Sun's marketshare.
on
Sun's Last Stand
·
· Score: 2, Interesting
In 2002, everyone lost sales, except for Dell's dekstops. Sun lost fewer sales than IBM and HP and everyone else in the industry, so it picked up marketshare, a critical yardstick of competitive performance in down markets. So, while Sun is fuxxored, it is less fuxxored than IBM, HP, SGI, Unisys, etc.
The piece was a hatchetjob that displayed very little deep understanding of the IT market.
Sun's throughput has always been phenomenal, and it's mostly due to hardware... the glue logic's on the processor itself, so it doesn't need to hit a chip outside the processor to do I/O and memory access. Solaris is written with a complete understanding of the hardware, so it's latency is unbelieveably low on Sun hardware, where it's raw number crunching ability lags. In a server environment, it's a good trade-off.
Japanese steel was of poor quality, which is why it needed to be folded so many times when swordmaking... the excessive forging worked out some of the impurities. European steel, apart and aside from the "fabled" Toledo and Damascus steel, was of very high quality. Modern tool steels are orders of magnitude stronger than any of the above, and generally what the modern SCA armorers use.
If anything, the modern recreations are more potent weapons than the originals... this is due to better material available to metalcrafters.
This is an extremely good idea, but it needs to leverage device integration.
The "personal server" would, essentially, allow you to take all of your personal data with you wherever you went... no matter what computer you sat down in front of, you would have access to your files and even the applications needed to access and edit them. Imagine an iPod with 802.11 and Appleshare (or NFS or Samba).
The downside is that away from a computer, you would have no way to access or modify those files. So you'd need a PDA and a personal server, anyway.
The logical extension is to combine the personal server with a PDA. Hell, might as well build the whole damn thing into a cell phone... an iPod sized cellphone with 5 gigs of data for MP3s, client databases, config files, what have you.
Add bluetooth to this thing, and you wouldn't even need a PC, just a screen and keyboard that will talk to the personal server/pda/phone. Instant desktop, without taking the computer from your geeky belt clip. Perfect for businesses, but not so hot for home gaming... hence the server. Wander over to a real PC, and you're still in business, with the same set of user information.
Base the thing around open standards, and you should be able to wander from a Mac to a PC to a Sun workstation and back again, with the same set of files...
Dance Dance Revolution is a wildly popular video-game, even with demographics that normally wouldn't be caught dead in an arcade (ie: girls.)
One of the keys to its success is user-prompting... an interface where visual and audio cues (the arrows and the music) indicate what the user should do next to achieve a goal without rote memorization or exhaustive trial and error. This is something that will revolutionize UI design once it is better understood... a graphics program walking you through the steps needed to resize a photo without leaving the programs interface to refer to a help file... or an IDE that can ease newbies into coding.
Speaking of newbie coders, Squeak is a very interesting and active project that combines an "educational" IDE with a completely portable run-time environment that also has its own user interface... and everything in it is an object than can be modified on the fly. Extraordinarily powerful while being simple to learn. You can and do jump right into GUI, graphics and network programming almost from the beginning, from a compltely OO standpoint. It's interface needs a ton of work, tho... Smalltalk 80 was ugly and awkward to start with, and it has not improved with time. It also needs more interesting high level objects in its arsenal like those available for Java.
Cox has some really sharp data engineers (I know a few of them), and the resi accounts block all kinds of ports and traffic shaping systems keep an eye out for server behavior, so I'm stuck. If I want to spend $110/month, I can get a 256k symetrical... yay.
I've got Cox, and to get a hard IP I can run a server on, I have to shell out $69/month for a 128k symetrical link. (It's 3mbps down and 256k up at $40 for a DHCP account.) Not too speedy compared to what Speakeasy's shilling for the same price... but Verizon has not upgraded the local switch to DSL-capable, and probably never will.
While we're on the topic of Speakeasy, I once had a 780k symmetrical link from them for $80/month. They no longer offer anything anywhere near that speed/price ratio... they've taken a huge step back. Yet, as has been noted elsewhere, DSL in Japan is dirt cheap for a pipe that can saturate a 10b-T link on the downstream.
This is what deregulation gets you.
The "game over" is in the next generation wireless. Goodbye, cable! Goodbye, POTS! Sprint already offers 155k symetrical links for $50/month... uncapped and unmetered. All you can eat. The other big wireless vendors will either quickly follow suit, or get eaten alive by Sprint. If they can even get a reliable link at 1.5mbps at $50/month, they'll steal huge share away from DSL and Cable at twice that speed. Everyone with a notebook will want it. Unlike on copper, wireless RF sees no cost benefit from throttling upload, so I can see hard IPs and servers being the deal maker for premium "geek packages." Then the other broadband vendors will either shape up, or get run out of town on a rail.
MacOS X is BSD based, without a lick of Sys V in it. AIX is it's own thing, derived from the old OSF standard that was never adopted. (Not to be confused with the OSF pushed by DEC, which became Digital Unix, which became True64, which is BSD-based.)
Palm and Symbian are most certainly not System V based. Nor is VMS or OS/400. QNX, Plan 9, OS/9, RTOS are not Sys V based. The Amiga OS and its derivatives, Pheonix and Morph/OS are not Sys V based. Atheos is not Sys V based.
NetBSD has picked up some Sys V flavor in the past couple of years, but it's still, well, BSD. The other BSDs are definitely not Sys V, and free and clear of it with BSD 4.4 Lite based implementations.
Sys V based:
SCO (Unixware and SCO Unix) Solaris (but with BSD flavor) IRIX (Ditto.) HP-UX (But with a Mach microkernel.) Whatever name Fujitsu-Siemens is shilling Pyramid Unix as these days. UNICOS Non-Stop Unix Doesn't Unisys still flog their own Unix?
You get the hint. Apart from Solaris and HP-UX, Sys V is only seen in ultra-tiny market segements boasting big server hardware.
Shock and outrage! Theo opens his mouth to bite the hand that feeds him, and so gets no bone? Who would have thought it would happen to such a sweet and affable fellow?
Bah.
Theo's legendary lack of tact and people-skills has sunk him... again. He can fork NetBSD and come out on top, he can fork OpenSSH and win the trademark dispute, he can fork IPfilter after alienating Darren Reed... I don't think he can fork the US Government. (Tho it would be a lot more stable and secure if he did... )
What illicit technology offers the most fun and challenge today... where are the new frontiers for today's hackers to push the bleeding edge, and what interesting directions do you see them taking with it?
We do this in the US, too. If you haven't been since highschool, take a visit to your local public library: even in small towns there are often a half dozen PCs or Macs set up for the express purpose of browsing the 'net. You'd be shocked at how busy they are. Way back in the '70s, there were "public terminals" set up in places like Berkely... kiosks where people off the street could get and recieve email and participate on messageboards. This really helped to prime the pump for the personal computer revolution, and the BBS and Internet revolutions that followed. No reason why it can't work in Africa...
Obsolte PCs make great web browsing stations with the addition of Linux and the free web browser of your choice... set up a network of three or four of them in the town school/library/bar/whatever, and you're all set.
It occurs to me that a standardized "Library Terminal" distro, designed for robustness and single-click installation off of one CD, would be a great idea...
Verifying reverse-DNS isn't a terribly good idea... you're blocking mail comming from sites that do virtual hosting. In the olden days of one-to-one mapping of hostname and IP, it was a smart move. Now, when a single box can host hundred domains or more, filtering out connections from domains without reverse-DNS is going to cause more problems than it solves.
A Meta-Study is the most inbred and brain-damaged excuse for science I've seen in a long, long time.
OK, my own interest in low-carb diets is very personal and direct: I've lost more than 70 lbs on one. Not exactly a scientific reason to believe weight loss through the restriction of carbohydrates works, but a compelling one all the same.
So! From my perspective, a more usefull study would be clinical, and ask the following: Is my experience typical? Will I keep this weight off? What percentage of people following carb-restricted diets loose weight and keep it off?
Because low-calorie or not, everyone, and I mean everyone, I know who follows the Atkins diet looses weight (so long as they don't cheat... and resisting temptation is usually pretty easy.)
I'd like some clinical studies to back up this informal observation with hard statistics. Then we can break out the lood samples to figure out the biochemistry.
Dredging throug unrelated past studies to re-interpret the data in a light friendly to "common knowlege" doesn't seem much like science to me. Especially since I'm down four pant sizes. I will admit that I'm eating less... but it's not a conscious change. Unlike regular calorie restricted diets, my body doesn't rebel, and I'm able to stick with the diet... I'm eating less because my body is telling me to eat less. Again, this is an informal observation, but a scientific study to determine if this is indeed what's happening, and how it happens, would be more useful than the Stanford "meta-study."
The author's claim that Java is a dead-end language like COBOL is patently ludicrous, and by dismissing Java out of hand, he undermines the rest of his article.
Java is essentially a re-implementation of C; it's a very C-like language in syntax and semantics. The goal was to do an OO language based on a familiar paradigm. In addition to eliminating some of C's less endearing traits, it brings two things to the table, both of which which will shape future languages:
1. Re-usability. Sun offers you a crapload of very usefull re-usable objects in their JDKs, and people like the Apache project offers you even more. The ability to do insanely complex prrojects with tiny amounts of effort is one reason why Java rules the corporate enterprise. Future languages will start looking more and more like a big box of legos: find the parts you need and plug them together.
2) A universal computing environment: you can't write to the metal in Java, but it's not as slow as interpreted languages like Python and TCL for gigantic computing tasks. Any project, no matter how monolithic and task-optimized, is as portable as the VM is. Anyone who's had to manage a platform migration for key buisiness applications, from VAX to Solaris, say, or worse, from S/360 to Windows, knows the pain of re-implementation. That pain is gone when you use a VM-based language like Java.
Projects like Squeak are looking more and more like Java these days, in terms of re-usability and VM-based platform independance. The only missing piece of the puzzle are popular VM environments not tied to any one vendor: Java needs to cut its dependancy on Sun JDKs, or it will be supplanted by another language that is independant and standards based.
This isn't to say that other languages aren't going to evolve, too, or are useless because they're not like Java. Ayone who programs in the new interpreted scripting languages: PHP, Python, Perl, Ruby, TCL, Scheme, etc, etc, can attest to the power of the that approach to modern computing.
On the other hand, I really don't see any new compiled-to-the-metal languages emerging. Fortran is used for high-performance computing, Forth is used for tiny computers, and C/C++ is used for system programming. It will very likely be the same way in another 20 years, or another 50. The difference is that applications will slowly drift to either VM languages or interpreted languages from binaries compiled from source.
MacOS X Jaguar install disks MacOS 9.2 install disk. MacOS 8.1 install disk (for Oooold hardware) System 7.1 boot floppy (for really, really old hardware.)
Software Utilities:
Alsoft DiskWarrior Norton Utilities for Mac Norton Antivirus Retrospect
Hardware:
Apple-branded Firewire emergency backup and restore device (Or as we like to call it when not adding it to our equipment req: the iPod.) Go for the big one, you'll be glad you had it when you need to rescue your data from a flaky powerbook. Use Retrospect to make sure you get everything backed up proper.
Firewire to SCSI adapter (for getting data to and from older Macs.)
If you're going to dealing with real old Macs: AAUI dongle, phone-net adapters, Mini DIN 8 to DB24 and DB9 serial cables.
Unix Survival Kit:
Hardware:
Powerbook or iBook, with aforementioned Firewire SCSI adapter, USB serial adapter and a nice terminal emu program.
SCSI external HDD
Bunch 'o SCSI adapters/cables/testers
A SCSI CDROM... if you deal with Sun equipment, make sure it's able to boot a SPARC box.
Software:
Install CDs for your Unix flavors of choice. CDs with the most current OS patch levels on it, one per OS. Another CD with your customized dot files, shell scripts and all the useful stuff you really wished came with your vendor's Unix, but didn't (GNU).
NetBSD install CDs "for when all else fails." Comes in handy when you need to repurpose an old Motorola VME system previously installed with telco switching software to interface with lab monitoring hardware with shell scripts, a serial port and a prayer.
Documentation:
Unix System Administrator's Handbook on CD. A copy of the "Fixing Solaris" howto in.TXT.
Linux Kit:
Mac Powerbook or iBook, haughty sneer.
Software:
Slackware to inspire feeling of inadequacy in self proclaimed Linux gurus. Gentoo Level 3 on a USB keychain drive... especially usefull if you're stuck with a 2400 baud modem in a jungle where the pbone only works for three hours alternating tuesdays. (Ah, sarcasm!)
Copy of NetBSD or OpenBSD install disks to get real work done.
The most hated company in all of Macdom is not the beast from Redmond, who makes the tasty, lickable Office, but Quark. User-hostile doesn't even begin to cover its marketing and support... user belligerent is more like it. They flat-out refused to port to OSX (they still haven't), and they openly despise the Macintosh platform and insult its adherents at trade shows.
They are this way because they believed they had an unbeatable product, a single killer app the world could not do without: Xpress. The Mac dweebs would buy and keep buying, because there was no credible choice.
Until Adobe came up with InDesign, which is easier, faster, every inch as powerful, compatible with Xpress "Xtensions" and runs on OS X. Adobe shows their users lots of lovin', with trade shows, rational support, and deep Mac roots. Now InDesign is poised to topple Xpress into irrelevancy.
Adobe does not have the only pro-caliber image editing app out there. If they're upset that iPhoto killed ImageReady, and incensed that FinalCut destroyed Premier, wait until Apple decides to buy the TIFF-any codebase, or Avisa Image, or just roll their own Photoshop killer based on the GIMP.
Adobe is playing a very dangerous game. If Quark can be dethroned, you better damn well believe Photoshop can be, too. Apple's got pockets deep enough to do it, and marketing savvy that put FinalCut Pro on a Powerbook in the news vans of every TV station in the civilized world.
64-bitness won't increase speed. It will increase the amount of RAM you can throw at a problem. 48bit drum scans eat lots and lots of RAM, and of you're working with multiple images, you can bump up against Apple's current 1.5GB limit, and then virtual memory kicks in and performance goes to hell.
64bits on the desktop will be a great and good thing for all sorts of content creation chores.
Yup. UC Berkely, BSDi and BSD/386, the progenitor of the three *BSDs, Open, Net and Free, already had that lawsuit. AT&T wrung 'em through the wringer, UC Berkely slapped back with a copyright countersuit, and after much legal arangling, BSD 4.4 Lite came to be. It was free of patent and IP trouble, and then went on to become the backbone of BSDi and FreeBSD, which begat NetBSD, which begat OpenBSD. MacOS X was based on NeXT, which was a strange interpretation of BSD before 4.4 Lite. It's unixy bits are now based around current forks of FreeBSD code, IIRC, but it's no less strange.
Unix is very, very well understood, as the source code was freely distributed and then forked to hell and gone. Since the early '90s, the only thing all the Unix vendors really needed from Unix was the name, which is kindly provided by the Open Group these days so long as you have the cash.
AIX is not an SVR4-based Unix. Anyone who's taken a look at the psychotic commands required to admin an AIX box clearly understands this, as does anyone who's ever ported SVR4 based software to it.
Solaris and Irix are a little closer to their SVR4 roots, but True64 is BSD based. HP-UX is based around a Mach microkernel. MacOS X is based around both Mach and BSD, doesn't use any of the traditional file system or configuration tools, and it's still not as weird as AIX. So it can be easily argued that IBM did, does and will continue to do it's onw weird, scary thing without the need for anybody's code, patents or input. Exhibits A, B and C: zOS, OS/400 and OS/2. IBM has OS style kung-fu like no other company in the industry.
The offspring of 386BSD; Free, Open and Net are cleared of patent problems after a legal tussle with AT&T. Yahoo's been a FreeBSD shop from day one, so it's an enterprise class OS that got there without the deep pockets of any Computer Giant. There goes the foundation of the lawsuit.
What's worse, if you have to fear just one legal department, fear IBM's. They have a portfolio of defensive patents that will bury SCO so deep in countersuits, they'll never see the light of day.
So SCO is going to crash and burn and die for being stupid.
SoupIsGood Food
Re:Worthless market analysis
on
The Faded Sun
·
· Score: 1
No counter argument, except it;s the same song that's been sung since the '70s. Personal comnputers were going to overthrow the old order, and relegate the mainframe to the dustbin of history! Didn't happen. From the early Apple II with Visicalc, to the "Standard" CP/M machines to the IBM PC, to the (IBM) Clone Wars, to the Mac and the 32bit broughaha, to NT, each new advance has been heralded as the ultimate commoditization of enterprise computing.
Bullshit. Enterprise desktops can't even be commoditized... when's the last time you saw a Fortune 500 company rolling out the generic white boxes to the desktop? Never. They all use top-tier desktop hardware vendors at three to five times markup over generic (or more). Meanwhile, VMS, AS/400 and HP3000 boxes still roll off the assmebly lines and into datacenters worldwide, despite their prospective vendors' best efforts to kill them as "obsolete"... there are many times when one big box that just works is more cost-competitive than lots of smaller, cheaper boxes.
Some applications can be commoditized... Google and Yahoo's server farms come to mind. Even so, Sun and IBM counter by putting the paralellization into software in their mainframe class systems, where the TCO and other spooooky financial metrics favor a few big, expensive boxes with solid service contracts and SLAs over bazillions of teensy ones.
The personal computer industry's been threatening to bury the big enterprise systems with an avalanche of commodified systems for a long, long time... and while commodity hardware has indeed found a permanent place in the datacenter, it's still a long, long, long ways off from supplanting everything else in there.
And Apple is Apple. So long as they're not bankrupt and being dismembered for liquidation, they live on, usually pulling in a nice, fat profit margin. (Didn't do so hot last year... but nobody except Dell was. Apple has 4 Billion bucks in cash, so a coupla million here or there means squat/all in the long range.)
SoupIsGood Food
Sunset a long, long ways off.
on
The Faded Sun
·
· Score: 4, Interesting
Utter horseshit. Everyone lost sales last year: IBM, HP/DECompaq, SGI, Fujitsu-Siemens, Bull, NEC... everyone. Sun lost fewer sales than the other major players, so they picked up marketshare. They hemmorhaged money because of spotty buisiness practices from the dot-com era comming up to catch them, but as Cringley says, they've got another five years to get that sorted out.
Sun's transformation from a king of the workstation vendor into a server powerhouse that only IBM has any real hope of competing with is nothing short of phenomenal. If it looks like Linux is going to kill the proprietary Unix market, then Sun will go Linux... they've made similar moves in the past. Sun switched their entire installed base from the BSD-derived SunOS to the SVR4 flavored Solaris, no small feat at the time. A switch to Linux will be a snap, and yes, Sun will still charge you as much for your Sun-branded Linux as they do for Solaris, and get it from satisfied CIOs, along with fat service contracts.
Sun is never the first to market. Sun is never the ideal solution. Sun never offers the highest performance. Sun is never the cheapest option. Sun doesn't offer the best service in the industry. But they come "close enough" on so many fronts, they're an unbeatable market force. Add in Java, which rules enterprise computing like no technology since COBOL, and Sun ain't going nowhere.
Apple bled billions in the '90s, but they rebounded. McNealy's at least as smart as Jobs, and his marketing instincts are almost as honed. Sun has replaced IBM, and even Windows, in the hearts and minds of every serious CIO and VAR. Give it a year for either the economy to have rebounded, or for Sun to have staunched the bleeding on its own with austerity measures and something innovative. This is the company who managed to launch a line of workstations at the height of the NT onslaught in the Workstation market, and managed to make a mint with them. (The Ultra5 and Ultra10.) They aren't out of tricks yet.
Tho it would be nice if they put the screws to Fujitsu-Siemens to get access to their SPARC design... call it "SuperhyperultraSPARC" or "BadAssSPARC" or "TotallyAwesomeSPARC" or somesuch, and use it to hold the Itanium/POWER dogs at bay while they ready the UltraV.
I'm a Gen-X Slacker, but I forgot I was one for almost 6 years. I worked my ass off for this company or that, trusting their management until I got my ass canned, when I'd get another gig paying twice as much. Well, last year the gigs ran out. If you're a Unix admin, you better damn well have at least a Masters in a technology field, 'cuz nobody's gonna look twice at you otherwise. Recruiters tell me there are 150-200 resumes submitted for every open position, and everyday a new tech company folds, or goes through a round of layoffs. While I'm in the top 10% based on my skillset and experience that means there's 15 to 20 people who are in line ahead of me. Some schmoe who spent the last six years figuring out Tetris was N-P for a sheepskin is gonna get the nod because all I ever did was flunk out of art school.
In short, I trusted that someone knew what they were doing in the big corner office, and now I'm fucked. And it's all my own damn fault, because I forgot the work ethics that brought me into computers and technology in the first place. Time for some of that olde timey religion:
1) SLACK! The world does owe you a living.
Screw 80+ hour workweeks for fat checks. I want 10 hour workweeks for fat checks, and a reasonable assumption that the fat checks will keep rolling in with minimal effort. I don't mind hellacious effort up front... being a lazy sysadmin, I know ten hours of scripting and testing can lead to a hundred hours or more of prime goof-off time. (Or, as the case was, time to write more scripts and to string more cable for my corporate overlords.)
So, by putting in a lot of work that's actually just fucking around with the computer up front, I can spend more time reading comics, posting trolls to/., and drinking. So! Cram a little code, email the hell out of reviewers and tech news blogs, set up a website with credit card ordering information, et voila! Software company. There's probably an open source way to scam money from people with big iron and deep pockets, but I'm too lazy to think of it. Go think of it yourself.
2) If it's worth doing, it's worth DIY.
The suit-wearing weasels sold me out. I'll never trust another. "Professional Management" means "porfessional backstabber." Screw 'em. I will learn the fundamentals of business management from one of the many excellent books at my local library, used book store or web bookseller.
I will learn how to market what I make, and how to balance the books. I will learn how to grow a company. I will never hire employees, but I will pay co-conspirators, and I will figure out a way to make this legal. I will figure out how to run a health insurance/HMO co-op with local small businesses. I will do all this with the meager funds from my teensy, just-above-minimum-wage non-computer job, and I will make any interested Venture Capitalists drink a bottle of robitussin, and I will laugh at them as they hallucnate and tell them to get the hell out. MY company, damnit!
Because if it's worth doing, it's worth doing my own damn self.
But first! I will learn Java and low-level C programming, for the things I am now interested in require that knowlege. I'll fake the rest as I need to. In a year, I'll come back to let you know how well I've succeeded. (I can't fail, as I'm already at baseline failure state right now. Any change is an improvement.)
Slack!
SoupIsGood Food
Re:No, Apple should continue to heed Intel
on
PowerPC Goes 64 bit
·
· Score: 5, Insightful
The battle for the 32bit desktop processor has been won, you mean. It's been a long, bloody battle, and for a time in the mid-'90s, when Apple was on the ropes, the victory looked complete. Intel and AMD have finally emerged victorious, laying claim to the most powerful 32bit processors on the planet.
But, to be frank, Itanium sucks... even Linus thinks so. Nobody wants to use it, Dell, SGI, even HP is still developing PA-RISC silicon, and is incredibly hesitant to commit to the "next generation" IA-64 chip it designed partly in-house. Yamhill is a nice idea, but Intel has no plans to go that route yet, and what's more, denies it's even considering them.
AMD's 64bit offering are, as yet, vapor... and unlikely to pack the punch of the Power4, nevermind a dual-core Power4 with Alti-Vec.
Meanwhile, PowerPC's been 64bit since '96.
Indeed, the PC will continue to kick Apple's butt in 32bit systems, except in notebook applications, which is the only place Apple will keep using 32bit PowerPC processors. D'oh.
So, yes, x86 is irrelevant and outclassed by PowerPC, Itanium is a floundering wreck, leaving Hammer to look very lonely and small up there all buy itself, shoulder to shoulder with UltraSPARCs, R2400s and Power4s. Economy of scale? What scale? When it comes to 64bit hardware, RISC/Unix =is= the scale.
Right now, the prices for digital systems are stratospheric, but the advantages could, in theory, more than make up for the cost. You can use it to hold telecasts: bring lectures in from Oxford, live bigscreen showings of away games for the University's football team. The cost of renting movies drops through the floor as the cost of distribution approaches nil.
The problem is, there are no clear standards yet, and a whole lot of competing ones: Boeing, Technicolor, Sony, DLP, etc. You choose one option, there is =zero= guarantee it will use the digital projection standards for distribution and format that the rest of the industry winds up settling on. Then you have a $150,000 betamax VCR, and the professional equivalent to the dwindling "BETA" section at the local Video Store. No Oxford lectures. No away games. Not even Spiderman II.
I'd give the industry a few years to decide which way it wants to jump, or, barring that, a stone-solid contract from whichever vendor you go with that they will provide you with an upgrade to the equipment to make it compatible with dominant standard in case you pick a looser. Of course, if a proprietary standard is settled upon, you're screwed anyway.
Slashdot Mirror
In 2002, everyone lost sales, except for Dell's dekstops. Sun lost fewer sales than IBM and HP and everyone else in the industry, so it picked up marketshare, a critical yardstick of competitive performance in down markets. So, while Sun is fuxxored, it is less fuxxored than IBM, HP, SGI, Unisys, etc.
The piece was a hatchetjob that displayed very little deep understanding of the IT market.
SoupIsGood Food
Sun's throughput has always been phenomenal, and it's mostly due to hardware... the glue logic's on the processor itself, so it doesn't need to hit a chip outside the processor to do I/O and memory access. Solaris is written with a complete understanding of the hardware, so it's latency is unbelieveably low on Sun hardware, where it's raw number crunching ability lags. In a server environment, it's a good trade-off.
SoupIsGood Food
Japanese steel was of poor quality, which is why it needed to be folded so many times when swordmaking... the excessive forging worked out some of the impurities. European steel, apart and aside from the "fabled" Toledo and Damascus steel, was of very high quality. Modern tool steels are orders of magnitude stronger than any of the above, and generally what the modern SCA armorers use.
If anything, the modern recreations are more potent weapons than the originals... this is due to better material available to metalcrafters.
SoupIsGood Food
This is an extremely good idea, but it needs to leverage device integration.
The "personal server" would, essentially, allow you to take all of your personal data with you wherever you went... no matter what computer you sat down in front of, you would have access to your files and even the applications needed to access and edit them. Imagine an iPod with 802.11 and Appleshare (or NFS or Samba).
The downside is that away from a computer, you would have no way to access or modify those files. So you'd need a PDA and a personal server, anyway.
The logical extension is to combine the personal server with a PDA. Hell, might as well build the whole damn thing into a cell phone... an iPod sized cellphone with 5 gigs of data for MP3s, client databases, config files, what have you.
Add bluetooth to this thing, and you wouldn't even need a PC, just a screen and keyboard that will talk to the personal server/pda/phone. Instant desktop, without taking the computer from your geeky belt clip. Perfect for businesses, but not so hot for home gaming... hence the server. Wander over to a real PC, and you're still in business, with the same set of user information.
Base the thing around open standards, and you should be able to wander from a Mac to a PC to a Sun workstation and back again, with the same set of files...
It's a really neat idea.
SoupIsGood Food
Dance Dance Revolution is a wildly popular video-game, even with demographics that normally wouldn't be caught dead in an arcade (ie: girls.)
One of the keys to its success is user-prompting... an interface where visual and audio cues (the arrows and the music) indicate what the user should do next to achieve a goal without rote memorization or exhaustive trial and error. This is something that will revolutionize UI design once it is better understood... a graphics program walking you through the steps needed to resize a photo without leaving the programs interface to refer to a help file... or an IDE that can ease newbies into coding.
Speaking of newbie coders, Squeak is a very interesting and active project that combines an "educational" IDE with a completely portable run-time environment that also has its own user interface... and everything in it is an object than can be modified on the fly. Extraordinarily powerful while being simple to learn. You can and do jump right into GUI, graphics and network programming almost from the beginning, from a compltely OO standpoint. It's interface needs a ton of work, tho... Smalltalk 80 was ugly and awkward to start with, and it has not improved with time. It also needs more interesting high level objects in its arsenal like those available for Java.
SoupIsGood Food
Cox has some really sharp data engineers (I know a few of them), and the resi accounts block all kinds of ports and traffic shaping systems keep an eye out for server behavior, so I'm stuck. If I want to spend $110/month, I can get a 256k symetrical... yay.
SoupIsGood Food
I've got Cox, and to get a hard IP I can run a server on, I have to shell out $69/month for a 128k symetrical link. (It's 3mbps down and 256k up at $40 for a DHCP account.) Not too speedy compared to what Speakeasy's shilling for the same price... but Verizon has not upgraded the local switch to DSL-capable, and probably never will.
While we're on the topic of Speakeasy, I once had a 780k symmetrical link from them for $80/month. They no longer offer anything anywhere near that speed/price ratio... they've taken a huge step back. Yet, as has been noted elsewhere, DSL in Japan is dirt cheap for a pipe that can saturate a 10b-T link on the downstream.
This is what deregulation gets you.
The "game over" is in the next generation wireless. Goodbye, cable! Goodbye, POTS! Sprint already offers 155k symetrical links for $50/month... uncapped and unmetered. All you can eat. The other big wireless vendors will either quickly follow suit, or get eaten alive by Sprint. If they can even get a reliable link at 1.5mbps at $50/month, they'll steal huge share away from DSL and Cable at twice that speed. Everyone with a notebook will want it. Unlike on copper, wireless RF sees no cost benefit from throttling upload, so I can see hard IPs and servers being the deal maker for premium "geek packages." Then the other broadband vendors will either shape up, or get run out of town on a rail.
SoupIsGood Food
MacOS X is BSD based, without a lick of Sys V in it. AIX is it's own thing, derived from the old OSF standard that was never adopted. (Not to be confused with the OSF pushed by DEC, which became Digital Unix, which became True64, which is BSD-based.)
Palm and Symbian are most certainly not System V based. Nor is VMS or OS/400. QNX, Plan 9, OS/9, RTOS are not Sys V based. The Amiga OS and its derivatives, Pheonix and Morph/OS are not Sys V based. Atheos is not Sys V based.
NetBSD has picked up some Sys V flavor in the past couple of years, but it's still, well, BSD. The other BSDs are definitely not Sys V, and free and clear of it with BSD 4.4 Lite based implementations.
Sys V based:
SCO (Unixware and SCO Unix)
Solaris (but with BSD flavor)
IRIX (Ditto.)
HP-UX (But with a Mach microkernel.)
Whatever name Fujitsu-Siemens is shilling Pyramid Unix as these days.
UNICOS
Non-Stop Unix
Doesn't Unisys still flog their own Unix?
You get the hint. Apart from Solaris and HP-UX, Sys V is only seen in ultra-tiny market segements boasting big server hardware.
SoupIsGood Food
Shock and outrage! Theo opens his mouth to bite the hand that feeds him, and so gets no bone? Who would have thought it would happen to such a sweet and affable fellow?
Bah.
Theo's legendary lack of tact and people-skills has sunk him... again. He can fork NetBSD and come out on top, he can fork OpenSSH and win the trademark dispute, he can fork IPfilter after alienating Darren Reed... I don't think he can fork the US Government. (Tho it would be a lot more stable and secure if he did... )
~Soop
What illicit technology offers the most fun and challenge today... where are the new frontiers for today's hackers to push the bleeding edge, and what interesting directions do you see them taking with it?
SoupIsGood Food
We do this in the US, too. If you haven't been since highschool, take a visit to your local public library: even in small towns there are often a half dozen PCs or Macs set up for the express purpose of browsing the 'net. You'd be shocked at how busy they are. Way back in the '70s, there were "public terminals" set up in places like Berkely... kiosks where people off the street could get and recieve email and participate on messageboards. This really helped to prime the pump for the personal computer revolution, and the BBS and Internet revolutions that followed. No reason why it can't work in Africa...
Obsolte PCs make great web browsing stations with the addition of Linux and the free web browser of your choice... set up a network of three or four of them in the town school/library/bar/whatever, and you're all set.
It occurs to me that a standardized "Library Terminal" distro, designed for robustness and single-click installation off of one CD, would be a great idea...
SoupIsGood Food
Verifying reverse-DNS isn't a terribly good idea... you're blocking mail comming from sites that do virtual hosting. In the olden days of one-to-one mapping of hostname and IP, it was a smart move. Now, when a single box can host hundred domains or more, filtering out connections from domains without reverse-DNS is going to cause more problems than it solves.
SoupIsGood Food
A Meta-Study is the most inbred and brain-damaged excuse for science I've seen in a long, long time.
OK, my own interest in low-carb diets is very personal and direct: I've lost more than 70 lbs on one. Not exactly a scientific reason to believe weight loss through the restriction of carbohydrates works, but a compelling one all the same.
So! From my perspective, a more usefull study would be clinical, and ask the following: Is my experience typical? Will I keep this weight off? What percentage of people following carb-restricted diets loose weight and keep it off?
Because low-calorie or not, everyone, and I mean everyone, I know who follows the Atkins diet looses weight (so long as they don't cheat... and resisting temptation is usually pretty easy.)
I'd like some clinical studies to back up this informal observation with hard statistics. Then we can break out the lood samples to figure out the biochemistry.
Dredging throug unrelated past studies to re-interpret the data in a light friendly to "common knowlege" doesn't seem much like science to me. Especially since I'm down four pant sizes. I will admit that I'm eating less... but it's not a conscious change. Unlike regular calorie restricted diets, my body doesn't rebel, and I'm able to stick with the diet... I'm eating less because my body is telling me to eat less. Again, this is an informal observation, but a scientific study to determine if this is indeed what's happening, and how it happens, would be more useful than the Stanford "meta-study."
SoupIsGood Food
The author's claim that Java is a dead-end language like COBOL is patently ludicrous, and by dismissing Java out of hand, he undermines the rest of his article.
Java is essentially a re-implementation of C; it's a very C-like language in syntax and semantics. The goal was to do an OO language based on a familiar paradigm. In addition to eliminating some of C's less endearing traits, it brings two things to the table, both of which which will shape future languages:
1. Re-usability. Sun offers you a crapload of very usefull re-usable objects in their JDKs, and people like the Apache project offers you even more. The ability to do insanely complex prrojects with tiny amounts of effort is one reason why Java rules the corporate enterprise. Future languages will start looking more and more like a big box of legos: find the parts you need and plug them together.
2) A universal computing environment: you can't write to the metal in Java, but it's not as slow as interpreted languages like Python and TCL for gigantic computing tasks. Any project, no matter how monolithic and task-optimized, is as portable as the VM is. Anyone who's had to manage a platform migration for key buisiness applications, from VAX to Solaris, say, or worse, from S/360 to Windows, knows the pain of re-implementation. That pain is gone when you use a VM-based language like Java.
Projects like Squeak are looking more and more like Java these days, in terms of re-usability and VM-based platform independance. The only missing piece of the puzzle are popular VM environments not tied to any one vendor: Java needs to cut its dependancy on Sun JDKs, or it will be supplanted by another language that is independant and standards based.
This isn't to say that other languages aren't going to evolve, too, or are useless because they're not like Java. Ayone who programs in the new interpreted scripting languages: PHP, Python, Perl, Ruby, TCL, Scheme, etc, etc, can attest to the power of the that approach to modern computing.
On the other hand, I really don't see any new compiled-to-the-metal languages emerging. Fortran is used for high-performance computing, Forth is used for tiny computers, and C/C++ is used for system programming. It will very likely be the same way in another 20 years, or another 50. The difference is that applications will slowly drift to either VM languages or interpreted languages from binaries compiled from source.
Survival Gear for Macheads:
.TXT.
System Software:
MacOS X Jaguar install disks
MacOS 9.2 install disk.
MacOS 8.1 install disk (for Oooold hardware)
System 7.1 boot floppy (for really, really old hardware.)
Software Utilities:
Alsoft DiskWarrior
Norton Utilities for Mac
Norton Antivirus
Retrospect
Hardware:
Apple-branded Firewire emergency backup and restore device (Or as we like to call it when not adding it to our equipment req: the iPod.) Go for the big one, you'll be glad you had it when you need to rescue your data from a flaky powerbook. Use Retrospect to make sure you get everything backed up proper.
Firewire to SCSI adapter (for getting data to and from older Macs.)
If you're going to dealing with real old Macs: AAUI dongle, phone-net adapters, Mini DIN 8 to DB24 and DB9 serial cables.
Unix Survival Kit:
Hardware:
Powerbook or iBook, with aforementioned Firewire SCSI adapter, USB serial adapter and a nice terminal emu program.
SCSI external HDD
Bunch 'o SCSI adapters/cables/testers
A SCSI CDROM... if you deal with Sun equipment, make sure it's able to boot a SPARC box.
Software:
Install CDs for your Unix flavors of choice.
CDs with the most current OS patch levels on it, one per OS.
Another CD with your customized dot files, shell scripts and all the useful stuff you really wished came with your vendor's Unix, but didn't (GNU).
NetBSD install CDs "for when all else fails." Comes in handy when you need to repurpose an old Motorola VME system previously installed with telco switching software to interface with lab monitoring hardware with shell scripts, a serial port and a prayer.
Documentation:
Unix System Administrator's Handbook on CD.
A copy of the "Fixing Solaris" howto in
Linux Kit:
Mac Powerbook or iBook, haughty sneer.
Software:
Slackware to inspire feeling of inadequacy in self proclaimed Linux gurus. Gentoo Level 3 on a USB keychain drive... especially usefull if you're stuck with a 2400 baud modem in a jungle where the pbone only works for three hours alternating tuesdays. (Ah, sarcasm!)
Copy of NetBSD or OpenBSD install disks to get real work done.
Windows Survival Kit:
Hardware:
Powerbook or iBook
Software:
Condescending sneer.
SoupIsGood Food
I wrote an
SoupIsGood Food
The most hated company in all of Macdom is not the beast from Redmond, who makes the tasty, lickable Office, but Quark. User-hostile doesn't even begin to cover its marketing and support... user belligerent is more like it. They flat-out refused to port to OSX (they still haven't), and they openly despise the Macintosh platform and insult its adherents at trade shows.
They are this way because they believed they had an unbeatable product, a single killer app the world could not do without: Xpress. The Mac dweebs would buy and keep buying, because there was no credible choice.
Until Adobe came up with InDesign, which is easier, faster, every inch as powerful, compatible with Xpress "Xtensions" and runs on OS X. Adobe shows their users lots of lovin', with trade shows, rational support, and deep Mac roots. Now InDesign is poised to topple Xpress into irrelevancy.
Adobe does not have the only pro-caliber image editing app out there. If they're upset that iPhoto killed ImageReady, and incensed that FinalCut destroyed Premier, wait until Apple decides to buy the TIFF-any codebase, or Avisa Image, or just roll their own Photoshop killer based on the GIMP.
Adobe is playing a very dangerous game. If Quark can be dethroned, you better damn well believe Photoshop can be, too. Apple's got pockets deep enough to do it, and marketing savvy that put FinalCut Pro on a Powerbook in the news vans of every TV station in the civilized world.
You don't take on Apple and win.
SoupIsGood Food
64-bitness won't increase speed. It will increase the amount of RAM you can throw at a problem. 48bit drum scans eat lots and lots of RAM, and of you're working with multiple images, you can bump up against Apple's current 1.5GB limit, and then virtual memory kicks in and performance goes to hell.
64bits on the desktop will be a great and good thing for all sorts of content creation chores.
SoupIsGood Food
Yup. UC Berkely, BSDi and BSD/386, the progenitor of the three *BSDs, Open, Net and Free, already had that lawsuit. AT&T wrung 'em through the wringer, UC Berkely slapped back with a copyright countersuit, and after much legal arangling, BSD 4.4 Lite came to be. It was free of patent and IP trouble, and then went on to become the backbone of BSDi and FreeBSD, which begat NetBSD, which begat OpenBSD. MacOS X was based on NeXT, which was a strange interpretation of BSD before 4.4 Lite. It's unixy bits are now based around current forks of FreeBSD code, IIRC, but it's no less strange.
SoupIsGood Food
Unix is very, very well understood, as the source code was freely distributed and then forked to hell and gone. Since the early '90s, the only thing all the Unix vendors really needed from Unix was the name, which is kindly provided by the Open Group these days so long as you have the cash.
AIX is not an SVR4-based Unix. Anyone who's taken a look at the psychotic commands required to admin an AIX box clearly understands this, as does anyone who's ever ported SVR4 based software to it.
Solaris and Irix are a little closer to their SVR4 roots, but True64 is BSD based. HP-UX is based around a Mach microkernel. MacOS X is based around both Mach and BSD, doesn't use any of the traditional file system or configuration tools, and it's still not as weird as AIX. So it can be easily argued that IBM did, does and will continue to do it's onw weird, scary thing without the need for anybody's code, patents or input. Exhibits A, B and C: zOS, OS/400 and OS/2. IBM has OS style kung-fu like no other company in the industry.
The offspring of 386BSD; Free, Open and Net are cleared of patent problems after a legal tussle with AT&T. Yahoo's been a FreeBSD shop from day one, so it's an enterprise class OS that got there without the deep pockets of any Computer Giant. There goes the foundation of the lawsuit.
What's worse, if you have to fear just one legal department, fear IBM's. They have a portfolio of defensive patents that will bury SCO so deep in countersuits, they'll never see the light of day.
So SCO is going to crash and burn and die for being stupid.
SoupIsGood Food
No counter argument, except it;s the same song that's been sung since the '70s. Personal comnputers were going to overthrow the old order, and relegate the mainframe to the dustbin of history! Didn't happen. From the early Apple II with Visicalc, to the "Standard" CP/M machines to the IBM PC, to the (IBM) Clone Wars, to the Mac and the 32bit broughaha, to NT, each new advance has been heralded as the ultimate commoditization of enterprise computing.
Bullshit. Enterprise desktops can't even be commoditized... when's the last time you saw a Fortune 500 company rolling out the generic white boxes to the desktop? Never. They all use top-tier desktop hardware vendors at three to five times markup over generic (or more). Meanwhile, VMS, AS/400 and HP3000 boxes still roll off the assmebly lines and into datacenters worldwide, despite their prospective vendors' best efforts to kill them as "obsolete"... there are many times when one big box that just works is more cost-competitive than lots of smaller, cheaper boxes.
Some applications can be commoditized... Google and Yahoo's server farms come to mind. Even so, Sun and IBM counter by putting the paralellization into software in their mainframe class systems, where the TCO and other spooooky financial metrics favor a few big, expensive boxes with solid service contracts and SLAs over bazillions of teensy ones.
The personal computer industry's been threatening to bury the big enterprise systems with an avalanche of commodified systems for a long, long time... and while commodity hardware has indeed found a permanent place in the datacenter, it's still a long, long, long ways off from supplanting everything else in there.
And Apple is Apple. So long as they're not bankrupt and being dismembered for liquidation, they live on, usually pulling in a nice, fat profit margin. (Didn't do so hot last year... but nobody except Dell was. Apple has 4 Billion bucks in cash, so a coupla million here or there means squat/all in the long range.)
SoupIsGood Food
Utter horseshit. Everyone lost sales last year: IBM, HP/DECompaq, SGI, Fujitsu-Siemens, Bull, NEC... everyone. Sun lost fewer sales than the other major players, so they picked up marketshare. They hemmorhaged money because of spotty buisiness practices from the dot-com era comming up to catch them, but as Cringley says, they've got another five years to get that sorted out.
Sun's transformation from a king of the workstation vendor into a server powerhouse that only IBM has any real hope of competing with is nothing short of phenomenal. If it looks like Linux is going to kill the proprietary Unix market, then Sun will go Linux... they've made similar moves in the past. Sun switched their entire installed base from the BSD-derived SunOS to the SVR4 flavored Solaris, no small feat at the time. A switch to Linux will be a snap, and yes, Sun will still charge you as much for your Sun-branded Linux as they do for Solaris, and get it from satisfied CIOs, along with fat service contracts.
Sun is never the first to market. Sun is never the ideal solution. Sun never offers the highest performance. Sun is never the cheapest option. Sun doesn't offer the best service in the industry. But they come "close enough" on so many fronts, they're an unbeatable market force. Add in Java, which rules enterprise computing like no technology since COBOL, and Sun ain't going nowhere.
Apple bled billions in the '90s, but they rebounded. McNealy's at least as smart as Jobs, and his marketing instincts are almost as honed. Sun has replaced IBM, and even Windows, in the hearts and minds of every serious CIO and VAR. Give it a year for either the economy to have rebounded, or for Sun to have staunched the bleeding on its own with austerity measures and something innovative. This is the company who managed to launch a line of workstations at the height of the NT onslaught in the Workstation market, and managed to make a mint with them. (The Ultra5 and Ultra10.) They aren't out of tricks yet.
Tho it would be nice if they put the screws to Fujitsu-Siemens to get access to their SPARC design... call it "SuperhyperultraSPARC" or "BadAssSPARC" or "TotallyAwesomeSPARC" or somesuch, and use it to hold the Itanium/POWER dogs at bay while they ready the UltraV.
SoupIsGood Food
I'm a Gen-X Slacker, but I forgot I was one for almost 6 years. I worked my ass off for this company or that, trusting their management until I got my ass canned, when I'd get another gig paying twice as much. Well, last year the gigs ran out. If you're a Unix admin, you better damn well have at least a Masters in a technology field, 'cuz nobody's gonna look twice at you otherwise. Recruiters tell me there are 150-200 resumes submitted for every open position, and everyday a new tech company folds, or goes through a round of layoffs. While I'm in the top 10% based on my skillset and experience that means there's 15 to 20 people who are in line ahead of me. Some schmoe who spent the last six years figuring out Tetris was N-P for a sheepskin is gonna get the nod because all I ever did was flunk out of art school.
/., and drinking. So! Cram a little code, email the hell out of reviewers and tech news blogs, set up a website with credit card ordering information, et voila! Software company. There's probably an open source way to scam money from people with big iron and deep pockets, but I'm too lazy to think of it. Go think of it yourself.
In short, I trusted that someone knew what they were doing in the big corner office, and now I'm fucked. And it's all my own damn fault, because I forgot the work ethics that brought me into computers and technology in the first place. Time for some of that olde timey religion:
1) SLACK! The world does owe you a living.
Screw 80+ hour workweeks for fat checks. I want 10 hour workweeks for fat checks, and a reasonable assumption that the fat checks will keep rolling in with minimal effort. I don't mind hellacious effort up front... being a lazy sysadmin, I know ten hours of scripting and testing can lead to a hundred hours or more of prime goof-off time. (Or, as the case was, time to write more scripts and to string more cable for my corporate overlords.)
So, by putting in a lot of work that's actually just fucking around with the computer up front, I can spend more time reading comics, posting trolls to
2) If it's worth doing, it's worth DIY.
The suit-wearing weasels sold me out. I'll never trust another. "Professional Management" means "porfessional backstabber." Screw 'em. I will learn the fundamentals of business management from one of the many excellent books at my local library, used book store or web bookseller.
I will learn how to market what I make, and how to balance the books. I will learn how to grow a company. I will never hire employees, but I will pay co-conspirators, and I will figure out a way to make this legal. I will figure out how to run a health insurance/HMO co-op with local small businesses. I will do all this with the meager funds from my teensy, just-above-minimum-wage non-computer job, and I will make any interested Venture Capitalists drink a bottle of robitussin, and I will laugh at them as they hallucnate and tell them to get the hell out. MY company, damnit!
Because if it's worth doing, it's worth doing my own damn self.
But first! I will learn Java and low-level C programming, for the things I am now interested in require that knowlege. I'll fake the rest as I need to. In a year, I'll come back to let you know how well I've succeeded. (I can't fail, as I'm already at baseline failure state right now. Any change is an improvement.)
Slack!
SoupIsGood Food
The battle for the 32bit desktop processor has been won, you mean. It's been a long, bloody battle, and for a time in the mid-'90s, when Apple was on the ropes, the victory looked complete. Intel and AMD have finally emerged victorious, laying claim to the most powerful 32bit processors on the planet.
But, to be frank, Itanium sucks... even Linus thinks so. Nobody wants to use it, Dell, SGI, even HP is still developing PA-RISC silicon, and is incredibly hesitant to commit to the "next generation" IA-64 chip it designed partly in-house. Yamhill is a nice idea, but Intel has no plans to go that route yet, and what's more, denies it's even considering them.
AMD's 64bit offering are, as yet, vapor... and unlikely to pack the punch of the Power4, nevermind a dual-core Power4 with Alti-Vec.
Meanwhile, PowerPC's been 64bit since '96.
Indeed, the PC will continue to kick Apple's butt in 32bit systems, except in notebook applications, which is the only place Apple will keep using 32bit PowerPC processors. D'oh.
So, yes, x86 is irrelevant and outclassed by PowerPC, Itanium is a floundering wreck, leaving Hammer to look very lonely and small up there all buy itself, shoulder to shoulder with UltraSPARCs, R2400s and Power4s. Economy of scale? What scale? When it comes to 64bit hardware, RISC/Unix =is= the scale.
Game on!
SoupIsGood Food
Right now, the prices for digital systems are stratospheric, but the advantages could, in theory, more than make up for the cost. You can use it to hold telecasts: bring lectures in from Oxford, live bigscreen showings of away games for the University's football team. The cost of renting movies drops through the floor as the cost of distribution approaches nil.
The problem is, there are no clear standards yet, and a whole lot of competing ones: Boeing, Technicolor, Sony, DLP, etc. You choose one option, there is =zero= guarantee it will use the digital projection standards for distribution and format that the rest of the industry winds up settling on. Then you have a $150,000 betamax VCR, and the professional equivalent to the dwindling "BETA" section at the local Video Store. No Oxford lectures. No away games. Not even Spiderman II.
I'd give the industry a few years to decide which way it wants to jump, or, barring that, a stone-solid contract from whichever vendor you go with that they will provide you with an upgrade to the equipment to make it compatible with dominant standard in case you pick a looser. Of course, if a proprietary standard is settled upon, you're screwed anyway.
SoupIsGood Food