Okay, truth is Microsoft hired the guy who was the architect of VMS to design WNT. It didn't start with a lawsuit, but there certainly was a legal team hammering down the rights before Digital handed him (Dave Cutler) over.
VMS was pretty much bulletproof by then (it was a *long* road to get there) but back then you could look at the VMS SYSGEN parameters related to the memory model and see the same cryptic names you saw in the depths of the NT Registry, and a couple of the kernel-mode bugs relating to system services and security made it across, too. The RTL became the model for DLL's, (.NET finally brought a lot of the DLL rubbish back into a single RTL, one of the principles that made VMS a fairly easy OS to code for - pick your language, forget the baggage, any language could call any other no sweat). The hardware abstraction layer was buggy, because the Intel instruction set didn't have the isolation the Vax chip set did; VMS had Kernel-Exec-Supervisor-User modes & associated instruction address ranges, which matched the hardware (they were designed together, like IBM's System 360 and OS/360) and the separate modes provided very effective execution isolation -- not a lot of opportunity for the hackers.
The memory model was brilliant and very tunable for different results, but was oriented toward extremely expensive memory -- 512 byte page sizes, for example. But you could get the sucker to run hard if you knew what to tweak. And IMNSHO DCL still beats Perl hands down (ok, a bit of personal preference there).
No, the OS2 stuff belonged to IBM, different team, different compost heap all together. Yes, the urban legend about the character shift is true, too.
It was a drag watching Digital fall apart, first selling off the excellent RDB to Oracle, then dumping AltaVista, then the pathetic attempts to capture the small market with the Rainbow and their own brand of brain-dead PC's, Compaq's ripoff of the clustering technology, then the labels coming off the buildings.
We may garner much intergalactic anthropological interest over the many and varied myths we still retain, the quaint science fiction, and the magnificent primitive smogs. Cool, huh? Way more marketable than religion.
Chuck another prawn on the barbie, come visit us on Outback Earth, y'hear?
Wasn't there a rules engine project some time ago that was trying to codify common sense? It was a few years ago, and I remember they had a few million common-sense statements encoded, but I don't remember many details of the project (like it's name, for starters).
Wouldn't Google and this sort of rather extraordinary database be a great combination? The parent to this post is pretty incredible, when you get down to it.
Of course, but the entire thing is silly, isn't it?
However, since we're in the market for a cheap supply of pure unobtanium, we might as well insist on the variant that thinks it's moving to a lower energy state while the glue cures & contracts, increasing the radial velocity by decreasing the moment arm.
Unobtanium is going to be hard to procure, it's true -- but I'm sure the necessary Impossibilium we're buying to glue it all together will only take a little longer.
Provided you could harness enough energy (big iff) you could probably build the ring a little larger than design radius, then cause the ring to shrink a bit by contraction of the ring material or it's microcrystalline structure. Perhaps by drawing in open lattices (like closing the gaps in a woven or perforated material). Conservation of momentum would increase the spin rate. (Sorry, I'm not up to the calcs with the caffinometer so far into the red (where's my caffinated shower soap...)).
Tux is the quintessential Open Source mascot; not the result of a billion pesetas thrown to a marketing firm for a symbol chosen by a well-paid committee after a subsidised liquid lunch.
Smaller, more active groups is the ticket I think; easier to get a high active-to-passive ratio, easier to find a common purpose.
Technology brings empowerment, can help things happen -- cases in point, Bert Rutan and the people at Scaled Composites, Linus Torvaldis or Tim Berners-Lee -- you don't need to be a government to get highly dramatic things done. Brains and willpower do the work, and communications bring in the funding.
The only real difficulty is weaning those-who-care from the cheap, easy and compelling entertainment offered by that same technology.
I would be glad to help! For a low fee of US$1.00 per file, you may e-mail me the file and I will send you a zip file with your file in it! Aren't I nice?
Okay, you're on. I not only read it, but I built one. 3mm brass sheet, 10 degree resolution, but only one climate plate so far (the SCA has it's share of geeks -- well known fact).
It's a little primitive, but out of the box it boots faster than my Palm Pilot. Slightly less accurate than my vintage ACME digital watch, though, and a lot more waterproof. You do have to keep ahead of the verdigris though...
Actually, one of the big obstacles to using hydrogen as a fuel is that it ISN'T very easily transportable.
It leaks, too, quite easily. Leetle tiny molecule, H2.
Wasn't there a method bandied about some years back about cracking H2 from ammonia using spare ergs from the back end of a nuclear power plant? I know, fission-based power is dirty (yadda yadda) but some day we'll be able to press those little puppies together in Mr.Fusion and we may yet find an alternative to burning all our valuable hydrocarbon chemical manufacturing feedstocks. Looking forward to it, I am.
Some did... the microwave at Logisticon (Sunnyvale, late 70's) was much higher powered. That is, until that particular experiment was tried. Scrambled egg everywhere, door blown across the room and seriously warped.
I'm glad I saw that before I tried drying the stray puppy.
If they're found, they'll be lucky if they only get sued.
Thought -- Imagine if they end up in jail; considering how many inmates' only contact with the outside world is via the Internet, what would be the inside lifespan of a convicted spammer?
Filled out a few thousand Fortran coding forms (remember them?) for a satellite project with a Pentel mechanical pencil, one of the early ones -- had a crosshatched tungsten barrel, and it was very heavy. Heavy barrel made for smooth writing, important after that fourth or fifth pot of coffee, sort of like a heavy flywheel on a Moto Guzzi.
Sigh. No, I am not assuming Open Office is not scriptable. Yes, I speak multiple languages, picked up a few in 20 years as a systems programmer. Yes, I prefer Python to VBA. Yes, I agree vendor lock-in is bad, having spent the last ten years as an architect and development manager.
By the way, when you refer to vendor lock-in do you mean some vendors and not others? Would you include small vendors who write mods to open source apps? It's quite easy to be locked in by them, too -- been there, done that, ate the shirt, that problem has a lineage dating back to the days when software was distributed as Cobol source. Everyone could read it, didn't help because nobody wanted the job of trying to save somebody else's crap code.
I was referring to certain macros that revector the "save as" options into SAN instead of allowing an army of clerks to write their own filing system. Distasteful to me, godsend to large organisations. And I'm not speaking for any CIO in particular, so I'm afraid I can't give you a single target upon which to vent your wrath. Customers of document management systems are usually organisations constrained by law to have their docs carefully controlled, and that means governments, so you're a stockholder whether you want to be or not.
Actually, large organisations have more clout with Microsoft than you might believe, and are not as locked in as you think. Think of Telstra, or the government of India.
Ah, such insight. I can assure you that the bayesean modifier of path-darkness is given very little weight against the cost of support in a large, enterprise-level COTS-based solution, although expensive-ness certainly is.
I am not talking about a few blocks of VBA script hacked in a back room, I am talking about multi-million dollar applications from vendors who build the support component (i.e. paying programmer's wages for upgrade work) into the solution up-front, sold to customers for whom spending 40% or more on apps support is considered a bargain compared to the large multiplier placed against not having a solution in place.
In the overall context of a commercial solution, the spreadsheet rules. Lots of effort goes into examining the pros and cons of each selection under the heading of due-diligence. This means paying for the opinions of engineers and strategists to weigh the value of this aspect of support costs. Scripting a MS app with VBA or adding a few macros is changing a small piece of a pre-existing package in a fairly non-invasive way, as opposed to an OS-approach, which allows change to all of it. Is this necessarily the worst way to achieve the lowest total cost of ownership for a package? On the other hand, how many wheels do they have to pay to re-invent before an open-source variant becomes no longer economically viable? The market will decide, as you imply -- e.g. as soon as a better, cheaper, well-packaged Open Source document managment system appears on the market you can be sure it will get equal air play in the bid responses. In that context, whether or not Open Source efforts can duplicate the functionality is commercially irrelevant. I know of some deep and complex DMS' that are entirely J2EE except for the MS-doc & macro bits, which are considered presentation-layer only & not worth the effort to worry about. Does this mean they're pandering to Microsoft? I don't think so.
I am not an apologist for Microsoft, neither am I a believer in the sanctity of Open Source. Code is code, and this is a competitive world. Get over it.
There are lots of excellent reasons for moving to open source. Emotional appeal isn't one of them.
Not all macros are virii. There are some very large, very expensive 3rd party document management suites that rely on them, and without those products thousands of sysadmins are condemned to explain to millions of public servants why they shouldn't store essential documents on their C drive.
Slashdot Mirror
VMS was pretty much bulletproof by then (it was a *long* road to get there) but back then you could look at the VMS SYSGEN parameters related to the memory model and see the same cryptic names you saw in the depths of the NT Registry, and a couple of the kernel-mode bugs relating to system services and security made it across, too. The RTL became the model for DLL's, (.NET finally brought a lot of the DLL rubbish back into a single RTL, one of the principles that made VMS a fairly easy OS to code for - pick your language, forget the baggage, any language could call any other no sweat). The hardware abstraction layer was buggy, because the Intel instruction set didn't have the isolation the Vax chip set did; VMS had Kernel-Exec-Supervisor-User modes & associated instruction address ranges, which matched the hardware (they were designed together, like IBM's System 360 and OS/360) and the separate modes provided very effective execution isolation -- not a lot of opportunity for the hackers.
The memory model was brilliant and very tunable for different results, but was oriented toward extremely expensive memory -- 512 byte page sizes, for example. But you could get the sucker to run hard if you knew what to tweak. And IMNSHO DCL still beats Perl hands down (ok, a bit of personal preference there).
No, the OS2 stuff belonged to IBM, different team, different compost heap all together. Yes, the urban legend about the character shift is true, too.
It was a drag watching Digital fall apart, first selling off the excellent RDB to Oracle, then dumping AltaVista, then the pathetic attempts to capture the small market with the Rainbow and their own brand of brain-dead PC's, Compaq's ripoff of the clustering technology, then the labels coming off the buildings.
The icon is dead, long live the icons!
-- old, sad, unemployed systems programmer
Chuck another prawn on the barbie, come visit us on Outback Earth, y'hear?
Wouldn't Google and this sort of rather extraordinary database be a great combination? The parent to this post is pretty incredible, when you get down to it.
"Will Code For Food"
However, since we're in the market for a cheap supply of pure unobtanium, we might as well insist on the variant that thinks it's moving to a lower energy state while the glue cures & contracts, increasing the radial velocity by decreasing the moment arm.
Unobtanium is going to be hard to procure, it's true -- but I'm sure the necessary Impossibilium we're buying to glue it all together will only take a little longer.
Provided you could harness enough energy (big iff) you could probably build the ring a little larger than design radius, then cause the ring to shrink a bit by contraction of the ring material or it's microcrystalline structure. Perhaps by drawing in open lattices (like closing the gaps in a woven or perforated material). Conservation of momentum would increase the spin rate. (Sorry, I'm not up to the calcs with the caffinometer so far into the red (where's my caffinated shower soap...)).
In Tasmania, there's a building marked "Penguin Fire Station". Anybody there care to make a Tux pose?
Do you know why they called it the Pentium instead of the 586? They added 100 to 486 and got 585.9999897332413...
I once read that the recoil of the cannon matched the thrust produced by the engines. Scary!
Read "Karma" by Arsen Darney for an excellent treatment of this subject.
Technology brings empowerment, can help things happen -- cases in point, Bert Rutan and the people at Scaled Composites, Linus Torvaldis or Tim Berners-Lee -- you don't need to be a government to get highly dramatic things done. Brains and willpower do the work, and communications bring in the funding.
The only real difficulty is weaning those-who-care from the cheap, easy and compelling entertainment offered by that same technology.
Send your file to not@clue.com
It's a little primitive, but out of the box it boots faster than my Palm Pilot. Slightly less accurate than my vintage ACME digital watch, though, and a lot more waterproof. You do have to keep ahead of the verdigris though...
It leaks, too, quite easily. Leetle tiny molecule, H2.
Wasn't there a method bandied about some years back about cracking H2 from ammonia using spare ergs from the back end of a nuclear power plant? I know, fission-based power is dirty (yadda yadda) but some day we'll be able to press those little puppies together in Mr.Fusion and we may yet find an alternative to burning all our valuable hydrocarbon chemical manufacturing feedstocks. Looking forward to it, I am.
I'm glad I saw that before I tried drying the stray puppy.
Thought -- Imagine if they end up in jail; considering how many inmates' only contact with the outside world is via the Internet, what would be the inside lifespan of a convicted spammer?
It's dead now, but I can't bear to throw it away.
Not all that ripples and is pretty is high-tech.
I'm a mouse potato. Could I be identified and logged in by my hand pressure on the mouse?
By the way, when you refer to vendor lock-in do you mean some vendors and not others? Would you include small vendors who write mods to open source apps? It's quite easy to be locked in by them, too -- been there, done that, ate the shirt, that problem has a lineage dating back to the days when software was distributed as Cobol source. Everyone could read it, didn't help because nobody wanted the job of trying to save somebody else's crap code.
I was referring to certain macros that revector the "save as" options into SAN instead of allowing an army of clerks to write their own filing system. Distasteful to me, godsend to large organisations. And I'm not speaking for any CIO in particular, so I'm afraid I can't give you a single target upon which to vent your wrath. Customers of document management systems are usually organisations constrained by law to have their docs carefully controlled, and that means governments, so you're a stockholder whether you want to be or not.
Actually, large organisations have more clout with Microsoft than you might believe, and are not as locked in as you think. Think of Telstra, or the government of India.
I am not talking about a few blocks of VBA script hacked in a back room, I am talking about multi-million dollar applications from vendors who build the support component (i.e. paying programmer's wages for upgrade work) into the solution up-front, sold to customers for whom spending 40% or more on apps support is considered a bargain compared to the large multiplier placed against not having a solution in place.
In the overall context of a commercial solution, the spreadsheet rules. Lots of effort goes into examining the pros and cons of each selection under the heading of due-diligence. This means paying for the opinions of engineers and strategists to weigh the value of this aspect of support costs. Scripting a MS app with VBA or adding a few macros is changing a small piece of a pre-existing package in a fairly non-invasive way, as opposed to an OS-approach, which allows change to all of it. Is this necessarily the worst way to achieve the lowest total cost of ownership for a package? On the other hand, how many wheels do they have to pay to re-invent before an open-source variant becomes no longer economically viable? The market will decide, as you imply -- e.g. as soon as a better, cheaper, well-packaged Open Source document managment system appears on the market you can be sure it will get equal air play in the bid responses. In that context, whether or not Open Source efforts can duplicate the functionality is commercially irrelevant. I know of some deep and complex DMS' that are entirely J2EE except for the MS-doc & macro bits, which are considered presentation-layer only & not worth the effort to worry about. Does this mean they're pandering to Microsoft? I don't think so.
I am not an apologist for Microsoft, neither am I a believer in the sanctity of Open Source. Code is code, and this is a competitive world. Get over it.
There are lots of excellent reasons for moving to open source. Emotional appeal isn't one of them.
Not all macros are virii. There are some very large, very expensive 3rd party document management suites that rely on them, and without those products thousands of sysadmins are condemned to explain to millions of public servants why they shouldn't store essential documents on their C drive.