You did say the compatability was different, your parent post explicitly states that the compatability subsystem in NT is different from the 'virtualization system' (as you put it) within OS/2.
And I'm not just talking about the posix subsystem, there were at least two comercial unix systems you could layer on OS/2, again, you didn't actually go Read about what I pointed you at. you are hopeless. Are you going to start claiming that IBM and MS had a lawsuit again over OS/2?
Print drivers. Go try it.
You did say Os/2 was a 16b OS, go back a few posts. It was your point #3. Then, when I said it was not 16b, you turned around and said it was a kludge, which was just a reiteration of your earlier statements.
You really keep trying to hammer home all these arguments that NT was/is better than os/2 - which if you actualy read what I've been saying, I never said.
The NT kernel, and much of the subsystems, have a common anscenstry with the OS/2 ver 3 project that MS and IBM worked on. When the split happened, they both took copies of that code. MS renamed is OS/2 NT, then just NT by the time of release. Reguardless of what changes MS may have wrought in the meantime, and reguardless of what IBM may have done with its copy, they both had the same basic archetecture going into the final rounds to finish ver 3 of each product.
You keep trying to educate me on the NT kernel, as if I don't know how it works. Now you claim it has a client-server model - which is incorrect. The OS Layers provide a client-server model, the NT kernel by itself does provide only a monolithic set of interfaces as with OS X (which you dragged into this conversation why?) - only with the addition of the win32 subsystem (or the posix, etc) does the 'kernel' provide a set of flexible, extendable interfaces in a client-server fashion. If you are going to include such layers, then a monolithic kernel design such as OS X can make the same claims - the GUI on OS X could be described as a client of the kernel's services, and yet the GUI libraries and manager provide a much broader se of interfaces than the OS X kernel itself.
You tell me to go look at microsoft.com, and yet you can't even sem to pull any references there for use in this conversation? Once again, no references, no believability.
You say I should buy the inside wNT books, which I have behind me on a shelf, and yet you apparently have never realized something very basic about many MS books - they are filled with propaganda that can turn out to be twisted and just plain wrong. this is why when you do research, you get more than one source for your information.
For example, I have a book on exchange here that attempts to describe mail systems before exchange was created, and it's description of original client-server systems is hideously generalized. Ignoring this generalization, it continues to enumerate the 'cons' of these types of systems - but due to its over generalization about how such systems worked, it misses the mark completely. Instead, it is only describing how a sub-set of client-server mail systems worked, and then the cons of that subset. However, from reading the book and never having exposed yourself to other (non-ms) sources of information, you would think that exchange was the end-all-be-all of mail systems (which, even from an objective standpoint its prety damn close) - and that all previous mail systems were writen by people with very small brains.
This is propaganda, the MS way. It has very little to do with how they are doing things on a technical level, and much more todo with how they want you to think their way is better than the competition. A non experienced user, or one with a selective memory, would read the book and then sneer at every pre-exchange mail system, making false assumptions. this sort of belief system is reinforced via the Lack of qualifiers in the text. No where does it say 'most client-server mail systems' or 'the majority of clie
OS/2 Code is used in the OS/2 subsystem. But if you understood NT architecture
Had you actually read what I had stated previously? Your entire rant (about #1) is based upon the belief that I misunderstood how the campatability layer worked. If you read my statement again you will see that I'm referencing what the book Told Me. Quite clearly, in the Introductory chapter, it explained where NT came from, and that it was its roots in OS/2 that lead to the compatability. Did I claim that NT had a compatability layer and that was proof? No, I referenced what a microsoft document stated.
As for the 'subsystem' being ripped out and the 'subsystem' being required - these statements I never made, you are once again reading too much into a very simple statement. Imagine you are a Kernel developer at microsoft, and you are working on the next NT kernel based OS. Os/2 compatability isn't required anymore, so what do you do? You remove the api functions it used from the Kernel (the ones not used by other systems - the OS/2 exclusive ones), and then you modify the executable loader to not recognize OS/2 executables as being properly runable applications. that was hard, wasn't it? Now, where in that example was OS/2 source code removed, that was in use? Source code from the original OS/2 ver 3 microkernel that was used by any other part of the OS remained. In fact, simply by modifying the executable loader you could 'remove' compatability.
As for what I know about the NT achetecture, thats off base, and rude, since you do _not_ know anything about what I do or do not know. So far, I know that you ignore facts and references from me, and claim to know everything. This wouldn't be so bad, except you don't seem to have references of your own to backup what you think.
Moving on.
For #2 & 3 - you seem to be going on a rant about what a kludge the OS was. This has little to do with my statements, that the OS was 32 bit. You could use 16 bit drivers with NT 3.0, if you were stupid, but that didn't make the OS 16 bit did it?
The GUI core of OS/2 and NT are different, but that has nothing to do with wether they had a common heritage - MS took the microkernel and ran with it, extending the win api to create the win32 api. The hardware abstraction layer was a feature being built into OS/2 3.0, as well as the move to a microkernel design (which, btw, anyone who really knows OS design will tell you that the NT kernel (and the OS/2 kernel) are not true microkernels in the classic sense.), and the OS/2 ver3 project was when the split happened.
You claim I should go do a web search, I did, thats where I got the links for my other response. It really was hard, I went to google and typed 'OS/2 NT history'. You claim I should go do some reading - I throw down the guantlet to this (because I _have_ done reading) and ask you to do the same. You repeatedly, through your statements, show your lack of knowledge, and then give no supporting evidence. It would actually be _less_ embarassing for your ignorance to come from bad sources, then for it to be sprouted as if it were the truth.
Let me give you some pointers on things to go read about:
Windows 3.x compatability versus the WOW system of NT - you claim they are different, but apparently that means you didn't read how they worked very well, especially in early versions of NT.
Could you buy a unix or other subsystem for OS/2? Yes. Relatedly, OS/2 got a posix subsystem as well, and its just as buggy and incomplete as the NT one. And yes, the NT one is incomplete - try doing anything that requires a fork() call on NT, and thats only the most blatent one.
Prove to me you aren't a troll, use some links in your next post, back up your statements.
No, you apparently didn't read them, from the second one:
"By late 1990, Microsoft had intensified its disagreements with IBM to the point where IBM decided that it would have to take some overt action to ensure that OS/2 development continued at a reasonable pace. IBM, therefore, took over complete development responsibility for OS/2 1.x, even though it was in its dying days, and OS/2 2.00. Microsoft would continue development on Windows and OS/2 3.00. Shortly after this split, Microsoft renamed OS/2 V3 to Windows NT."
This clearly shows that NT came from the os/2 ver 3 project - which was well under way by the time the split occured, as a paragraph a few before clearly states, from the 1989 days:
"Work had also begun on two new OS/2 products. Work on OS/2 2.0 was well underway. This product would be the first true 32 bit operating system for personal computers. Designed to work on the Intel 80386 and its follow on processors which were still in development, OS/2 2.00 would no longer be compatible with the 80286 processor.
OS/2 3.0 was in the very early stages of development and was intended at the time to be a network server version of the operating system. It was also intended to be platform independent. Because the operating system would be built on top of a microkernel, it would not need to be aware of the type of hardware on which it was running and therefore could run on Intel processors as well as Motorola, SUN, and DEC, chips with only a change of the microkernel hardware abstraction layer."
Note the microkernel. Note the hardware abstraction layer.
Now, I've found links to backup my side, and all you do is spout crap about 'everybody knows' - when in fact, you are wrong. Go get a book on the archetechture of windows NT4, or preferably, NT 3.5, from microsoft. goto a used bookstore. Read. Being able to backup your side with facts is elemental to an educated debate.
Here are references for the truth behind the NT and OS/2 origins, and how they DID share a common codebase. Note that one of these showed me something I did not know: OS/2 ver 1 was 16 bit, but in 89 was being rewritten fully 32-bit. this was Before the split with MS, which took place in 90, at which time, ver 3.0 was on its way to completion, designed as a network server, it was 32-bit. It was this version MS forked and renamed NT. I'm only listing the majorly relevant links, and unfortunatly, the interview linked to by the slashdot story is now gone, though you can find the text if you dig a little.
Ok, lets go point by point.
#1 - in correct. I have MCSE manuals here next to me for NT4. OS/2 compatability is there, the roots of the OS/2 compatibility came from the OS/2 project. this is a microsoft press book.
#2) See above, and, I think you know a lot less about OS/2 then you think you do... OS/2 warp server has just about all the features you just mentioned.
#3) OS/2 was and is 32-bit. Go look it up.
As an expansion on this, why don't you go look up how OS/2 ran 16-bit windows 3.x applications - it ran it via an isolated virtual dos machine process that ran true dos and windows. this is the same system that MS used in NT to run 16-bit application, they called it Windows on Windows. This, by the way, contradicts your assertion that what OS/2 had was a compatability layer - it was a full virtual machine subsystem, exactly as the one MS had in NT. In fact, a version of OS/2 was released that was cheaper than the full version - but required you to have a copy of window 3.0 for compatability - because by not including the actual windows files with OS/2, IBM didn't have to pay MS for a windows liscense.
As for it being a 'poor IBM story' - how did you get that from my post? I called NT the ugly stepsister of OS/2, I was refering directly to its lineage, not making a reference to a fairy tale or implying that OS/2 was better.
Get a brain. go read some books.
incorrect on the NT citation.
NT's security system was designed by the same man who worked on VMS and defined it's security system (mind you, this guy hates it when people link him to NT security, since MS took what he gave them and did it badly).
NT was the ugly stepsister to OS/2 - both OS/2 and NT originally had the same codebase and where a MS-IBM joint project. NT was the fork MS ran with after they left IBM out in the cold.
Also, if MS 'borrowed' the gui from apple, it should be noted that apple 'borrowed' it from Xerox parc.
Having said all of that:
Viso: bought
Frontpage: bought
SQL Server: forked from 'joint project cross liscensing agrement' with Sybase (ie, stolen underhandedly)
USDOJ: bought
It must be doing really well then - because all I can get ahold of are hardbacks, and I had to go to three bookstores just to find one in stock:)
Good news for the publisher and the author.
less than 30 days ago, each store which is now sold-out had many copies of Redemption Ark(hardback) and Chasm City in stock - I know this because I was hunting around for Rev Space, which everyone was sold out of at the time.
I just finished Revelation Space and Chasm City (A related novel, but not a direct sequel or prequel to rev space), and I must say, these were the best novels sci-fi novels I've read in some time. Reynolds plays a little fast and lose with the explainations of the technology at times (usually writing it off as the POV character doesn't know the details) - but that only enhances the realism of the world, since the POV character is rarely someone who would know how things work.
I'm going to buy the latest, Redemption Ark later today. It's still in hardback - but the wait to see what happens to the characters from Rev Space would be too long if I waited for a paperback. Normally I don't buy hardbacks...
Any speakers that do not include circuitry to self-limit thier output (which only some very high end speakers have) will output distorted sound when fed too much power.
The amp in these speakers is too generous with it's output, and is sending too much power to the speakers when set to a high volumn.
This is common place in all commodity speakers and speaker systems. Manufacturers make them this way because the source audio you are playing may have been recorded at a low volumn - and thus needs to be amplified to a higher degree than the 'average' sound being put through the system. IE, sound recordings are not normalized for volumn to a 'standard'.
Go get an 'executive' sterero for a couple of hundred dollars. turn it up too loud, and you'll get distortion. Take a factory stereo in a car, and turn it up too loud, and you'll get distortion.
I guess my point is: how is this a 'con' for this set of speakers, when 99.99% of all speakers bought suffer frm the same problem? It's like complaining that cars have only four wheels.
My old 1.1GHz athlon got fried.
Occasionaly the CD-rom drive would spin up for no (apparent) reason, and would vibrate like mad - the sound it made was very annoying.
I'd sometimes hit the top of my case, which would cause the sound to stop as the cd-rom drive was jolted into a slightly different position.
I apparently hit my case a lot harder then I thought when I'd do this - because one time, the last time, my screen all of a sudden went blank about 10 seconds after I'd hit the case. No signal. I looked down under my desk at the tower in confusion - looking at the power LED. It was on.
Then I smelled something... like something burning. Not a woodsy smell, like a wood fire, but the kind of fried electronics smell a power supply makes as it goes. I quickly turned off my machine - thinking the power supply had gone bad.
Nope. The CPU heatsink was sitting on top of my agp card from where it had fallen off. Apparently, I had knocked it loose over time, and the last jolt was too much.
The CPU enver worked again. No flames, no obvious burn marks or sillyness like that, it just refused to post ever again.
Maybe it's just me, but I've never heard of this sort of take on multiple monitors - that they are for gamers. I've been using multiple monitors (In fact, I've insisted on them) since 1998 at home and the office, and I've never heard this argument.
If I'd had heard it, I would of laughed - dual (or triple) monitors are usually a pain when it comes to gaming, because most games (for windows) don't handle it very well if you leave the other monitors on and 'attached' via the display settings panel. Thus, you have to go and set each extra monitor as 'unattached' before launching starcraft or some other games. Forgetting to do so can cause big issues, since in the middle of a good game of craft, moving your mouse into the other monitor's 'space' and clicking, usually by accident, can throw you back to your windows desktop and minimize the game - a very Bad Thing. Even if all you played were games that didn't suffer from this limitation, running the second monitor while gaming puts extra load on your machine - slowing down the response of your game.
Has anyone else ever heard this sort of BS from an 'IT manager' - and if so, do you know _why_ they'd have this odd misconception?
He states: I don't run any windows computers, and/dev/null'ing viruses are trivial.
And then at the end he states: save us all money on bandwidth, hard disk space, and general anguish.
Your 'email system' is a CLIENT - he is talking about email servers. He never mentions using outlook, the term 'Outlook' was used to describe his opinion/request, as-in 'his outlook on the subject'.
Even if his server is configured to stop the spam and viruses by piping them to/dev/null, his bandwidth is still being eaten up (you can't scan what you haven't yet recieved). People pay for bandwidth, especially people who have dedicated servers or colo - and just because someone else's server bounced the message to yours, doesn't mean that you don't get 'digned' at bill time for the bandwidth.
In addition to the extra cost, and it can add up if you run a server that has many email users (all of whom may be being sent the virus, and whom may be recieving bounces from forged virus emails), not only can the virus eat up a lot of bandwidth over time, it can Slow You Down. 100k a virus email or bounce... 200 users... 1 bounce or attempted deliver to one of those users every 40 seconds or so... your pipe is full. Qmail is busy. Its an email slashdot effect. It can slow everything down.
Remember, QT is a library, and trolltech makes their money from it. I'm pretty sure that all that will be needed for most apps using current QT is a recompile with the new tools (QT has a tool used as part of the Make process). To use the new features might require changes to code, but thats a different story - you're already changing your code to add new features.
Oh, I'm not talking about closing and opening accounts, I'm just talking about reporting the card stolen. That card gets deactivated and they send you a new one, with a new expiration date, and sometimes a new number on the card itself.
Gee, and having a fake ID made up is easier todo then just use an already signed card? The whole point is to make it tougher for the criminal so that you have more time to report the card stolen.
I'm willing to take steps now to increase my security at the cost of convenience. (emphasis mine)
Not to rain on your parade or anything, but from the trouble you've taken (different passwords everywhere, spending money on periodic credit reports, premptive fraud alarms attached to your credit that will make applying for any sort of acredit a Pain for you) you already ARE at the point where you've given up a lot of convienence.
That being said, the one thing I've done that is 'inconvienent' is I don't sign my credit cards. Now, I don't mean I leave them blank - thats asking to get ripped off (anyone who steals the card can sign your name) - I put 'SEE ID' in the signature area. Mind you, a few places don't even bother to check the sig area, so you're still SOL if someone steals your card and uses it at a lax restraunt or gas station, but having the guy behind the counter ask to see a photo ID every time I buy something expensive feels like a good tradeoff to me.
Of course, someone could always make a fake ID with their photo and my name on it, but thats a lot of effort, and frankly, I'm not That paranoid. I have fraud insurance on all of my accounts, and have very clean credit. If I loose a card at an ATM (by forgetting it), or loose a imprinted recipt, I call the card company immediatly. Having a track record of getting new cards whenever something like that happens does wonders when there is something questionable on your statement and you call about it.
Along that vein, a friend of mine recommends reporting your card lost once or twice a year, just to get new cards with different numbers.
Then again, that friend is a little bit more paranoid then I am... He's about as paranoid as you are... Hey, wait a second, you're name isn't Bryan is it?
Yes, I could. But to charge them each infringment would still require going through a lawyer to try to negotiate (they don't respond to my calls/letters otherwise), which incurs legal fees, which rapidly skyrocket out of control.
Oddly enough, the Best thing I ever found for possible use against them was the DMCA - it has a provision in it stating that removal of a copy-protection feature without permission is criminal, and can lead to civil damages. It also expands on that provision so that every infringing use of the product that has had the feature removed can be treated as an additional infringement, criminal and finable. Why was this applicable? - Because I made a backdoor. When things first started to get funky, I put in a special username and password combination - 50+ characters each, so almost impossible to crack - that if ever used would cause the software to delete itself. It was specifically designed to only affect the software, not the data, so their 100k user records in the email system would still be there. My software would be gone, but email and the user data would continue as normal. However, when I mentioned to them (late in the game, after they had rolled the software 'live' in multiple locations and were trying to pay me 1/4th the bill) that this existed and that if they don't pay it would go bye-bye, they hired someone to track it down and remove it. By removing this backdoor - a copyright mechanism I had put in place to protect myself - they violated the DMCA. Lovely, eh?
These are all fine ideas, and ones I now use regularly in my business - I don't think I mentioned that despite the BS I went through I ended up deciding to be my own boss.
I think in my case, the bigest things I could of done differently were to have opened my eyes over two things:
1) Company Y had already proven itself, by how it shut down and sold off my employeer X, to be a ruthless and rude entity.
2) Person A had a bad reputation outside the tech departments of X and Y for being an ass, and being untrustworthy.
With #1, I had blinders on and saw the oppourtunity for making money. With #2, I had my blinders on and thought my previous experiences with person A entitled him to 'friend' status. In both instances I was dead wrong, and If I'd of taken a step back I'd of given the whole thing more thought ahead of time.
sometimes some perspective is all we need, and I've found that keeping your perspective and seeing the whole, big picture can be an important part of keeping a biz afloat.
For example: I recently met with a client who had grand ideas, and wanted all sorts of things done. I wrote a proposal and sent it to them, a short proposal, very informal. considering the features involved, this could be a huge project, so why didn't I write a detailed proposal and send over a work order? Because the (potential) client is a startup, and has money problems. I know that If I spend too much time on proposals for them and they go under I've lost potential business I could of been landing. The (potential) client is bothered by the informal proposal - even though it was informal it has a price range in it. We keep in contact and if they want my services they'll either try to haggle or cut some features. Or, they'll come back 6 months down the line and say 'we're ready to do this now' - which happens more often then you'd believe.
If it took the well-publicized OJ trial to shake your faith, then you weren't paying much attention. It's been f-ed up for awhile.
Heres a post-OJ example: I used to work for company X, which was owned by company Y. Company X laid me off. The very next day I get contacted by a big honcho (lets call him A) at X (and Y, CTO for both) to do some contracting work on an unfinished project for Y. I agree and get to work. I work, and I work, and I rack up hours, all billable via the agreement I had with A. I fix all the bugs, I deliver the system. And then Y doesn't pay....and doesn't pay....and leaves me in the lurch. I find out the software I'd written is about to be taken from its single location and rolled out nationaly by Y, and I get pissed and start bugging the hell out of my contact at Y (person B). Person B tells me that A was never allowed to authorize work for Y, and puts me in contact with a different honcha, person C. Person C claims they are evaluating my software and comparing it against other off-the-shelf systems. I make it very clear to C that this matters not: I agreed to do work for a specific price per hour, and they are using the fruits of my labor. He shrugs it off and says he is not bound by the agreements A made. Over the next two months, I get the runaround, and then finally C offers to pay less than 1/4rd of the bill as 'a fair liscensing fee based upon our research'. By this time, I had debt upto my eyeballs, but I still had some sense of reason and pride. I made it clear to C that I, not they, get to set liscensing fees for any product I decide to sell, and also stress that liscense was never an issue: I did contract work for them modifying an existing piece of software and fixing bugs, and I expected to get paid. C said the 1/4th was the best he could do. I said I'd go talk to a lawyer.
Here's where the legal system steps in, the above was backstory. I go talk to a lawyer. With my documentation in hand, he said I have a very strong case - should it ever go to court. But he sugests we do some research and try other tactics (letter from his law firm, etc) first, because of a (to me) glaring problem with how these sort of civil cases go: No matter how big the organization in the wrong is, or how obviously in the wrong they are, it is soley upto the judges discretion as to wether the party in the wrong has to pay legal fees for the complaining party (should the complaining party win). Added to this fact, was that in the conservative city I live in, judges rarely make defendants pay for the complaining party's fees.
This upset me quite a bit.
Then the research into the company came down - just googling stuff I and the lawyer did. Y was run by lawyers and ex lawyers, and no suit brought against them ever made it out of court - they dragged their feet for years on anything they could, and used heavy-handed tactics and threats of countersuits whenever possible to stop suits dead in their tracks by scaring off the person or company making the complaint.
So, here I am, almost a year later, still out TONS of $$ (>20k), and I can't afford to pursue the matter. For every letter they write in response to something my lawyer would send, I'd be out 100-250$ (assuimng it was a short letter my lawyer coudl respond to quickly). If I were to take the case to court, depositions alone could bankrupt me. I entered into a binding agreement with an officer of company Y, and I can't collect squat from it.
Meanwhile, company Y is doing well. They have over 100000 users nationwide using the software I fixed and improved for them.
What it comes down to is this (as the prosecution found out the hard way in the OJ trial): the legal system protects those who have money. If you have tons of money (for lawyers), you can get away with almost anything, or at least make it so painful, expensive and drawn-out for anyone to come after you that it isn't worth it.
Back to the story at hand though: IBM has very deep pockets. But SCO is going to spend eery penny it has on lawyers, and can draw this out for quite awhile. Realisticaly, we wont have an answer to the is-the-gpl-enforcable question for Years.
Re:Black, white, and 10,000 shades of grey
on
Mac OS X Power Tools
·
· Score: 3, Funny
shouldn't that be 16384 (2^14) or 8192 (2^13) shades of grey? I mean, this is slashdot...
Right now, thier shares are at 14.71USD a piece. Assuming thier stock prices rise as we buy, lets project a 15.5 avg price per share. The question is, and I can't find this offhand, is how many shares of stock we'd need to buy? I know I'd chip in for 6 shares, and if 100,000 linux users/slashdot readers all bought an avg of two shares, we're looking at 2.5 million shares purchased... Which gets us what % of the company, exactly? Ok, after digging on finance.yahoo I've found that there are 13.1 million shares outstanding. that means we need to buy 6.56 million shares. We need 101.68 million dollars. Ouch. That really throws my numbers out of whack. Ok then, we need 6.56 million linux users to buy a share a piece, or 1 million users to buy just over 6 shares a piece. Thats a big organizational challenge... Anyone at FSF or GNU ready to take this on? (the buy-shares forms/donations system and handing over rights for voting with said stock to said org. Realisticly, it might be easier for people to donate cash towards buying stock, not buying it direct, thatway FSF or whomever gets full control of whats bought)
Also: I don't think selling the operation to redhat for a dollar would be the right solution though - thats too politically (in linux circles) charged (to let one distro get thier stuff) - I think thier assests should be liquified, and the results split 50-50 with the fsf getting half to fund dev projects, and 50% going back to the stockholders. Of course, there wouldn't be much left once you consider that many customers with support contracts (and yes, some people have paid sco for support, even pre lawsuit) would want refunds.
Also Also: Yes, I know many people would be against this on moral grounds - 'its the principle of the thing!' - but how many hours and how much money will we waste trying to promote linux against this FUD-crap from SCO? And, what sort of message would a SCO buyout send the world... It'd tell M$ and everyone else that f-ing with Linux is a dangerous proposition, because the community will unite and topple all who oppose us...
Uhm, WHY would an organization have a table filled with names that isn't indexed? I agree with in concept, but any organization that has its act together enough to consider using honeytokens and honeypots should also know better than to have crappy db schemas where something like a customer's name isn't indexed...
Slashdot Mirror
Ah, here we go again.
You did say the compatability was different, your parent post explicitly states that the compatability subsystem in NT is different from the 'virtualization system' (as you put it) within OS/2.
And I'm not just talking about the posix subsystem, there were at least two comercial unix systems you could layer on OS/2, again, you didn't actually go Read about what I pointed you at. you are hopeless. Are you going to start claiming that IBM and MS had a lawsuit again over OS/2?
Print drivers. Go try it.
You did say Os/2 was a 16b OS, go back a few posts. It was your point #3. Then, when I said it was not 16b, you turned around and said it was a kludge, which was just a reiteration of your earlier statements.
You really keep trying to hammer home all these arguments that NT was/is better than os/2 - which if you actualy read what I've been saying, I never said.
The NT kernel, and much of the subsystems, have a common anscenstry with the OS/2 ver 3 project that MS and IBM worked on. When the split happened, they both took copies of that code. MS renamed is OS/2 NT, then just NT by the time of release. Reguardless of what changes MS may have wrought in the meantime, and reguardless of what IBM may have done with its copy, they both had the same basic archetecture going into the final rounds to finish ver 3 of each product.
You keep trying to educate me on the NT kernel, as if I don't know how it works. Now you claim it has a client-server model - which is incorrect. The OS Layers provide a client-server model, the NT kernel by itself does provide only a monolithic set of interfaces as with OS X (which you dragged into this conversation why?) - only with the addition of the win32 subsystem (or the posix, etc) does the 'kernel' provide a set of flexible, extendable interfaces in a client-server fashion. If you are going to include such layers, then a monolithic kernel design such as OS X can make the same claims - the GUI on OS X could be described as a client of the kernel's services, and yet the GUI libraries and manager provide a much broader se of interfaces than the OS X kernel itself.
You tell me to go look at microsoft.com, and yet you can't even sem to pull any references there for use in this conversation? Once again, no references, no believability.
You say I should buy the inside wNT books, which I have behind me on a shelf, and yet you apparently have never realized something very basic about many MS books - they are filled with propaganda that can turn out to be twisted and just plain wrong. this is why when you do research, you get more than one source for your information.
For example, I have a book on exchange here that attempts to describe mail systems before exchange was created, and it's description of original client-server systems is hideously generalized. Ignoring this generalization, it continues to enumerate the 'cons' of these types of systems - but due to its over generalization about how such systems worked, it misses the mark completely. Instead, it is only describing how a sub-set of client-server mail systems worked, and then the cons of that subset. However, from reading the book and never having exposed yourself to other (non-ms) sources of information, you would think that exchange was the end-all-be-all of mail systems (which, even from an objective standpoint its prety damn close) - and that all previous mail systems were writen by people with very small brains.
This is propaganda, the MS way. It has very little to do with how they are doing things on a technical level, and much more todo with how they want you to think their way is better than the competition. A non experienced user, or one with a selective memory, would read the book and then sneer at every pre-exchange mail system, making false assumptions. this sort of belief system is reinforced via the Lack of qualifiers in the text. No where does it say 'most client-server mail systems' or 'the majority of clie
Once again, no references, unless you want to claim all of this comes from that one edition.
First, you claim the articles prove your side right, then you claim they are wrong in general - which is it? You cant have it both ways.
Your response is the typical 'I know because I know' troll, and once again fails to take into account what anyone else may or may not know.
OS/2 Code is used in the OS/2 subsystem. But if you understood NT architecture
Had you actually read what I had stated previously? Your entire rant (about #1) is based upon the belief that I misunderstood how the campatability layer worked. If you read my statement again you will see that I'm referencing what the book Told Me. Quite clearly, in the Introductory chapter, it explained where NT came from, and that it was its roots in OS/2 that lead to the compatability. Did I claim that NT had a compatability layer and that was proof? No, I referenced what a microsoft document stated.
As for the 'subsystem' being ripped out and the 'subsystem' being required - these statements I never made, you are once again reading too much into a very simple statement. Imagine you are a Kernel developer at microsoft, and you are working on the next NT kernel based OS. Os/2 compatability isn't required anymore, so what do you do? You remove the api functions it used from the Kernel (the ones not used by other systems - the OS/2 exclusive ones), and then you modify the executable loader to not recognize OS/2 executables as being properly runable applications. that was hard, wasn't it? Now, where in that example was OS/2 source code removed, that was in use? Source code from the original OS/2 ver 3 microkernel that was used by any other part of the OS remained. In fact, simply by modifying the executable loader you could 'remove' compatability.
As for what I know about the NT achetecture, thats off base, and rude, since you do _not_ know anything about what I do or do not know. So far, I know that you ignore facts and references from me, and claim to know everything. This wouldn't be so bad, except you don't seem to have references of your own to backup what you think.
Moving on.
For #2 & 3 - you seem to be going on a rant about what a kludge the OS was. This has little to do with my statements, that the OS was 32 bit. You could use 16 bit drivers with NT 3.0, if you were stupid, but that didn't make the OS 16 bit did it?
The GUI core of OS/2 and NT are different, but that has nothing to do with wether they had a common heritage - MS took the microkernel and ran with it, extending the win api to create the win32 api. The hardware abstraction layer was a feature being built into OS/2 3.0, as well as the move to a microkernel design (which, btw, anyone who really knows OS design will tell you that the NT kernel (and the OS/2 kernel) are not true microkernels in the classic sense.), and the OS/2 ver3 project was when the split happened.
You claim I should go do a web search, I did, thats where I got the links for my other response. It really was hard, I went to google and typed 'OS/2 NT history'. You claim I should go do some reading - I throw down the guantlet to this (because I _have_ done reading) and ask you to do the same. You repeatedly, through your statements, show your lack of knowledge, and then give no supporting evidence. It would actually be _less_ embarassing for your ignorance to come from bad sources, then for it to be sprouted as if it were the truth.
Let me give you some pointers on things to go read about:
Windows 3.x compatability versus the WOW system of NT - you claim they are different, but apparently that means you didn't read how they worked very well, especially in early versions of NT.
Could you buy a unix or other subsystem for OS/2? Yes. Relatedly, OS/2 got a posix subsystem as well, and its just as buggy and incomplete as the NT one. And yes, the NT one is incomplete - try doing anything that requires a fork() call on NT, and thats only the most blatent one.
Prove to me you aren't a troll, use some links in your next post, back up your statements.
Read your own damn articles...
No, you apparently didn't read them, from the second one:
"By late 1990, Microsoft had intensified its disagreements with IBM to the point where IBM decided that it would have to take some overt action to ensure that OS/2 development continued at a reasonable pace. IBM, therefore, took over complete development responsibility for OS/2 1.x, even though it was in its dying days, and OS/2 2.00. Microsoft would continue development on Windows and OS/2 3.00. Shortly after this split, Microsoft renamed OS/2 V3 to Windows NT."
This clearly shows that NT came from the os/2 ver 3 project - which was well under way by the time the split occured, as a paragraph a few before clearly states, from the 1989 days:
"Work had also begun on two new OS/2 products. Work on OS/2 2.0 was well underway. This product would be the first true 32 bit operating system for personal computers. Designed to work on the Intel 80386 and its follow on processors which were still in development, OS/2 2.00 would no longer be compatible with the 80286 processor.
OS/2 3.0 was in the very early stages of development and was intended at the time to be a network server version of the operating system. It was also intended to be platform independent. Because the operating system would be built on top of a microkernel, it would not need to be aware of the type of hardware on which it was running and therefore could run on Intel processors as well as Motorola, SUN, and DEC, chips with only a change of the microkernel hardware abstraction layer."
Note the microkernel. Note the hardware abstraction layer.
Now, I've found links to backup my side, and all you do is spout crap about 'everybody knows' - when in fact, you are wrong. Go get a book on the archetechture of windows NT4, or preferably, NT 3.5, from microsoft. goto a used bookstore. Read. Being able to backup your side with facts is elemental to an educated debate.
Here are references for the truth behind the NT and OS/2 origins, and how they DID share a common codebase. Note that one of these showed me something I did not know: OS/2 ver 1 was 16 bit, but in 89 was being rewritten fully 32-bit. this was Before the split with MS, which took place in 90, at which time, ver 3.0 was on its way to completion, designed as a network server, it was 32-bit. It was this version MS forked and renamed NT. I'm only listing the majorly relevant links, and unfortunatly, the interview linked to by the slashdot story is now gone, though you can find the text if you dig a little.
e ID/13464/1 3464.htmlr y.html2 09.shtml
Enjoy.
http://www.winnetmag.com/Article/Articl
http://www.os2bbs.com/os2news/OS2Histo
http://slashdot.org/articles/99/08/25/124
Ok, lets go point by point. #1 - in correct. I have MCSE manuals here next to me for NT4. OS/2 compatability is there, the roots of the OS/2 compatibility came from the OS/2 project. this is a microsoft press book. #2) See above, and, I think you know a lot less about OS/2 then you think you do... OS/2 warp server has just about all the features you just mentioned. #3) OS/2 was and is 32-bit. Go look it up. As an expansion on this, why don't you go look up how OS/2 ran 16-bit windows 3.x applications - it ran it via an isolated virtual dos machine process that ran true dos and windows. this is the same system that MS used in NT to run 16-bit application, they called it Windows on Windows. This, by the way, contradicts your assertion that what OS/2 had was a compatability layer - it was a full virtual machine subsystem, exactly as the one MS had in NT. In fact, a version of OS/2 was released that was cheaper than the full version - but required you to have a copy of window 3.0 for compatability - because by not including the actual windows files with OS/2, IBM didn't have to pay MS for a windows liscense. As for it being a 'poor IBM story' - how did you get that from my post? I called NT the ugly stepsister of OS/2, I was refering directly to its lineage, not making a reference to a fairy tale or implying that OS/2 was better. Get a brain. go read some books.
incorrect on the NT citation. NT's security system was designed by the same man who worked on VMS and defined it's security system (mind you, this guy hates it when people link him to NT security, since MS took what he gave them and did it badly). NT was the ugly stepsister to OS/2 - both OS/2 and NT originally had the same codebase and where a MS-IBM joint project. NT was the fork MS ran with after they left IBM out in the cold. Also, if MS 'borrowed' the gui from apple, it should be noted that apple 'borrowed' it from Xerox parc. Having said all of that: Viso: bought Frontpage: bought SQL Server: forked from 'joint project cross liscensing agrement' with Sybase (ie, stolen underhandedly) USDOJ: bought
It must be doing really well then - because all I can get ahold of are hardbacks, and I had to go to three bookstores just to find one in stock :)
Good news for the publisher and the author.
less than 30 days ago, each store which is now sold-out had many copies of Redemption Ark(hardback) and Chasm City in stock - I know this because I was hunting around for Rev Space, which everyone was sold out of at the time.
I just finished Revelation Space and Chasm City (A related novel, but not a direct sequel or prequel to rev space), and I must say, these were the best novels sci-fi novels I've read in some time. Reynolds plays a little fast and lose with the explainations of the technology at times (usually writing it off as the POV character doesn't know the details) - but that only enhances the realism of the world, since the POV character is rarely someone who would know how things work.
I'm going to buy the latest, Redemption Ark later today. It's still in hardback - but the wait to see what happens to the characters from Rev Space would be too long if I waited for a paperback. Normally I don't buy hardbacks...
Any speakers that do not include circuitry to self-limit thier output (which only some very high end speakers have) will output distorted sound when fed too much power.
The amp in these speakers is too generous with it's output, and is sending too much power to the speakers when set to a high volumn.
This is common place in all commodity speakers and speaker systems. Manufacturers make them this way because the source audio you are playing may have been recorded at a low volumn - and thus needs to be amplified to a higher degree than the 'average' sound being put through the system. IE, sound recordings are not normalized for volumn to a 'standard'.
Go get an 'executive' sterero for a couple of hundred dollars. turn it up too loud, and you'll get distortion.
Take a factory stereo in a car, and turn it up too loud, and you'll get distortion.
I guess my point is: how is this a 'con' for this set of speakers, when 99.99% of all speakers bought suffer frm the same problem? It's like complaining that cars have only four wheels.
My old 1.1GHz athlon got fried.
Occasionaly the CD-rom drive would spin up for no (apparent) reason, and would vibrate like mad - the sound it made was very annoying.
I'd sometimes hit the top of my case, which would cause the sound to stop as the cd-rom drive was jolted into a slightly different position.
I apparently hit my case a lot harder then I thought when I'd do this - because one time, the last time, my screen all of a sudden went blank about 10 seconds after I'd hit the case. No signal. I looked down under my desk at the tower in confusion - looking at the power LED. It was on.
Then I smelled something... like something burning. Not a woodsy smell, like a wood fire, but the kind of fried electronics smell a power supply makes as it goes. I quickly turned off my machine - thinking the power supply had gone bad.
Nope. The CPU heatsink was sitting on top of my agp card from where it had fallen off. Apparently, I had knocked it loose over time, and the last jolt was too much.
The CPU enver worked again. No flames, no obvious burn marks or sillyness like that, it just refused to post ever again.
Maybe it's just me, but I've never heard of this sort of take on multiple monitors - that they are for gamers. I've been using multiple monitors (In fact, I've insisted on them) since 1998 at home and the office, and I've never heard this argument.
If I'd had heard it, I would of laughed - dual (or triple) monitors are usually a pain when it comes to gaming, because most games (for windows) don't handle it very well if you leave the other monitors on and 'attached' via the display settings panel. Thus, you have to go and set each extra monitor as 'unattached' before launching starcraft or some other games. Forgetting to do so can cause big issues, since in the middle of a good game of craft, moving your mouse into the other monitor's 'space' and clicking, usually by accident, can throw you back to your windows desktop and minimize the game - a very Bad Thing. Even if all you played were games that didn't suffer from this limitation, running the second monitor while gaming puts extra load on your machine - slowing down the response of your game.
Has anyone else ever heard this sort of BS from an 'IT manager' - and if so, do you know _why_ they'd have this odd misconception?
how could it be the nebuchadnezzar? It got blown away in the last movie.
Besides, the ship is _not_ the typical neb-type hovership when you look close at it.
He states: /dev/null'ing viruses are trivial.
/dev/null, his bandwidth is still being eaten up (you can't scan what you haven't yet recieved). People pay for bandwidth, especially people who have dedicated servers or colo - and just because someone else's server bounced the message to yours, doesn't mean that you don't get 'digned' at bill time for the bandwidth.
I don't run any windows computers, and
And then at the end he states:
save us all money on bandwidth, hard disk space, and general anguish.
Your 'email system' is a CLIENT - he is talking about email servers. He never mentions using outlook, the term 'Outlook' was used to describe his opinion/request, as-in 'his outlook on the subject'.
Even if his server is configured to stop the spam and viruses by piping them to
In addition to the extra cost, and it can add up if you run a server that has many email users (all of whom may be being sent the virus, and whom may be recieving bounces from forged virus emails), not only can the virus eat up a lot of bandwidth over time, it can Slow You Down. 100k a virus email or bounce... 200 users... 1 bounce or attempted deliver to one of those users every 40 seconds or so... your pipe is full. Qmail is busy. Its an email slashdot effect. It can slow everything down.
Now, do you still think he's whining?
Get a brain.
Remember, QT is a library, and trolltech makes their money from it. I'm pretty sure that all that will be needed for most apps using current QT is a recompile with the new tools (QT has a tool used as part of the Make process). To use the new features might require changes to code, but thats a different story - you're already changing your code to add new features.
Oh, I'm not talking about closing and opening accounts, I'm just talking about reporting the card stolen. That card gets deactivated and they send you a new one, with a new expiration date, and sometimes a new number on the card itself.
Gee, and having a fake ID made up is easier todo then just use an already signed card? The whole point is to make it tougher for the criminal so that you have more time to report the card stolen.
Brain, meet faedle.
No, not 'extra' fraud insurance - I just make sure I never accept a card that doesn't offer fraud protection.
Sorry, I guess I phrased that badly.
I'm willing to take steps now to increase my security at the cost of convenience.
(emphasis mine)
Not to rain on your parade or anything, but from the trouble you've taken (different passwords everywhere, spending money on periodic credit reports, premptive fraud alarms attached to your credit that will make applying for any sort of acredit a Pain for you) you already ARE at the point where you've given up a lot of convienence.
That being said, the one thing I've done that is 'inconvienent' is I don't sign my credit cards. Now, I don't mean I leave them blank - thats asking to get ripped off (anyone who steals the card can sign your name) - I put 'SEE ID' in the signature area. Mind you, a few places don't even bother to check the sig area, so you're still SOL if someone steals your card and uses it at a lax restraunt or gas station, but having the guy behind the counter ask to see a photo ID every time I buy something expensive feels like a good tradeoff to me.
Of course, someone could always make a fake ID with their photo and my name on it, but thats a lot of effort, and frankly, I'm not That paranoid. I have fraud insurance on all of my accounts, and have very clean credit. If I loose a card at an ATM (by forgetting it), or loose a imprinted recipt, I call the card company immediatly. Having a track record of getting new cards whenever something like that happens does wonders when there is something questionable on your statement and you call about it.
Along that vein, a friend of mine recommends reporting your card lost once or twice a year, just to get new cards with different numbers.
Then again, that friend is a little bit more paranoid then I am... He's about as paranoid as you are...
Hey, wait a second, you're name isn't Bryan is it?
Yes, I could. But to charge them each infringment would still require going through a lawyer to try to negotiate (they don't respond to my calls/letters otherwise), which incurs legal fees, which rapidly skyrocket out of control.
Oddly enough, the Best thing I ever found for possible use against them was the DMCA - it has a provision in it stating that removal of a copy-protection feature without permission is criminal, and can lead to civil damages. It also expands on that provision so that every infringing use of the product that has had the feature removed can be treated as an additional infringement, criminal and finable.
Why was this applicable? - Because I made a backdoor. When things first started to get funky, I put in a special username and password combination - 50+ characters each, so almost impossible to crack - that if ever used would cause the software to delete itself. It was specifically designed to only affect the software, not the data, so their 100k user records in the email system would still be there. My software would be gone, but email and the user data would continue as normal. However, when I mentioned to them (late in the game, after they had rolled the software 'live' in multiple locations and were trying to pay me 1/4th the bill) that this existed and that if they don't pay it would go bye-bye, they hired someone to track it down and remove it. By removing this backdoor - a copyright mechanism I had put in place to protect myself - they violated the DMCA. Lovely, eh?
These are all fine ideas, and ones I now use regularly in my business - I don't think I mentioned that despite the BS I went through I ended up deciding to be my own boss.
I think in my case, the bigest things I could of done differently were to have opened my eyes over two things:
1) Company Y had already proven itself, by how it shut down and sold off my employeer X, to be a ruthless and rude entity.
2) Person A had a bad reputation outside the tech departments of X and Y for being an ass, and being untrustworthy.
With #1, I had blinders on and saw the oppourtunity for making money. With #2, I had my blinders on and thought my previous experiences with person A entitled him to 'friend' status.
In both instances I was dead wrong, and If I'd of taken a step back I'd of given the whole thing more thought ahead of time.
sometimes some perspective is all we need, and I've found that keeping your perspective and seeing the whole, big picture can be an important part of keeping a biz afloat.
For example: I recently met with a client who had grand ideas, and wanted all sorts of things done. I wrote a proposal and sent it to them, a short proposal, very informal. considering the features involved, this could be a huge project, so why didn't I write a detailed proposal and send over a work order? Because the (potential) client is a startup, and has money problems. I know that If I spend too much time on proposals for them and they go under I've lost potential business I could of been landing. The (potential) client is bothered by the informal proposal - even though it was informal it has a price range in it. We keep in contact and if they want my services they'll either try to haggle or cut some features. Or, they'll come back 6 months down the line and say 'we're ready to do this now' - which happens more often then you'd believe.
If only perspective came in a can.
If it took the well-publicized OJ trial to shake your faith, then you weren't paying much attention. It's been f-ed up for awhile.
...and doesn't pay. ...and leaves me in the lurch.
Heres a post-OJ example:
I used to work for company X, which was owned by company Y. Company X laid me off. The very next day I get contacted by a big honcho (lets call him A) at X (and Y, CTO for both) to do some contracting work on an unfinished project for Y. I agree and get to work. I work, and I work, and I rack up hours, all billable via the agreement I had with A. I fix all the bugs, I deliver the system.
And then Y doesn't pay.
I find out the software I'd written is about to be taken from its single location and rolled out nationaly by Y, and I get pissed and start bugging the hell out of my contact at Y (person B). Person B tells me that A was never allowed to authorize work for Y, and puts me in contact with a different honcha, person C. Person C claims they are evaluating my software and comparing it against other off-the-shelf systems. I make it very clear to C that this matters not: I agreed to do work for a specific price per hour, and they are using the fruits of my labor. He shrugs it off and says he is not bound by the agreements A made.
Over the next two months, I get the runaround, and then finally C offers to pay less than 1/4rd of the bill as 'a fair liscensing fee based upon our research'. By this time, I had debt upto my eyeballs, but I still had some sense of reason and pride. I made it clear to C that I, not they, get to set liscensing fees for any product I decide to sell, and also stress that liscense was never an issue: I did contract work for them modifying an existing piece of software and fixing bugs, and I expected to get paid. C said the 1/4th was the best he could do. I said I'd go talk to a lawyer.
Here's where the legal system steps in, the above was backstory.
I go talk to a lawyer. With my documentation in hand, he said I have a very strong case - should it ever go to court. But he sugests we do some research and try other tactics (letter from his law firm, etc) first, because of a (to me) glaring problem with how these sort of civil cases go: No matter how big the organization in the wrong is, or how obviously in the wrong they are, it is soley upto the judges discretion as to wether the party in the wrong has to pay legal fees for the complaining party (should the complaining party win). Added to this fact, was that in the conservative city I live in, judges rarely make defendants pay for the complaining party's fees.
This upset me quite a bit.
Then the research into the company came down - just googling stuff I and the lawyer did. Y was run by lawyers and ex lawyers, and no suit brought against them ever made it out of court - they dragged their feet for years on anything they could, and used heavy-handed tactics and threats of countersuits whenever possible to stop suits dead in their tracks by scaring off the person or company making the complaint.
So, here I am, almost a year later, still out TONS of $$ (>20k), and I can't afford to pursue the matter. For every letter they write in response to something my lawyer would send, I'd be out 100-250$ (assuimng it was a short letter my lawyer coudl respond to quickly). If I were to take the case to court, depositions alone could bankrupt me. I entered into a binding agreement with an officer of company Y, and I can't collect squat from it.
Meanwhile, company Y is doing well. They have over 100000 users nationwide using the software I fixed and improved for them.
What it comes down to is this (as the prosecution found out the hard way in the OJ trial): the legal system protects those who have money. If you have tons of money (for lawyers), you can get away with almost anything, or at least make it so painful, expensive and drawn-out for anyone to come after you that it isn't worth it.
Back to the story at hand though: IBM has very deep pockets. But SCO is going to spend eery penny it has on lawyers, and can draw this out for quite awhile. Realisticaly, we wont have an answer to the is-the-gpl-enforcable question for Years.
shouldn't that be 16384 (2^14) or 8192 (2^13) shades of grey? I mean, this is slashdot...
Right now, thier shares are at 14.71USD a piece. Assuming thier stock prices rise as we buy, lets project a 15.5 avg price per share. The question is, and I can't find this offhand, is how many shares of stock we'd need to buy?
I know I'd chip in for 6 shares, and if 100,000 linux users/slashdot readers all bought an avg of two shares, we're looking at 2.5 million shares purchased... Which gets us what % of the company, exactly?
Ok, after digging on finance.yahoo I've found that there are 13.1 million shares outstanding. that means we need to buy 6.56 million shares.
We need 101.68 million dollars.
Ouch. That really throws my numbers out of whack.
Ok then, we need 6.56 million linux users to buy a share a piece, or 1 million users to buy just over 6 shares a piece.
Thats a big organizational challenge... Anyone at FSF or GNU ready to take this on? (the buy-shares forms/donations system and handing over rights for voting with said stock to said org. Realisticly, it might be easier for people to donate cash towards buying stock, not buying it direct, thatway FSF or whomever gets full control of whats bought)
Also: I don't think selling the operation to redhat for a dollar would be the right solution though - thats too politically (in linux circles) charged (to let one distro get thier stuff) - I think thier assests should be liquified, and the results split 50-50 with the fsf getting half to fund dev projects, and 50% going back to the stockholders. Of course, there wouldn't be much left once you consider that many customers with support contracts (and yes, some people have paid sco for support, even pre lawsuit) would want refunds.
Also Also: Yes, I know many people would be against this on moral grounds - 'its the principle of the thing!' - but how many hours and how much money will we waste trying to promote linux against this FUD-crap from SCO? And, what sort of message would a SCO buyout send the world... It'd tell M$ and everyone else that f-ing with Linux is a dangerous proposition, because the community will unite and topple all who oppose us...
just my 3.14 cents...
Uhm, WHY would an organization have a table filled with names that isn't indexed? I agree with in concept, but any organization that has its act together enough to consider using honeytokens and honeypots should also know better than to have crappy db schemas where something like a customer's name isn't indexed...