for OpenGL, and I see better performance out of OpenGL.
I don't know about this. OpenGL seems to be all about having the driver authors do more work, presuming they're much more knowledgable about how their product wants to draw a primitive.
Direct3D, as far as I can tell, is just much, much easier to write drivers for--more stuff is left to the game developer to implement. Are there any experts out there who can verify?
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Suppose this kid did go ahead and shoot up his school. Suddenly, this teacher would lose her job for not doing anything after such an obvious threat. 20/20 hindsight means you gotta cover your hind quarters.
The administrators had no choice--nobody wanted to be the scapegoat who didn't do anything, not the teacher, not the principal, not the judge.
Grammatical errors aside, it sure was a scary story. Scared the entire political and educational infrastructure half to death about losing their job.
One has to wonder how wrong they were to. Things are rarely so simple, Jon.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Most of Microsoft Games (the good ones anyway, not the "Return of the Arcade" serieses) are licenced by Microsoft or published through them. I wonder why such a large company can't get its shit together when it comes to actually making a good game.
According to John Carmack, size is actually a detractor in making really good games. He's been quoted as saying that "Three programmers, three artists, and three level designers can make the greatest games in the world."
Too many designers spoil the vision.
(Although they do make a good flight game, MS Flight Simulator being this one exception, though my coworker who works on the project says there are so many people working on it, it's any wonder things get done. It would be cheaper for them to outsource it as well.)
Flight Sim is a legacy game--it's part of the definition of what Microsoft is. It may be an albatross of a division, but most of the people in there probably grew up playing it. It's part of their corporate identity, I'm sure.
I don't know in fact, but it's likely the hardware is made by a seperate entity.
It's definitely separately managed; they don't do everything perfect but their quality is noticably higher than the software divisions'.
Anyway, the hardware is decent, but definitely not ergonomic, it just looks that way.
No, the iMac mouse is definitely not ergonomic. What complaints do you levy against MS hardware?
The scrollwheel is just brilliantly implemented, down to the steppage to provide the stop.
The Windows key is tragically named but is an amazing concession to a design concept I wish Linux subscribed more to--everything should be doable via keyboard. I use the Windows key all the time in 98(is there a KDE patch yet?), and respect the Right Mouse Key for what it is.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Re:Why Linux Supporters Should Be Excited About X-
on
More on the MS "X-Box"
·
· Score: 2
I'm confused.
Don't worry about it. Understanding the "enemy" is much more fulfilling than hating a bogeyman. I actually agree with most of your analysis.
You think MS should be credited for competing?
Heh, when did I say this? I said it's cool that they're creating a much more open platform than what we've got now.
What's so noble about that?
Cool Product != Noble Aims. Embrace and Extend is about as ignoble as it gets.
The only reason MS is offering this "open and free" development platform is because they have no choice. Sony and Sega have the market locked down. Microsoft's original plan to make the PC the premier gaming platform hasn't exactly panned out and they're not too happy about that.
Sure, they've got a choice. Kick Sega, Sony, and Nintendo around for their relatively ridiculous developer relation$, or don't.
Look. The X-Box isn't something like...say, removing Knowledge Base entries that help Windows work with Samba. Or removing critical files from their online archives, demanding that you call tech $upport to get the patch to a PPTP bug. Or revoking Compaq's right to sell Windows because they dared to remove the Internet Explorer icon.
Complaining about the X-Box because it's a Windows compatible platform is kinda like complaining about those FreePC deals because, my god, they're free PC's that can run WINDOWS!
MS is in this to make sure that developers work on MS platforms, using MS development products. They only care about openess and freedom when it's the only way to enter a market.
MS keeps you as open and free as possible as long as you're using their tools. Openness and Freedom is historically their major selling point, and they'll be as closed and violent as they can get away with to prevent others from intruding on their domain in this regard.
[WARNING: THE FOLLOWING ANALOGY IS OFFENSIVE BEYOND BELIEF.] Microsoft is really no more complicated than your average hooker--she's got the goods, it's $49.95, and she'll claw the eyes out of anyone else who tries to do the same on her corner.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
No it won't. The idea is that it will be on millions and millions of homes and will run _Windows_.
It'll run an open PC platform. Whatever OS is loaded is whatever OS is loaded.
The anticompetitive hardware that would prevent this would attract farrrrrr too much legal/PR flack that MS doesn't need for any reason.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Re:Why Linux Supporters Should Be Excited About X-
on
More on the MS "X-Box"
·
· Score: 2
I disagree that this will ever be as 'open' as you say, and certainly no more open then Sony using Linux as it's primary development platform for the PS 2.
You automagically locked in to, more then likely, Visual Studio for development, and Direct X as the primary graphics library, neither of which are noted for being superb for game development.
Those are the standard development tools. If other people want to use other environments, it would take some serious anticompetitive hardware that wouldn't survive a moment of legal/PR scrutiny to literally MANDATE the usage of the Windows OS on otherwise standard PC Hardware.
Microsoft is likely to actually Do The Right Thing for one reason or another here and posit that game developers would rather develop their high end code using the many services embedded in DirectX7 than use whatever the OSS guys are doing.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Whoaoooaaaaa, Sharky's missing a bit of history in their report.
3dfx' partnership with SEGA for the Dural (Sega's code name two years ago for the Dreamcast) fell apart some 18 months ago and although both parties seemed to be satisfied with the end result the real reason for the 'split' then came down to capabilities in terms of supply and yield levels.
Oh, my ass both sides were satisfied!
This was nothing short of a minor scandal some time back. Turned out Sega got some pressure from someone, somewhere, and decided to go with a Japanese company(NEC, in a beautiful twist of irony for those who remember the NEC TurboGraphix 16) for the core chipset in their Dreamcast.
This was after signing with 3DFX and extracting detailed technical specifications out of them.
Needless to say, 3DFX went ballistic, and (probably correctly) accused Sega of delivering their prized designs to NEC, who at the time was still theoretically going to release their tile based PowerVR2 monster chip.
I actually don't think that chipset ended being used in the DreamCast(I lost track--so sue me), but I definitely remember a pretty massive settlement for breach of contract etc. against Sega for their 3DFX fiasco.
Lots of drama, somebody go post some links.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Why Linux Supporters Should Be Excited About X-Box
on
More on the MS "X-Box"
·
· Score: 5
I've been saying this for a while: Most actions of Microsoft can be analyzed under a "Chains For Freedom" Philosophy: Microsoft wants to make the general public as free as possible, as long as they're the ones to deliver the freedom.
That in mind, consider the values of the X-Box. Whereas most console makers tightly restrict development on their machines, game developers will likely be free to release whatever they like for the X-Box.
Expect, of course, an extremely quickloading and game-customized version of Windows to have a per unit cost on each Game DVD. Also expect this to be significantly cheaper than Sony or Sega mandates per unit.
Don't expect non-standard media. Microsoft, for all the embracing and extending it's accused of(justifiably), has an excellent hardware department. I've never, ever been more embarassed of Slashdotters as when I watched them rip apart the touch-mouse--this was an impressive technology that would have been worshipped had it come from Logitech and shipped with Linux drivers. Anyway, expect their department to build in an industry standard DVD-ROM, likely one that can play DVD-RAM disks so as to facilitate ease of use for system developers.
Developers become "free" from propietary startup hardware costs, and "shackled" by having to code to Microsoft standards. It's pretty interesting to watch.
Linux is the wildcard. While Sony is releasing Linux development code for PS2, Linux should actually *run* on this K6(!!!) based system. This, above all else, should excite the heck out of us. Here we have a cheap system that will very likely be in millions and millions of homes, have 10/100 Ethernet built in(oops, Sega!), significant processing power($20 says they put the SB Live audio chipset in it, btw), and will run Linux.
Beyond the fact that we will see developers using Linux as their game OS rather than the one-off Windows from MS, here's a console that one can actually pawn computational work on. It's a console free of almost any legal demands and enforcements of its creator.
It's an open platform, from Microsoft.
Recognizing that this is a cool thing gives us more credibility when we accuse MS of the kind of junk described in the Halloween papers.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
You obviously work for RealNetworks or otherwise benefit from their largess, so why don't you stop polluting this thread with your corporate PR -/. is a forum for people, not corporations.
Unfair. Corporations have every right to defend themselves, and there's no reason to believe that A Nonymous Coward is really a RealNetworks employee. (Yes, people can doubt me without having an ulterior motive.)
His point is rational--the claim could be taken to mean that RealNetworks reports all MP3s encoded by them and nothing else. It's plausable, but I'd be qiote pissed at the Times--Number of MP3s Encoded != Number of MP3s on the Hard Drive. (Still, there's a pretty reasonable amount of privacy violation even without the extra-software spying.)
The only way to check is to rip out a copy of FileMon and see what RealNetworks is really up to. If I get some free time, I'll do this myself.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
(BTW: No reason to be anonymous. I prefer to respond to people, not "entities"--You Are Your Words. Own them.)
Richard Smith, a Brookline, Massachusetts-based independent security consultant, said the numbers of songs stored on a user's hard drive, the kind of file formats in which the songs are stored, the user's preferred genre of music, and the type of portable music player, if any, the user has connected to the computer are sent to the company, the Times said.
This is my evidence(and my first paragraph from the post you responded to). If it's wrong, I self-flagellate myself upon the battered journalistic integrity of the above. RealNetworks didn't particularly refute any of this, and I'm sure they'd be screaming bloody f*cking murder if they were accused of taking one iota of extra data.
AC, I would be laughing myself to tears if this was all about mere listening patterns. That's NOT what the evidence suggests.
Do you have any evidence we don't know about?
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
I have only one question here: Did the company listen to the outrage of thousands of customers over the privacy violation or the 1-7/8 drop in their stock?
And me without my moderator points. Ah well, such is the pain for posting in this discussion.
Excellent observation.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Sorry, Mike. You summarized things a bit too well:
"If I decide to put up mikenash.com and I want to sell T-Shirts with my picture on them, for something uninteresting like me five CALs is all I need since I probably won't have more than five people buying at one time," Nash predicted.
Probably, eh? And what if I do? What if, say, Slashdot links to the T-Shirt site I'm going to open up someday and--amazingly enough--I have some T-Shirt that's surprisingly popular. Far more than its been. Are you telling me that, while Apache-SSL would be more than happy to accept as many credit card orders as the server could possibly handle, Windows 2000 would tell my customers to go away because I didn't give Microsoft enough money?
Are you kidding me?
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Deep within the nestled chroot()'d directories of Rob Malda's Precision Cut Head lies Slashdot...but nobody knew where this ever growing tumor of a side project turned major stock came from from. Nobody knew...until:
And if any of Townshend's PR people stumble on this, I wanna interview Pete for Slashdot.
Slashdot is just a gigantic ploy to let Rob Malda satisfy his lifelong wish to talk to Pete Townshend!
"...and thus it was written in the Book Of Dot, 'O Lord, How May I Speak To Your Musical Prophet', and the Lord replied, 'Yea, ye shall verily speak with thy prophet, but spread far and wide my message of openness and birds that do fail to fly. For that is the way to happiness; the road to bethlehem is with the Geek!"
Yours (Hopefully) Amusingly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
But that's the general rule of consumer computing hardware - don't expect any real quality unless it translates directly into higher Quake benchmarks.
Oh, you'd be surprised how amazingly correct you are. I blame the death of Cyrix on John Carmack and the amazing fact that Quake--the only application to ever demand serious FPU speeds, and the most Pentium-tuned app in history--ran like a dog on anything but Intel. That 3DFX was the only credible 3D company for quite a few years(an amazing amount of time for a peripheral manufacturer in this industry) was also Carmack's doing.
Yes, Carmack is a programming god. Just look up the Law of Unintended Consequences, however.
In contrast, Carmack's utter rejection of the Direct3D Execute Buffer disaster(tuned specifically for MS's Talisman architecture) translated directly into a reasonably predictable redesign of Direct3D.
All parts that are likely to fail in a system should have a standard of reliability they are required to meet in order to reach some government/industry certification. Couple this with an mandate to advertise consumer awareness at the level of "Intel Inside" and you force the market to a minimum but acceptable level of quality.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Richard Smith, a Brookline, Massachusetts-based independent security consultant, said the numbers of songs stored on a user's hard drive, the kind of file formats in which the songs are stored, the user's preferred genre of music, and the type of portable music player, if any, the user has connected to the computer are sent to the company, the Times said.
People, this isn't just RealNetworks incidentally receiving information on what CDs you have by nature of that being the only way to send back the track titles.
RealNetworks invasively scanned millions of American's computers for content that had nothing to do with the functioning behavior of RealNetworks software. We're talking about code that looked for MP3s, music applications, hardware interface tools, and who else knows--I wouldn't look for RealNetworks to tell.
Open Source is many things, but I'd seriously rather it not degrade into the only way to trust that code isn't Trojan'd. I expect that kind of paranoia for my cryptology of choice, not to play some Garbage!
This isn't an issue about a few missing lines from a privacy statement. Should RealNetworks be able to upload any interesting file on your hard drive to the corporate servers as long as they mention that "From time to time, RealNetworks may request feedback from your internal storage systems according to specific parameters to be determined according to your usage profile"? Maybe it'd be fine for them to tap into your computer's microphone, as long as they don't neglect to tack on "User agrees to indemnify RealNetworks from any liability in relation to any data flowing through said user's Sound Card"?
This isn't about legality, at least, not yet. It's about trust, and RealNetworks is losing mine fast.
The real question is, whether TrustE will follow.
I'm no history expert, but there's an aspect of TrustE that just smacks of the ill-fated League of Nations from the first part of the century. Namely, the well-intentioned but utterly toothless, powerless, and secretly mocked nature of it. I think TrustE actually has enough Respect Capital(if there is such a thing) with the press to actually do something, this one time...
Or never again, because nobody will listen anymore.
TrustE needs to set up guidelines of what may be buried in the fine print and what needs explicit and large dialogs before the function is completed--yes, this includes specifications like "Default must be no, and the software must still run even if it isn't allowed to insert seven links to the audio playing software like RealPlayer G2 does--we counted." That's clear, from RealNetwork's rather shocking behavior.
The bottom line is TrustE simply needs to file suit for breach of contract and reach a settlement where RealNetworks needs to contact all possible users, mass deploy a tremendous upgrade, and notify victims of the violations in both online and TV/Magazine forums.
That, or some combination with what I'd like to call TrustEeth: Privacy Protected for x Days.
If you think about it, it's really just a much more positive version of "This Site Accident Free for x Days" signs. The system encourages TrustE certification, since the longer one puts it off, the longer it will take to get to privacy levels respected by customers. It will make it progressively more expensive over time for large companies to allow their ego to overpower the rights of their customers--the CEO will be quite peeved at the middle manager who took the nationwide corporation down to one day of privacy protection.
If not a system using literal days, then an accumulation of points, lowered by violations, maintained by fair and quick resolution of privacy concerns, and accelerated by respectful "voluntary" policies could also be functional.
The key is, people need to have a gauge by which they can determine whether or not to trust a site and the code it asks them to download, and managers need to know they could get called on the carpet if they try a stunt like RealNetworks did.
The irony is truly remarkable, if you ask me. The CEO of RealNetworks(then Progressive Networks, if I remember correctly) went and testified in front of The United States House Of Representatives, arguing against everybody's favorite monopolist, Microsoft, was making the playing field unfair.
Meanwhile, here we are in November of 1999, and RealNetworks is repeating the sin that Microsoft did wayyyy back in the day with its overly nosy Registration Wizard that reported if software like Wordperfect was installed. Incidentally, the above dig at RealPlayer G2 for the seven links it litters all over your desktop(collect them all) is even more beautifully ironic considering the now strangely difficult to find position paper regarding asking the user before doing anything of import.
On a plus note, I don't think the US Patent Office had anything to do with this one.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
DirectSound3D was hacked together by the engineers at Aureal. They did it, they wrote it, MS used it. At least that's how it was explained to me back at WinHEC. (Ironic. The only convention I've ever paid money to attend was WinHEC '97. Definitely was worth it though, if only to see the Rendition guy get disturbingly flustered when I started comparing their product to the Voodoo solutions coming out of 3DFX. Incidentally, this was also the first the world had seen of the Riva 128, A.K.A. the return of nVidia. nVidia was the first 3D card manufacturer for PCs, but their card was so slow that it was better just to operate in software. Telling people nVidia was coming back from the grave was like announcing PC Chips as the one of the higher quality manufacturers in the industry...heh, wait a second...)
As for me, I use a SB Live Value because I compose music and the SB Live's MIDI support is light years beyond anything the rest of the industry has supplied us with. I'm truly looking forward to seeing what audio hackers can do with the power of that DSP.
I've still always lusted after a Vortex 2, though, and might just grab one for the hell of it.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
John Dvorak wrote an article around 1989 lambasting the computer industry for the complete and total lack of quality that 5.25" floppies possessed. They were slow, they were loud, and most disturbingly, they lost data with ridiculous ease. (Somebody find me a link to this article. I need it for my files.)
Ten years later, we're using 3.5" disks, and we're in the same situation.
Steve Jobs and the iMac development team had a chance to fix the situation. They recognized that floppies were garbage. They knew they weren't stable by any means. In response...they chose nothing.
Thanks, Steve.
So, floppy disks persist as the one and only truly standard way for moving documents and small files from one machine to another, and maybe even storing them off a hard drive for some period of time. Most of us who build our own computers know--buy Teac, or you might as well throw away your data. Most of us who build our own computers do so because ten more dollars for a part that can be trusted is worth it for us, but for a large scale company churning out thousands of boxes, ten more dollars per unit is suicide if you can get away with using it to pad the bottom line.
This is one of those "diseconomies of scale" when it comes to large computer companies--saving pennies on a box to remove a feature that everybody expects but only 1% know to check for(like supporting more than 32MB RAM on a certain motherboard, Compaq)--that keep the little guys in business.
Anyway, I must say I'm not entirely displeased with the tremendous pain about to be inflicted on those who have been knowingly distributing fatally flawed 3.5" disk technology in an effort to save money. I spent two years as a volunteer tech at my university--the amount of raw labor I saw go down the drain because of floppies gone terribly wrong was shocking, as was the amount of disks that simply couldn't handle a Linux boot disk. Imagining large corporations going through the same kind of pain I've watched a relatively small population of students go through is frightening, to say the least.
Of course, there will be some companies defending their negligence by saying the minimal quality standards were "common industry practice". So too were the falsified 15"/17" monitor statistics that lead to the "Viewable Size in Every Advertisement" agreement. Lets not even begin to mention the coming slap down of the entire 56K modem industry, which was all too happy to claim speeds twice as fast when even in the most generous contexts it wasn't the case. And, of course, sooner or later, Inkjet Printer manufacturers will get their due--8 pages a minute? Yeah, if you're printing a period.
But what's critical is that while faking monitor sizes only strained a few eyeballs, and slow modems and printers maybe caused a missed deadline or two, substandard(hell, plain old standard) floppy drives caused data loss that directly led to wasted employee hours and lost property.
I can't have too much mercy here. I've done the recoveries, I've gotten the pleading phone calls, I've thanked Word's Autorecovery feature innumerable times(and cursed Windows' awkwardness at saving to hard drive and making only a copy for the floppy).
It may have taken a very long while for the industry to be taken to task on this, but hopefully we'll finally see a stable portable media standard arise from the legal ashes.
About time.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
Aureal has always been one of my favortie companies, and their recent announcement is quite intriguing.
Much of the industry is obsessed with "meeting the marks"--x amount of video RAM, AGP/PCI/VLB(way back in the day), Direct3D, OpenGL, etc.
Essentially, the complexities of various products are simplified down to a set of necessary functions supported and some "speed scale" created by a semiobjective source.
Creative Labs' announcement signified not only a new level of support for Open Source driver development models, but also a relatively drastic turnaround in company policy. It wasn't that long ago that Creative was flatly refusing to open their drivers to anyone but OSS Sound Inc. under NDA. Now they're saying they'll support any and all comers, and even overcome the starting barriers(CVS/Bugzilla) so that development can begin ASAP.
Such full fledged support is reserved only for items considered integral for the success of a product--in other words, a line item, a necessary feature.
A checkmark.
That Aureal(whose stock symbol always seems to confuse the hell out of every engine I check it with) is following suit means that, for the first time, not only Linux support but full, open access to driver development infrastructure is becoming a line item checkmark for a segment of the PC Hardware Expansion industry.
Such developments bode well for future developers, who will hopefully not need to painfully reverse engineer nearly as much in their attempts to get network cards, 3D Graphics boards, or any other custom hardware to function now and long after a company decides to cease support for a given product.
When ZD Net reports closed drivers as a downside to a given piece of hardware, we'll have truly won.
However, it should be noted that while the SB Live is a full DSP architecture that developers should be able to exploit to unimaginable degrees(though I'm not expecting a Perl RegEx module using the SB Live *LOL*), much more content is hardcoded within the Aureal chipset. Indeed, this is a risk of a open development process--companies may feel that, to protect their intellectual property, they need to lock it up within the chipset instead of releasing it in the drivers.
Of course, a chipset and a card that remains viable for significantly longer periods of time is far more appealing to both OEMs and consumers, and this is a benefit that both programmable circuitry and open development foster.
This isn't to say, though, that I'm not extremely excited about Aureal's announcement. I first heard Aureal's A3D at WinHEC '97 a few years back, and it was the first time I had truly heard 3D Sound that actually worked. Aureal is one of the great stock price tragedies of the last few years, mainly because of some rather nasty innuendo and patent litigation. (Yes Virginia, that ugly USPTO spectre pops up again.) Their technology is sound(no pun intended), and their cards are uniformly high quality. I'm quite interested at seeing what the Linux community will do with access to the extensive 3D Sound modeling that Aureal has implemented--literal 3D environment models are used to determine reflections, reverbs, etc. It's all quite amazing, and very, very well implemented.
It's definitely an interesting time to be a sound engineer in the computer industry. One thought--BeOS ought to be supporting some new very powerful sound processing cards very soon, if it doesn't support them already.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
A bit more thought on this: Wow, creative is setting up CVS/Bugzilla. They're not merely opening the source; they're not just trying to grasp a bit of extra PR out of the Linux mindshare gods(Taco and Hemos:-). They're actually going the extra mile and providing not only the source but a development environment for coders to come, watch, and learn. This is amazing, and deserves a retrospective profile in around six months to see how this great, precedent setting experiment panned out. Of course, Creative isn't dumb. As I mentioned in another post, Creative stands to have their card become the standard DSP component in innumerable Linux machines--their foresight in developing a programmable sound card is very likely to pay off handsomely in increased sales. The economics of Open Source just got much more interesting. Yours Truly, Dan Kaminsky DoxPara Research http://www.doxpara.com
First of all, there's quite a bit of difference between fastest and fastest known. I can't imagine both the chinese AND the american governments from having some exceedingly classified hardware that blows the pants off the open stuff(read: governmental phallus-phlashing.) Second, the meaning of fastest is very unclear. I'd go so far to say that any system that implements a given function in software instead of hardware is going to be orders of magnitude slower than the state of the art. Witness the EFF DES cracking machine, 3D Graphics Accelerators, even Math Coprocessors. Fitting a square peg into a round hole is actually a pretty common occurance in the computer world, but it takes a relatively tortoise-like rate compared to what can be pulled off with raw gates. That's why XISC--Extensible Instruction Set Computing--is probably the upcoming processor paradigm. Programmers need the ability to redefine round holes into square ones, so the square pegs fit right in.
Yours Truly, Dan Kaminsky DoxPara Research http://www.doxpara.com
Alot of people aren't going to get the significance of this announcement--they'll think, "Cool, another sound card that I can play with on Linux w/o resorting to closed source drivers." This is beyond that. The SB Live is based on a single, ridiculously powerful and extremely programmable DSP. Almost all functions the chipset performs are executed in DSP level software, meaning suddenly Linux is getting full specs on a complete digital processing environment. The impacts of this are substantial. The SB Live chipset is cheap enough that it's a contender for "standard sound" in many machines. Open source algorithms for everything from MP3 encoding to analog synth simulation to the more esoteric, non-sound related stuff(GIMP graphical filters, datastream analysis, etc.) should, if the drivers are clean enough, start popping up over time. The uses of such a powerful digital signal processor on an open platform are honestly unpredictable at this point in time. While there are hardcoded design issues in the SB Live chipset(most notably, all signals are upsampled to 48khz before processing may occur), the sheer flexibility of this chipset will blow Linux programmers out of the water. This is truly excellent news, and shouldn't be ignored as a mere fun thing for the gamers to play with. If only 3D graphics hardware was as programmable...or as open. Yours Truly, Dan Kaminsky DoxPara Research http://www.doxpara.com
That's like saying that cars these days have gotten quite good at protecting their drivers from fatal crashes. The statement may be true, but that sort of thing should still NEVER happen. It is to be avoided at almost all costs.
No. You don't understand.
If you slam the power button on a FAT/FAT32 box, you're not gonna lose the partition.
You can't say the same for a Linux box using ext2, or even a Solaris box using UFS. From *VERY PAINFUL PERSONAL EXPERIENCE*, you have quite a decent chance of damaging some serious stuff, and way more than an unheard of possibility of just completely losing the filesystem.
FAT/FAT32 can recover from random reboots without a problem. It's simple enough to just not have the same kind of problems as Linux w/ ext2.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
When was the last time you sat through a Linux boot and upon execution of "scandisk" were required to hit "Fix" fourty-one-thousand and three times because the UI designers decided it would be too hard to add a "Fix All" button? e2fsck -p, my friend.
Be fair. (It's good for credibility.) Any time there's a semi-serious problem, you're gonna be hitting y for quite a while w/ fsck.
I've lost entire file systems more than a few times because of an unscheduled reboot, incidentally. The same has happened, incidentally, w/ NTFS, but never, ever, ever with FAT/FAT32.
It's actually enough that there's a semi-decent chance I'll make my MP3 partition a Fat32 one.
Yours Truly,
Dan Kaminsky DoxPara Research http://www.doxpara.com
P.S. Don't tell me Solaris is any better; it made some noises significantly scarier than "extra bytes discovered" when I recently bungled a shutdown.
Slashdot Mirror
How about for overreacting to the nth degreee. Would a parent/teacher sitdown and talk meeting have accomplished much more. Panic is the downfall of sentient species. The teacher, principal and judge, all panicked (sp).
Ah, but why?
See, everyone knew it was wrong. Instead of crying, it was wrong, it was obvious, lets find out why they believed a sit-down wouldn't be enough, and deal with that problem.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
No, it'd simply take Microsoft to add a few hardware specific catches in there, and not publish them.
Hardware specific catches break compatibility, lead to lots of headaches down the road, and don't even stand up to a decent amount of reverse engineering on a (wide) open platform like Windows, where every call is just a SoftICE away.
Offer an SDK, and do the same thing 3dfx does with Glide.. FORCE them to use it. If another company truies to write something compatible, sue their pants off.
Not their style. They're using standard hardware for a reason.
This isn't the first time MS has entered a market with overly restrictive incumbents. They're counting on the fact that they'll be much more open than their competitors to make them money.
And this has NOTHING TO DO with OSS. Personally, I see more promise in OpenGL then DirectX. I see better hardware support for OpenGL, and I see better performance out of OpenGL.
I don't know about this. OpenGL seems to be all about having the driver authors do more work, presuming they're much more knowledgable about how their product wants to draw a primitive.
Direct3D, as far as I can tell, is just much, much easier to write drivers for--more stuff is left to the game developer to implement. Are there any experts out there who can verify?
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Suppose this kid did go ahead and shoot up his school. Suddenly, this teacher would lose her job for not doing anything after such an obvious threat. 20/20 hindsight means you gotta cover your hind quarters.
The administrators had no choice--nobody wanted to be the scapegoat who didn't do anything, not the teacher, not the principal, not the judge.
Grammatical errors aside, it sure was a scary story. Scared the entire political and educational infrastructure half to death about losing their job.
One has to wonder how wrong they were to. Things are rarely so simple, Jon.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Most of Microsoft Games (the good ones anyway, not the "Return of the Arcade" serieses) are licenced by Microsoft or published through them. I wonder why such a large company can't get its shit together when it comes to actually making a good game.
According to John Carmack, size is actually a detractor in making really good games. He's been quoted as saying that "Three programmers, three artists, and three level designers can make the greatest games in the world."
Too many designers spoil the vision.
(Although they do make a good flight game, MS Flight Simulator being this one exception, though my coworker who works on the project says there are so many people working on it, it's any wonder things get done. It would be cheaper for them to outsource it as well.)
Flight Sim is a legacy game--it's part of the definition of what Microsoft is. It may be an albatross of a division, but most of the people in there probably grew up playing it. It's part of their corporate identity, I'm sure.
I don't know in fact, but it's likely the hardware is made by a seperate entity.
It's definitely separately managed; they don't do everything perfect but their quality is noticably higher than the software divisions'.
Anyway, the hardware is decent, but definitely not ergonomic, it just looks that way.
No, the iMac mouse is definitely not ergonomic. What complaints do you levy against MS hardware?
The scrollwheel is just brilliantly implemented, down to the steppage to provide the stop.
The Windows key is tragically named but is an amazing concession to a design concept I wish Linux subscribed more to--everything should be doable via keyboard. I use the Windows key all the time in 98(is there a KDE patch yet?), and respect the Right Mouse Key for what it is.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
I'm confused.
Don't worry about it. Understanding the "enemy" is much more fulfilling than hating a bogeyman. I actually agree with most of your analysis.
You think MS should be credited for competing?
Heh, when did I say this? I said it's cool that they're creating a much more open platform than what we've got now.
What's so noble about that?
Cool Product != Noble Aims. Embrace and Extend is about as ignoble as it gets.
The only reason MS is offering this "open and free" development platform is because they have no choice. Sony and Sega have the market locked down. Microsoft's original plan to make the PC the premier gaming platform hasn't exactly panned out and they're not too happy about that.
Sure, they've got a choice. Kick Sega, Sony, and Nintendo around for their relatively ridiculous developer relation$, or don't.
Look. The X-Box isn't something like...say, removing Knowledge Base entries that help Windows work with Samba. Or removing critical files from their online archives, demanding that you call tech $upport to get the patch to a PPTP bug. Or revoking Compaq's right to sell Windows because they dared to remove the Internet Explorer icon.
Complaining about the X-Box because it's a Windows compatible platform is kinda like complaining about those FreePC deals because, my god, they're free PC's that can run WINDOWS!
MS is in this to make sure that developers work on MS platforms, using MS development products. They only care about openess and freedom when it's the only way to enter a market.
MS keeps you as open and free as possible as long as you're using their tools. Openness and Freedom is historically their major selling point, and they'll be as closed and violent as they can get away with to prevent others from intruding on their domain in this regard.
[WARNING: THE FOLLOWING ANALOGY IS OFFENSIVE BEYOND BELIEF.] Microsoft is really no more complicated than your average hooker--she's got the goods, it's $49.95, and she'll claw the eyes out of anyone else who tries to do the same on her corner.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
No it won't. The idea is that it will be on millions and millions of homes and will run _Windows_.
It'll run an open PC platform. Whatever OS is loaded is whatever OS is loaded.
The anticompetitive hardware that would prevent this would attract farrrrrr too much legal/PR flack that MS doesn't need for any reason.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
I disagree that this will ever be as 'open' as you say, and certainly no more open then Sony using Linux as it's primary development platform for the PS 2.
You automagically locked in to, more then likely, Visual Studio for development, and Direct X as the primary graphics library, neither of which are noted for being superb for game development.
Those are the standard development tools. If other people want to use other environments, it would take some serious anticompetitive hardware that wouldn't survive a moment of legal/PR scrutiny to literally MANDATE the usage of the Windows OS on otherwise standard PC Hardware.
Microsoft is likely to actually Do The Right Thing for one reason or another here and posit that game developers would rather develop their high end code using the many services embedded in DirectX7 than use whatever the OSS guys are doing.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Whoaoooaaaaa, Sharky's missing a bit of history in their report.
3dfx' partnership with SEGA for the Dural (Sega's code name two years ago for the Dreamcast) fell apart some 18 months ago and although both parties seemed to be satisfied with the end result the real reason for the 'split' then came down to capabilities in terms of supply and yield levels.
Oh, my ass both sides were satisfied!
This was nothing short of a minor scandal some time back. Turned out Sega got some pressure from someone, somewhere, and decided to go with a Japanese company(NEC, in a beautiful twist of irony for those who remember the NEC TurboGraphix 16) for the core chipset in their Dreamcast.
This was after signing with 3DFX and extracting detailed technical specifications out of them.
Needless to say, 3DFX went ballistic, and (probably correctly) accused Sega of delivering their prized designs to NEC, who at the time was still theoretically going to release their tile based PowerVR2 monster chip.
I actually don't think that chipset ended being used in the DreamCast(I lost track--so sue me), but I definitely remember a pretty massive settlement for breach of contract etc. against Sega for their 3DFX fiasco.
Lots of drama, somebody go post some links.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
I've been saying this for a while: Most actions of Microsoft can be analyzed under a "Chains For Freedom" Philosophy: Microsoft wants to make the general public as free as possible, as long as they're the ones to deliver the freedom.
That in mind, consider the values of the X-Box. Whereas most console makers tightly restrict development on their machines, game developers will likely be free to release whatever they like for the X-Box.
Expect, of course, an extremely quickloading and game-customized version of Windows to have a per unit cost on each Game DVD. Also expect this to be significantly cheaper than Sony or Sega mandates per unit.
Don't expect non-standard media. Microsoft, for all the embracing and extending it's accused of(justifiably), has an excellent hardware department. I've never, ever been more embarassed of Slashdotters as when I watched them rip apart the touch-mouse--this was an impressive technology that would have been worshipped had it come from Logitech and shipped with Linux drivers. Anyway, expect their department to build in an industry standard DVD-ROM, likely one that can play DVD-RAM disks so as to facilitate ease of use for system developers.
Developers become "free" from propietary startup hardware costs, and "shackled" by having to code to Microsoft standards. It's pretty interesting to watch.
Linux is the wildcard. While Sony is releasing Linux development code for PS2, Linux should actually *run* on this K6(!!!) based system. This, above all else, should excite the heck out of us. Here we have a cheap system that will very likely be in millions and millions of homes, have 10/100 Ethernet built in(oops, Sega!), significant processing power($20 says they put the SB Live audio chipset in it, btw), and will run Linux.
Beyond the fact that we will see developers using Linux as their game OS rather than the one-off Windows from MS, here's a console that one can actually pawn computational work on. It's a console free of almost any legal demands and enforcements of its creator.
It's an open platform, from Microsoft.
Recognizing that this is a cool thing gives us more credibility when we accuse MS of the kind of junk described in the Halloween papers.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
You obviously work for RealNetworks or otherwise benefit from their largess, so why don't you stop polluting this thread with your corporate PR - /. is a forum for people, not corporations.
Unfair. Corporations have every right to defend themselves, and there's no reason to believe that A Nonymous Coward is really a RealNetworks employee. (Yes, people can doubt me without having an ulterior motive.)
His point is rational--the claim could be taken to mean that RealNetworks reports all MP3s encoded by them and nothing else. It's plausable, but I'd be qiote pissed at the Times--Number of MP3s Encoded != Number of MP3s on the Hard Drive. (Still, there's a pretty reasonable amount of privacy violation even without the extra-software spying.)
The only way to check is to rip out a copy of FileMon and see what RealNetworks is really up to. If I get some free time, I'll do this myself.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
AC--
(BTW: No reason to be anonymous. I prefer to respond to people, not "entities"--You Are Your Words. Own them.)
Richard Smith, a Brookline, Massachusetts-based independent security consultant, said the numbers of songs stored on a user's hard drive, the kind of file formats in which the songs are stored, the user's preferred genre of music, and the type of portable music player, if any, the user has connected to the computer are sent to the company, the Times said.
This is my evidence(and my first paragraph from the post you responded to). If it's wrong, I self-flagellate myself upon the battered journalistic integrity of the above. RealNetworks didn't particularly refute any of this, and I'm sure they'd be screaming bloody f*cking murder if they were accused of taking one iota of extra data.
AC, I would be laughing myself to tears if this was all about mere listening patterns. That's NOT what the evidence suggests.
Do you have any evidence we don't know about?
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
I have only one question here: Did the company listen to the outrage of thousands of customers over the privacy violation or the 1-7/8 drop in their stock?
And me without my moderator points. Ah well, such is the pain for posting in this discussion.
Excellent observation.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Sorry, Mike. You summarized things a bit too well:
"If I decide to put up mikenash.com and I want to sell T-Shirts with my picture on them, for something uninteresting like me five CALs is all I need since I probably won't have more than five people buying at one time," Nash predicted.
Probably, eh? And what if I do? What if, say, Slashdot links to the T-Shirt site I'm going to open up someday and--amazingly enough--I have some T-Shirt that's surprisingly popular. Far more than its been. Are you telling me that, while Apache-SSL would be more than happy to accept as many credit card orders as the server could possibly handle, Windows 2000 would tell my customers to go away because I didn't give Microsoft enough money?
Are you kidding me?
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Deep within the nestled chroot()'d directories of Rob Malda's Precision Cut Head lies Slashdot...but nobody knew where this ever growing tumor of a side project turned major stock came from from. Nobody knew...until:
And if any of Townshend's PR people stumble on this, I wanna interview Pete for Slashdot.
Slashdot is just a gigantic ploy to let Rob Malda satisfy his lifelong wish to talk to Pete Townshend!
"...and thus it was written in the Book Of Dot, 'O Lord, How May I Speak To Your Musical Prophet', and the Lord replied, 'Yea, ye shall verily speak with thy prophet, but spread far and wide my message of openness and birds that do fail to fly. For that is the way to happiness; the road to bethlehem is with the Geek!"
Yours (Hopefully) Amusingly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
But that's the general rule of consumer computing hardware - don't expect any real quality unless it translates directly into higher Quake benchmarks.
Oh, you'd be surprised how amazingly correct you are. I blame the death of Cyrix on John Carmack and the amazing fact that Quake--the only application to ever demand serious FPU speeds, and the most Pentium-tuned app in history--ran like a dog on anything but Intel. That 3DFX was the only credible 3D company for quite a few years(an amazing amount of time for a peripheral manufacturer in this industry) was also Carmack's doing.
Yes, Carmack is a programming god. Just look up the Law of Unintended Consequences, however.
In contrast, Carmack's utter rejection of the Direct3D Execute Buffer disaster(tuned specifically for MS's Talisman architecture) translated directly into a reasonably predictable redesign of Direct3D.
All parts that are likely to fail in a system should have a standard of reliability they are required to meet in order to reach some government/industry certification. Couple this with an mandate to advertise consumer awareness at the level of "Intel Inside" and you force the market to a minimum but acceptable level of quality.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Richard Smith, a Brookline, Massachusetts-based independent security consultant, said the numbers of songs stored on a user's hard drive, the kind of file formats in which the songs are stored, the user's preferred genre of music, and the type of portable music player, if any, the user has connected to the computer are sent to the company, the Times said.
People, this isn't just RealNetworks incidentally receiving information on what CDs you have by nature of that being the only way to send back the track titles.
RealNetworks invasively scanned millions of American's computers for content that had nothing to do with the functioning behavior of RealNetworks software. We're talking about code that looked for MP3s, music applications, hardware interface tools, and who else knows--I wouldn't look for RealNetworks to tell.
Open Source is many things, but I'd seriously rather it not degrade into the only way to trust that code isn't Trojan'd. I expect that kind of paranoia for my cryptology of choice, not to play some Garbage!
This isn't an issue about a few missing lines from a privacy statement. Should RealNetworks be able to upload any interesting file on your hard drive to the corporate servers as long as they mention that "From time to time, RealNetworks may request feedback from your internal storage systems according to specific parameters to be determined according to your usage profile"? Maybe it'd be fine for them to tap into your computer's microphone, as long as they don't neglect to tack on "User agrees to indemnify RealNetworks from any liability in relation to any data flowing through said user's Sound Card"?
This isn't about legality, at least, not yet. It's about trust, and RealNetworks is losing mine fast.
The real question is, whether TrustE will follow.
I'm no history expert, but there's an aspect of TrustE that just smacks of the ill-fated League of Nations from the first part of the century. Namely, the well-intentioned but utterly toothless, powerless, and secretly mocked nature of it. I think TrustE actually has enough Respect Capital(if there is such a thing) with the press to actually do something, this one time...
Or never again, because nobody will listen anymore.
TrustE needs to set up guidelines of what may be buried in the fine print and what needs explicit and large dialogs before the function is completed--yes, this includes specifications like "Default must be no, and the software must still run even if it isn't allowed to insert seven links to the audio playing software like RealPlayer G2 does--we counted." That's clear, from RealNetwork's rather shocking behavior.
The bottom line is TrustE simply needs to file suit for breach of contract and reach a settlement where RealNetworks needs to contact all possible users, mass deploy a tremendous upgrade, and notify victims of the violations in both online and TV/Magazine forums.
That, or some combination with what I'd like to call TrustEeth: Privacy Protected for x Days.
If you think about it, it's really just a much more positive version of "This Site Accident Free for x Days" signs. The system encourages TrustE certification, since the longer one puts it off, the longer it will take to get to privacy levels respected by customers. It will make it progressively more expensive over time for large companies to allow their ego to overpower the rights of their customers--the CEO will be quite peeved at the middle manager who took the nationwide corporation down to one day of privacy protection.
If not a system using literal days, then an accumulation of points, lowered by violations, maintained by fair and quick resolution of privacy concerns, and accelerated by respectful "voluntary" policies could also be functional.
The key is, people need to have a gauge by which they can determine whether or not to trust a site and the code it asks them to download, and managers need to know they could get called on the carpet if they try a stunt like RealNetworks did.
The irony is truly remarkable, if you ask me. The CEO of RealNetworks(then Progressive Networks, if I remember correctly) went and testified in front of The United States House Of Representatives, arguing against everybody's favorite monopolist, Microsoft, was making the playing field unfair.
Meanwhile, here we are in November of 1999, and RealNetworks is repeating the sin that Microsoft did wayyyy back in the day with its overly nosy Registration Wizard that reported if software like Wordperfect was installed. Incidentally, the above dig at RealPlayer G2 for the seven links it litters all over your desktop(collect them all) is even more beautifully ironic considering the now strangely difficult to find position paper regarding asking the user before doing anything of import.
On a plus note, I don't think the US Patent Office had anything to do with this one.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
DirectSound3D was hacked together by the engineers at Aureal. They did it, they wrote it, MS used it. At least that's how it was explained to me back at WinHEC. (Ironic. The only convention I've ever paid money to attend was WinHEC '97. Definitely was worth it though, if only to see the Rendition guy get disturbingly flustered when I started comparing their product to the Voodoo solutions coming out of 3DFX. Incidentally, this was also the first the world had seen of the Riva 128, A.K.A. the return of nVidia. nVidia was the first 3D card manufacturer for PCs, but their card was so slow that it was better just to operate in software. Telling people nVidia was coming back from the grave was like announcing PC Chips as the one of the higher quality manufacturers in the industry...heh, wait a second...)
As for me, I use a SB Live Value because I compose music and the SB Live's MIDI support is light years beyond anything the rest of the industry has supplied us with. I'm truly looking forward to seeing what audio hackers can do with the power of that DSP.
I've still always lusted after a Vortex 2, though, and might just grab one for the hell of it.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
John Dvorak wrote an article around 1989 lambasting the computer industry for the complete and total lack of quality that 5.25" floppies possessed. They were slow, they were loud, and most disturbingly, they lost data with ridiculous ease. (Somebody find me a link to this article. I need it for my files.)
Ten years later, we're using 3.5" disks, and we're in the same situation.
Steve Jobs and the iMac development team had a chance to fix the situation. They recognized that floppies were garbage. They knew they weren't stable by any means. In response...they chose nothing.
Thanks, Steve.
So, floppy disks persist as the one and only truly standard way for moving documents and small files from one machine to another, and maybe even storing them off a hard drive for some period of time. Most of us who build our own computers know--buy Teac, or you might as well throw away your data. Most of us who build our own computers do so because ten more dollars for a part that can be trusted is worth it for us, but for a large scale company churning out thousands of boxes, ten more dollars per unit is suicide if you can get away with using it to pad the bottom line.
This is one of those "diseconomies of scale" when it comes to large computer companies--saving pennies on a box to remove a feature that everybody expects but only 1% know to check for(like supporting more than 32MB RAM on a certain motherboard, Compaq)--that keep the little guys in business.
Anyway, I must say I'm not entirely displeased with the tremendous pain about to be inflicted on those who have been knowingly distributing fatally flawed 3.5" disk technology in an effort to save money. I spent two years as a volunteer tech at my university--the amount of raw labor I saw go down the drain because of floppies gone terribly wrong was shocking, as was the amount of disks that simply couldn't handle a Linux boot disk. Imagining large corporations going through the same kind of pain I've watched a relatively small population of students go through is frightening, to say the least.
Of course, there will be some companies defending their negligence by saying the minimal quality standards were "common industry practice". So too were the falsified 15"/17" monitor statistics that lead to the "Viewable Size in Every Advertisement" agreement. Lets not even begin to mention the coming slap down of the entire 56K modem industry, which was all too happy to claim speeds twice as fast when even in the most generous contexts it wasn't the case. And, of course, sooner or later, Inkjet Printer manufacturers will get their due--8 pages a minute? Yeah, if you're printing a period.
But what's critical is that while faking monitor sizes only strained a few eyeballs, and slow modems and printers maybe caused a missed deadline or two, substandard(hell, plain old standard) floppy drives caused data loss that directly led to wasted employee hours and lost property.
I can't have too much mercy here. I've done the recoveries, I've gotten the pleading phone calls, I've thanked Word's Autorecovery feature innumerable times(and cursed Windows' awkwardness at saving to hard drive and making only a copy for the floppy).
It may have taken a very long while for the industry to be taken to task on this, but hopefully we'll finally see a stable portable media standard arise from the legal ashes.
About time.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Aureal has always been one of my favortie companies, and their recent announcement is quite intriguing.
Much of the industry is obsessed with "meeting the marks"--x amount of video RAM, AGP/PCI/VLB(way back in the day), Direct3D, OpenGL, etc.
Essentially, the complexities of various products are simplified down to a set of necessary functions supported and some "speed scale" created by a semiobjective source.
Creative Labs' announcement signified not only a new level of support for Open Source driver development models, but also a relatively drastic turnaround in company policy. It wasn't that long ago that Creative was flatly refusing to open their drivers to anyone but OSS Sound Inc. under NDA. Now they're saying they'll support any and all comers, and even overcome the starting barriers(CVS/Bugzilla) so that development can begin ASAP.
Such full fledged support is reserved only for items considered integral for the success of a product--in other words, a line item, a necessary feature.
A checkmark.
That Aureal(whose stock symbol always seems to confuse the hell out of every engine I check it with) is following suit means that, for the first time, not only Linux support but full, open access to driver development infrastructure is becoming a line item checkmark for a segment of the PC Hardware Expansion industry.
Such developments bode well for future developers, who will hopefully not need to painfully reverse engineer nearly as much in their attempts to get network cards, 3D Graphics boards, or any other custom hardware to function now and long after a company decides to cease support for a given product.
When ZD Net reports closed drivers as a downside to a given piece of hardware, we'll have truly won.
However, it should be noted that while the SB Live is a full DSP architecture that developers should be able to exploit to unimaginable degrees(though I'm not expecting a Perl RegEx module using the SB Live *LOL*), much more content is hardcoded within the Aureal chipset. Indeed, this is a risk of a open development process--companies may feel that, to protect their intellectual property, they need to lock it up within the chipset instead of releasing it in the drivers.
Of course, a chipset and a card that remains viable for significantly longer periods of time is far more appealing to both OEMs and consumers, and this is a benefit that both programmable circuitry and open development foster.
This isn't to say, though, that I'm not extremely excited about Aureal's announcement. I first heard Aureal's A3D at WinHEC '97 a few years back, and it was the first time I had truly heard 3D Sound that actually worked. Aureal is one of the great stock price tragedies of the last few years, mainly because of some rather nasty innuendo and patent litigation. (Yes Virginia, that ugly USPTO spectre pops up again.) Their technology is sound(no pun intended), and their cards are uniformly high quality. I'm quite interested at seeing what the Linux community will do with access to the extensive 3D Sound modeling that Aureal has implemented--literal 3D environment models are used to determine reflections, reverbs, etc. It's all quite amazing, and very, very well implemented.
It's definitely an interesting time to be a sound engineer in the computer industry. One thought--BeOS ought to be supporting some new very powerful sound processing cards very soon, if it doesn't support them already.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
A bit more thought on this: :-). They're actually going the extra mile and providing not only the source but a development environment for coders to come, watch, and learn.
Wow, creative is setting up CVS/Bugzilla. They're not merely opening the source; they're not just trying to grasp a bit of extra PR out of the Linux mindshare gods(Taco and Hemos
This is amazing, and deserves a retrospective profile in around six months to see how this great, precedent setting experiment panned out.
Of course, Creative isn't dumb. As I mentioned in another post, Creative stands to have their card become the standard DSP component in innumerable Linux machines--their foresight in developing a programmable sound card is very likely to pay off handsomely in increased sales.
The economics of Open Source just got much more interesting.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
First of all, there's quite a bit of difference between fastest and fastest known. I can't imagine both the chinese AND the american governments from having some exceedingly classified hardware that blows the pants off the open stuff(read: governmental phallus-phlashing.)
Second, the meaning of fastest is very unclear. I'd go so far to say that any system that implements a given function in software instead of hardware is going to be orders of magnitude slower than the state of the art. Witness the EFF DES cracking machine, 3D Graphics Accelerators, even Math Coprocessors. Fitting a square peg into a round hole is actually a pretty common occurance in the computer world, but it takes a relatively tortoise-like rate compared to what can be pulled off with raw gates.
That's why XISC--Extensible Instruction Set Computing--is probably the upcoming processor paradigm. Programmers need the ability to redefine round holes into square ones, so the square pegs fit right in.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Alot of people aren't going to get the significance of this announcement--they'll think, "Cool, another sound card that I can play with on Linux w/o resorting to closed source drivers."
This is beyond that.
The SB Live is based on a single, ridiculously powerful and extremely programmable DSP. Almost all functions the chipset performs are executed in DSP level software, meaning suddenly Linux is getting full specs on a complete digital processing environment.
The impacts of this are substantial. The SB Live chipset is cheap enough that it's a contender for "standard sound" in many machines. Open source algorithms for everything from MP3 encoding to analog synth simulation to the more esoteric, non-sound related stuff(GIMP graphical filters, datastream analysis, etc.) should, if the drivers are clean enough, start popping up over time.
The uses of such a powerful digital signal processor on an open platform are honestly unpredictable at this point in time. While there are hardcoded design issues in the SB Live chipset(most notably, all signals are upsampled to 48khz before processing may occur), the sheer flexibility of this chipset will blow Linux programmers out of the water.
This is truly excellent news, and shouldn't be ignored as a mere fun thing for the gamers to play with. If only 3D graphics hardware was as programmable...or as open.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
Duh! that 115k2 *is* the serial port speed, my dear Evil friend... don't trust what Windows says.
Nah, don't trust how 3Com manipulated the Windows drivers. There's a difference.
Most 56K modems actually report the real connection speed. 3Com is doing a bit of phallus-phlashing, if you know what I mean.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
That's like saying that cars these days have gotten quite good at protecting their drivers from fatal crashes. The statement may be true, but that sort of thing should still NEVER happen. It is to be avoided at almost all costs.
No. You don't understand.
If you slam the power button on a FAT/FAT32 box, you're not gonna lose the partition.
You can't say the same for a Linux box using ext2, or even a Solaris box using UFS. From *VERY PAINFUL PERSONAL EXPERIENCE*, you have quite a decent chance of damaging some serious stuff, and way more than an unheard of possibility of just completely losing the filesystem.
FAT/FAT32 can recover from random reboots without a problem. It's simple enough to just not have the same kind of problems as Linux w/ ext2.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
When was the last time you sat through a Linux boot and upon execution of "scandisk" were required to hit "Fix" fourty-one-thousand and three times because the UI designers decided it would be too hard to add a "Fix All" button? e2fsck -p, my friend.
Be fair. (It's good for credibility.) Any time there's a semi-serious problem, you're gonna be hitting y for quite a while w/ fsck.
I've lost entire file systems more than a few times because of an unscheduled reboot, incidentally. The same has happened, incidentally, w/ NTFS, but never, ever, ever with FAT/FAT32.
It's actually enough that there's a semi-decent chance I'll make my MP3 partition a Fat32 one.
Yours Truly,
Dan Kaminsky
DoxPara Research
http://www.doxpara.com
P.S. Don't tell me Solaris is any better; it made some noises significantly scarier than "extra bytes discovered" when I recently bungled a shutdown.