The LGPL was originally called the "Library" GPL, and then later on was backronymed to the "Lesser" GPL by RMS.
Its purpose is to allow closed-source applications to use open-source libraries without becoming "infected" by copyleft source publication requirements.
So if you write a C program that links against the LGPL-licenced glibc, you are not forced to adopt copyleft for your program.
If, however, you modify the actual library code, you are required to publish source to your changes.
If WINE were to be LGPL-ed, you could write a program that would run on both Windows and [any x86 OS with a WINE port] by linking against WINE. Your program could be licenced however you wish, as the act of linking against an LGPL-ed resource does not incurr the responsibility of copyleft.
However, if you discovered that you really needed the as-yet WINE-unimplemented Windows API call foo(), and then did the work to implement the foo() call in WINE, the LGPL would force you to release the source to those changes to the public.
This is, IMHO, a REALLY REALLY good idea. The nature of the WINE project is that once a certain core of the API is ported, the rest of the work is really very modular, but very broad. Certain companies have been completing work on various APIs needed to get their pet projects working (like core gaming APIs) and then refusing to turn these changes back in to the core WINE project for "competitive" reasons - ie, if they have the only working version of these core APIs, then only they can publish software that uses these APIs (until someone re-does the port work and releases the API in a Free manner)
Result: uncecessary duplication of effort, and bad feelings all 'round.
I don't contribute to WINE, so I don't get a vote (which is as it should be) but I'm sure as hell cheering for the LGPL people.:)
Lemme ask you this: it seems that with the previous generation of 3D cards, the technology had reached the point where any game with a reasonable game engine could be run at 1024X768x32bit with all the detail goodies turned on at framerates that were completely playable.
(Perhaps this is a mistaken assumption?)
If so, then what does this card bring to the table from a game designer/coder's perspective?
If there's no point in driving a Quake3 style engine any faster (because it's already fast enough) then what will you be able to do with this new hardware that you couldn't do with older stuff?
Or to rephrase, what hardware feature do you most wish was availible on the current generation of 3D cards, and does this new card have that feature?
I think - I THINK - that Miquel's goal is to out-Microsoft Microsoft, by beating MS to the punch on.NET.
If you can make.NET as widespread as HTTP _before_ MS can get a signifigant foothold, then you have a certain element of control over the Beast From Redmond.
But that's a really dangerous game to be playing, methinks.
So there you are, some Big Company, with this nifty Linux cluster running Oracle, saving money on the OS and servers, but being bled white by Oracle's per-transaction fee structure.
And then somebody discovers this "PostgresSQL" thing....
The program I really cut my 3D teeth on was a program called Imagine. Imagine was Blender-like, in so far that it had an interface and a bunch of concepts that you really had to know before you could make good use of it; stuff that wasn't immediately obvious.
I did up a fully-detailed, to-scale Klingon D7 battlecruiser using it that was indistinguisable from the "real thing". It took about 3 months of after-work work. (Why is it that newbie 3D modelers always do Star Trek models?:)
Then, on the advice of a friend of mine, I got ahold of Lightwave.
In some ways it was more primitive than Imagine. No procedural textures to speak of, and no free-form bitmap textures (so "decals" were a serious PITA) But Oh-My-Lord it was SO SIMPLE to both make models and stage scenes.
Lightwave uses film paradigms for a lot of its user interface. It stops being a "3D tool" and starts being something more like an interactive film studio. With Imagine, 3D work was like being a cross between a computer programmer and an engineer. With Lightwave, it was more like being a film director.
Both my throughput and my quality went way, WAY up when I switched to Lightwave.
Imagine was capable of generating prettier stills (the renderer was more powerful) but Lightwave made better MOVIES - because the interface was more conducive to the real task at hand.
I don't doubt that people can do good work with Blender, and that with time, one can become proficiant in its use. But with Lightwave, that proficiancy step wasn't needed. The interface was transparent.
I guess you really have to try it and see for yourself.
The high-tech "telepresence" stuff is great - when it is working. The US Army has had the luxury of fighting against opponents that have not had either the equipment or the desire (or perhaps failed to see the need to) disrupt the communications networks.
Had Iraq been capable of jamming communications, locating and targeting transmitters, or throwing up a better deception plan, the Yanks might have had a tougher time.
Not to mention that sometimes the damned stuff just breaks down. Radios will do funky things out in the bush.
In my armoured recce troup, we used to practice communicating with hand signals and lights all the time. We assumed that any radio transmission longer than about a second would give away your exact position, and that any transmission _at all_ would give away that you were there, somewhere.
The high-tech stuff is really, really nice when you can use it, but don't assume that it will always be availible to you.
A few years ago, there used to be this NATO-wide tank gunnery competition called the Canadian Army Trophy. Part of it involved a move-and-shoot range where pop-up targets were shot at.
When the Yanks first brought out the M1 Abrams, its thermal sight was better than anything anybody had fielded before. It was so good that it picked up the heat from the motors that lifted the pop-up targets - and the Yanks cleaned house.
The next year, the motors on the targets were insulated, and a large number of fake-motor-heat-sources were seeded on the range. Suddenly, the fancy-shamcy thermal sights were feeding _disinformation_ - and the American performance dropped off considerably.:)
The bottom line is that technology is no substitute for sound procedures. It may augment them, but it cannot replace them.
Assuming the kernal can get to the point where it can be successfully modularized to the point where changes in a given module are localized to the scope of that module (and let's be honest, that's a Mighty Big "If") then Linus doesn't have to give up the level of fine-grained control he likes. He can focus his attention on individual trouble spots, secure in the knowledge that the "mini-Linusen" are holding the fort while his attention is elsewhere.
Successful Generals lead from the front. Really successful Generals (Patton!) are not only at the front, but have the ability to select the portion of the front which most needs them at that particular instant.
WWI taught that lesson. Generals who sit behind the front lines waiting for information to reach them are too far away in both distance and TIME to be able to affect the outcome of the battle.
But I agree that the sticky part is getting this structure set up. Once in place, it works, and it offers the best chance of success from all points of view (quality control for Linus, patch responsiveness for individual coders)
I think the first step is for everyone working on the kernel to - on their own - determine who their subsection maintainer is, and only submit patches to them: this is another principle of leadership called "chain of command" and frees up Linus to do more "battalion" orders and less "Bob" orders.
But it has to be done SOON. I think it's clear that Linus IS losing control of the kernel - no WAY should that bogus "preemptable kernel patch" that keeps circulating have the life that it does - and Linus' control has been key to the quality control that has worked so well to date. The less influence Linus has, the more willing people are to put up with bogus and wrong code in the kernel, and the more quality suffers.
This is a basic principle of leadership/management that the Army calls "span of control".
It has been proven (through all kinds of research) that any one person has difficulty controlling much more than 10 individuals, with 8 being a practical maximum.
This is why military units are organized in strict tree structures - at no point does any one leader exceed his span of control.
So an 8-man infantry section is controlled by a Section Commander (which can be further broken down to 2 4-man fire teams, but rarely). 4 sections report to a Platoon Commander. 4 Platoons report to a Company Commander. 4 Companies report to a Battalion Commander, and so on through Regiments, Brigades, Divsions, and so forth.
In this manner, the several thousand men in a large formation like a Division can be commanded by one man.
Further aiding the General in charge of the Division is that each sub-unit is granted a certain amount of autonomy within certain boundries. As you go up the chain, orders become more general (2 Battalion is to attack objective Oscar) and as you go down the chain, orders become more specific (Bob, you're carrying the machine gun)
While Linux development does not require this level of structure (and indeed, would probably suffer if this many authority-resistant cats were forced into such a regimented structure) the basic principle of "span of control" is still applicable.
The obvious correct solution is to modularize the kernel into subsections with clearly defined areas of responsibility, with a "mini-Linus" (who Linus trusts) granted control over each module.
Not suprisingly, this is what Linus has suggested.
The trick (and the catch) is two-fold. Firstly, you have to find people who are both responsible enough to be able to act as a "mini-Linus" without dropping the ball (which means Linus has to trust them too) and secondly, the kernel has to be modular enough so that changes in Module A do not step on or otherwise negatively impact code in Module B.
That's a hard row to hoe.
What I find a little distressing is that there appears to be a whiff of revolution in the air; of people talking about overthrowing Linus (and Alan Cox seems to be among them) This will not solve the problem. It is not that **Linus** doesn't scale, but rather that any single maintainer doesn't scale. Any revolution will just be "same shit, different name"
CVS doesn't solve the problem either, as anyone with commit privilages can jsut dump anything they want in there, and it becomes canon. Linus' function as a shit filter is very very important. every patch *must* be audited by someone with very high standards before it goes in.
I hope the crew rallies around Linus - as usual, he's right.
Maybe it's just me, but I don't understand the concept of streaming video to a phone.
For me, the ultimate cellphone would have the following:
1) A decent phone, with decent battery life, reception, and audio clarity (how many convergence devices overlook the fact that the primary purpose of the damn thing is a telephone?)
2) Palm Pilot-like functionality, with emphasis on providing space/ability for uploading apps of my own as well as the canned apps.
3) A GPS, with detailed street maps and wayfinding ability built in.
4) Integrate this stuff as tightly as possible, and keep as much of the data local as possible.
I can see, for example, having the complete North American phone directory on the phone, so I can look up numbers without hitting the network. Tie this into the GPS, and now I can do stuff like "get me the phone number for the house I'm standing in front of right now" or "Let me search the yellow pages for [whatever] and now that I've made a selection, give me driving directions to get there"
Or allow phones to transport GPS data on voice connections, and now I can get a map of where whoever it is I'm talking to is - geographic caller ID.
Network access is all well and good, but phones are phones first and foremost. Build in apps that support the "phone" part (things like searchable directories) and the "mobile" part (with the GPS) and now you're talking!
The Kyocera Smartphone (which is a Palm) seems close, and will probably be my next phone, but I'm still looking forward to a well-done phone+GPS combo.
There seems to be a fair number of people actually employed in these fields responding to this article, so this seems like a good place to ask this question.
It appears to me that we're pretty far along when it comes to the biology of sickness-by-infection, where an illness is caused by being attacked by other organisms. There's a long way to go, but it seems to an outsider that most of the fundimental processes are understood, and the lion's share of what remains is of the nature of "find germ, study germ, develop treatment that kills germ without killing host"
But it also seems that we're not very far along when it comes to understanding sickness-though-internal-breakdown, where actual body processes either fail to function or function abnormally.
It strikes me that understanding how human genetics really work is the key to all survival. If we knew how every gene and every internal process functioned, then we cound re-engineer our own genome to fix problems. Eliminate cancer, eliminate AGING, and so forth.
It would thus suggest to me that working on deciphering the human genome is the most important problem in human biology in history, and perhaps even the most important problem EVER.
We should have huge amounts of public money poured into this problem, with all results made public, and all information shared.
Would you agree? Have I made any erronious assumptions?
DG
For what it's worth, I bought a Loki game today
on
Last Word on Loki
·
· Score: 0, Flamebait
I've never been much of a fan of the concept of software for sale, as product. To my mind, writing software is a SERVICE, with the end result free for all to use.
So Loki was providing the service of selecting games of good quality, and porting them to Linux where I could use them with my OS of choice.
As such, I wasn't a Loki *customer* so much as I was a Loki *sponsor* or *patron*.
And accordingly, I paid for every Loki game I have. RT and M2 I bought direct from Loki. D3, SoF, and Q3 I bought from the local EB.
Myth 2 I bought yesterday, when I heard they were closing. If I had more ready cash, I would have picked up the rest of their ouvre.
Oh well. Farewell Loki, you did great work. I did what little I could to support your efforts.
The Motec ECU is a pretty serious system. It gets used in LeMans LMP cars, WRC cars, and the like.
http://www.motec.com to see. You can download evaluation copies of the programs to play with, if you like.
I have a similar system in my race car, made by GEMS (http://www.gems.co.uk) and it's awesome. Computerized fuel injection has revolutionized racing. You wouldn't believe the crap that thing can do.
Of course, the downside to having complete and total control over the engine is that you have to do all the calibration yourself. After having programmed the idle control/antistall stepper motor, and the accursed warmup maps, I have a huge amount of resepct for the OEM calibration guys.
Getting a car to idle smoothly under all weather conditions is tough work.
In any case, if Romero's Ferrari has a Motec system in it, then that ain't your garden variety Ferarri.
I'm kinda-sorta thinking about picking up a new processor/MB/case, and I know I want an AMD chip, but I've yet to find any easy to understand table that clearly deliniates what the differences are between the various Athlons.
Anybody got a source for such a beast? Or even feel like explaining it here?
I live in a Windows-free world, so I guess I'm a Linux gamer.
The downside is that there's not much in the way of games. The upside is that those that are there are pretty good.
My "game machine" is a P1-233 with a PCI GF-MX200 in it. Plays Q3 just fine. And I've been making my way through the Loki ouvre and having a ball with it.
I'm almost finished SOF, and then there's Descent 3 waiting in the wings.
Am I typical of the bleeding-edge, overclocked, 3000 FPS gamer that Windows seems to attract? Hell no. But I AM using Linux as my gaming OS, and I AM having fun with it, and I'm using a machine that's over 5 years old.
Well, there's a major difference between this technology and a lie detector.
Humans are not very good at detecting lies, and can in fact get very good at certain, specialized forms of deception, even in the face of equally speciallized deception-detectors. (I'm thinking here of professional poker players, who make their living on deceiving and detecting deception, in a very specialized manner)
So a lie detector is an attempt to augment an ability.
But humans are VERY good at facial recognition, much better than any machine is. The limitation here that the machine is trying to address is one of storage capacity (memory) and speed, not ability per sae.
I make the assumption here that the interface would display the picture of the person that it thinks the person in question matches alongside the picture from its camera (or print one for the officer to carry with him for an in-person comparison) so that the vastly superior human facial recognition abilities could be brought into play.
So, really, I don't see any reason to get upset. It's no different than if an officer saw your photo on a "wanted" poster in the post office, and then made the match from memory, only now the "memory" is much larger.
When an arrest warrent is cut, your driver's licence is flagged in the DMV computers. If you get pulled over (for any reason) and your licence is checked against this database, they get you. How is this any different?
As I understood it, the issue with facial recognition was the possibility of false positives; ie, I'm just trying to watch the SuperBowl. the FR system tags me as a known terrorist (incorrectly:) and the next thing I know I'm being dragged off to the can for some serious interregation (and not only unjustly tramatized, but I miss the game too)
But from the ACLU's press release, there was always a human step in the process, where a real live human being would examine each purported match before anybody got dragged off anywhere.
As such, all the face recognition software is is a _filter_, cutting down on the number of people a human agent must examine. Where's the problem?
After all, law enforcement officers have placed themselves in public places, looking for people they knew, for probably as long as there have been law enforcement officers.
A friend of mine was a sergent in the British Army, and he did a few tours in Northern Ireland. Part of his training was memorizing the faces of a large number of known IRA "players" (and apparently the IRA did the same thing with British soldiers' faces)
Aside from the usual privacy rants, there's another chilling effect here.
What if what I'm looking for is the info that is not normally availible to me through my local-access media.
Maybe I WANT info in not-my-language. Maybe I WANT the teasers from the other market. Maybe I WANT the spoilers. Etc. Etc Etc.
Something a lot of old-media people haven't seemed to grasp yet is that the Internet-as-media is an INDIVIDUAL DEMAND system, or if you prefer, a "pull" instead of a "push"
The more you try to force push-style paradiems on me, the less useful your outlet becomes.
The overriding consideration for any Internet media is "make it as easy to find what I'm looking for as possible" with things like clear indicies, functional and accurate searches, and so on. Don't make assumptions about what it is I want, and push it at me.
Here's the best analogy I can think of: A "push" media, like TV or Radio, can be used by a group. Assuming the audience has similar interests (and that the TV/speakers are large enough) a TV show can be watched alone, in a small group, or even a football stadium full of people.
Can you imagine a football stadium of people, watching the screen as one guy surfs the site?
Using the Internet is a personal and individual experience. It's a library, not a program. Trying to force broadcast-media concepts on it just doesn't work.
Part of the problem with revolutionary ideas is that there will forever be people who just don't understand; who cannot grasp the new concept, and who will attempt to recast it in terms they _do_ understand - only to miss the whole point all over again.
Such is Linux and Windows.
Windows is a PRODUCT. It is for sale, complete with sales reps, marketing budgets, and an army of lawyers to try and enforce the alien concept of "product scarcity" on a digital entity.
As a "product", it is subject to the rules of the market; the ebb and tide of economics.
Linux is NOT A PRODUCT, it is something else entirely. It's part common property, part social movement, part fun little hobby, and part irresistable juggernaut. In fact, I don't yet think there exists an English word that adequately expresses what Linux is. What do you call a tool that is owned by nobody, is constructed and maintained by many, and freely availible to all?
There are companies that produce products BASED on Linux, and these companies often subsidize contributions back to the greater whole, but these companies are no more "Linux" than Frito Lay or Doritos are "corn".
As long as the source code remains availible, and as long as it continues to function on existing hardware. Linux cannot "fail".
This is what the article author does not understand, and why Linux is so dangerous to Microsoft's monopoly. Linux, in some form, will _always_ be there. It will _never_ go away. It cannot be bought, swept under the rug, supressed, or otherwise made to go away.
The best you can do is to write code that does the same job, better - but we're seeing that Linux can develop every bit as fast (and oftentimes faster) as any proprietary product. No company, no matter how big, can muster a workforce as large as that actively working on Linux. Given enough time, Linux will eventually catch you and beat you on quality.
Bill Gates is often given credit for "inventing" the concept of software-for-sale, where previously, software was shared amongst users and developers free of cost. Well then, Bill has made his own bed. Linux is the ultimate competitor; the anti-Microsoft incarnate.
And a welcome CORRECTION, bringing software back from the artificial world of "product", to the real world of "service" where it originated and BELONGS.
Wouldn't clustering be a way to circumvent the law in the first place?
The problems that the law was intended to make difficult to solve (nuclear weapons simulation, aero flow analysis, cryptography, and so on) are, as far as I can tell, problems that can can be attacked in parallel, and so are good applications for clusters to tackle.
Well then, if the restriction prevented the export of any computer faster than x, couldn't a cluster of n export-legal computers of speed y (y less-than x ) produce a total throughput power Y (Y greater-than x)?
And for smaller values of y, substitute larger values of n to gain the same net power Y.
So really, I would think that clustering technology rendered (heh) the restriction moot a long time ago.
Wouldn't clustering be a way to circumvent the law in the first place?
The problems that the law was intended to make difficult to solve (nuclear weapons simulation, aero flow analysis, cryptography, and so on) are, as far as I can tell, problems that can can be attacked in parallel, and so are good applications for clusters to tackle.
Well then, if the restriction prevented the export of any computer faster than x, couldn't a cluster of n export-legal computers of speed y (y x ?
And for smaller values of y, substitute larger values of n to gain the same net power Y.
So really, I would think that clustering technology rendered (heh) the restriction moot a long time ago.
Slashdot Mirror
The LGPL was originally called the "Library" GPL, and then later on was backronymed to the "Lesser" GPL by RMS.
:)
Its purpose is to allow closed-source applications to use open-source libraries without becoming "infected" by copyleft source publication requirements.
So if you write a C program that links against the LGPL-licenced glibc, you are not forced to adopt copyleft for your program.
If, however, you modify the actual library code, you are required to publish source to your changes.
If WINE were to be LGPL-ed, you could write a program that would run on both Windows and [any x86 OS with a WINE port] by linking against WINE. Your program could be licenced however you wish, as the act of linking against an LGPL-ed resource does not incurr the responsibility of copyleft.
However, if you discovered that you really needed the as-yet WINE-unimplemented Windows API call foo(), and then did the work to implement the foo() call in WINE, the LGPL would force you to release the source to those changes to the public.
This is, IMHO, a REALLY REALLY good idea. The nature of the WINE project is that once a certain core of the API is ported, the rest of the work is really very modular, but very broad. Certain companies have been completing work on various APIs needed to get their pet projects working (like core gaming APIs) and then refusing to turn these changes back in to the core WINE project for "competitive" reasons - ie, if they have the only working version of these core APIs, then only they can publish software that uses these APIs (until someone re-does the port work and releases the API in a Free manner)
Result: uncecessary duplication of effort, and bad feelings all 'round.
I don't contribute to WINE, so I don't get a vote (which is as it should be) but I'm sure as hell cheering for the LGPL people.
DG
I know you're out there John. :)
Lemme ask you this: it seems that with the previous generation of 3D cards, the technology had reached the point where any game with a reasonable game engine could be run at 1024X768x32bit with all the detail goodies turned on at framerates that were completely playable.
(Perhaps this is a mistaken assumption?)
If so, then what does this card bring to the table from a game designer/coder's perspective?
If there's no point in driving a Quake3 style engine any faster (because it's already fast enough) then what will you be able to do with this new hardware that you couldn't do with older stuff?
Or to rephrase, what hardware feature do you most wish was availible on the current generation of 3D cards, and does this new card have that feature?
DG
I don't know if it has ever been done on a helicopter per sae, but there have been ejection seats that went out the _bottom_ of the aircraft.
:)
I think the F102 (or one of the other supersonic, large-single-vertical-fin fighters) had one of these.
Just don't pull the handle during taxi.
DG
http://streetmodified.org/books.html
I think - I THINK - that Miquel's goal is to out-Microsoft Microsoft, by beating MS to the punch on .NET.
.NET as widespread as HTTP _before_ MS can get a signifigant foothold, then you have a certain element of control over the Beast From Redmond.
If you can make
But that's a really dangerous game to be playing, methinks.
Miquel scares me sometimes.
DG
http://streetmodified.org/books.html
So there you are, some Big Company, with this nifty Linux cluster running Oracle, saving money on the OS and servers, but being bled white by Oracle's per-transaction fee structure.
And then somebody discovers this "PostgresSQL" thing....
Payback's a bitch, innit?
DG
http://autocross.dsm.org/books.html
The program I really cut my 3D teeth on was a program called Imagine. Imagine was Blender-like, in so far that it had an interface and a bunch of concepts that you really had to know before you could make good use of it; stuff that wasn't immediately obvious.
:)
I did up a fully-detailed, to-scale Klingon D7 battlecruiser using it that was indistinguisable from the "real thing". It took about 3 months of after-work work. (Why is it that newbie 3D modelers always do Star Trek models?
Then, on the advice of a friend of mine, I got ahold of Lightwave.
In some ways it was more primitive than Imagine. No procedural textures to speak of, and no free-form bitmap textures (so "decals" were a serious PITA) But Oh-My-Lord it was SO SIMPLE to both make models and stage scenes.
Lightwave uses film paradigms for a lot of its user interface. It stops being a "3D tool" and starts being something more like an interactive film studio. With Imagine, 3D work was like being a cross between a computer programmer and an engineer. With Lightwave, it was more like being a film director.
Both my throughput and my quality went way, WAY up when I switched to Lightwave.
Imagine was capable of generating prettier stills (the renderer was more powerful) but Lightwave made better MOVIES - because the interface was more conducive to the real task at hand.
I don't doubt that people can do good work with Blender, and that with time, one can become proficiant in its use. But with Lightwave, that proficiancy step wasn't needed. The interface was transparent.
I guess you really have to try it and see for yourself.
DG
I've tried to get something useful out of Blender on a couple of occasions, but the interface is just too obtuse.
I've been able to figure out Sculpt 3D, Turbo Silver, Imagine, Lightwave, and VariCAD without too much drama, but Blender defeats me.
I really, really miss Lightwave. It had the perfect mix of power and ease-of-use.
DG
The high-tech "telepresence" stuff is great - when it is working. The US Army has had the luxury of fighting against opponents that have not had either the equipment or the desire (or perhaps failed to see the need to) disrupt the communications networks.
:)
Had Iraq been capable of jamming communications, locating and targeting transmitters, or throwing up a better deception plan, the Yanks might have had a tougher time.
Not to mention that sometimes the damned stuff just breaks down. Radios will do funky things out in the bush.
In my armoured recce troup, we used to practice communicating with hand signals and lights all the time. We assumed that any radio transmission longer than about a second would give away your exact position, and that any transmission _at all_ would give away that you were there, somewhere.
The high-tech stuff is really, really nice when you can use it, but don't assume that it will always be availible to you.
A few years ago, there used to be this NATO-wide tank gunnery competition called the Canadian Army Trophy. Part of it involved a move-and-shoot range where pop-up targets were shot at.
When the Yanks first brought out the M1 Abrams, its thermal sight was better than anything anybody had fielded before. It was so good that it picked up the heat from the motors that lifted the pop-up targets - and the Yanks cleaned house.
The next year, the motors on the targets were insulated, and a large number of fake-motor-heat-sources were seeded on the range. Suddenly, the fancy-shamcy thermal sights were feeding _disinformation_ - and the American performance dropped off considerably.
The bottom line is that technology is no substitute for sound procedures. It may augment them, but it cannot replace them.
DG
Assuming the kernal can get to the point where it can be successfully modularized to the point where changes in a given module are localized to the scope of that module (and let's be honest, that's a Mighty Big "If") then Linus doesn't have to give up the level of fine-grained control he likes. He can focus his attention on individual trouble spots, secure in the knowledge that the "mini-Linusen" are holding the fort while his attention is elsewhere.
Successful Generals lead from the front. Really successful Generals (Patton!) are not only at the front, but have the ability to select the portion of the front which most needs them at that particular instant.
WWI taught that lesson. Generals who sit behind the front lines waiting for information to reach them are too far away in both distance and TIME to be able to affect the outcome of the battle.
But I agree that the sticky part is getting this structure set up. Once in place, it works, and it offers the best chance of success from all points of view (quality control for Linus, patch responsiveness for individual coders)
I think the first step is for everyone working on the kernel to - on their own - determine who their subsection maintainer is, and only submit patches to them: this is another principle of leadership called "chain of command" and frees up Linus to do more "battalion" orders and less "Bob" orders.
But it has to be done SOON. I think it's clear that Linus IS losing control of the kernel - no WAY should that bogus "preemptable kernel patch" that keeps circulating have the life that it does - and Linus' control has been key to the quality control that has worked so well to date. The less influence Linus has, the more willing people are to put up with bogus and wrong code in the kernel, and the more quality suffers.
DG
This is a basic principle of leadership/management that the Army calls "span of control".
It has been proven (through all kinds of research) that any one person has difficulty controlling much more than 10 individuals, with 8 being a practical maximum.
This is why military units are organized in strict tree structures - at no point does any one leader exceed his span of control.
So an 8-man infantry section is controlled by a Section Commander (which can be further broken down to 2 4-man fire teams, but rarely). 4 sections report to a Platoon Commander. 4 Platoons report to a Company Commander. 4 Companies report to a Battalion Commander, and so on through Regiments, Brigades, Divsions, and so forth.
In this manner, the several thousand men in a large formation like a Division can be commanded by one man.
Further aiding the General in charge of the Division is that each sub-unit is granted a certain amount of autonomy within certain boundries. As you go up the chain, orders become more general (2 Battalion is to attack objective Oscar) and as you go down the chain, orders become more specific (Bob, you're carrying the machine gun)
While Linux development does not require this level of structure (and indeed, would probably suffer if this many authority-resistant cats were forced into such a regimented structure) the basic principle of "span of control" is still applicable.
The obvious correct solution is to modularize the kernel into subsections with clearly defined areas of responsibility, with a "mini-Linus" (who Linus trusts) granted control over each module.
Not suprisingly, this is what Linus has suggested.
The trick (and the catch) is two-fold. Firstly, you have to find people who are both responsible enough to be able to act as a "mini-Linus" without dropping the ball (which means Linus has to trust them too) and secondly, the kernel has to be modular enough so that changes in Module A do not step on or otherwise negatively impact code in Module B.
That's a hard row to hoe.
What I find a little distressing is that there appears to be a whiff of revolution in the air; of people talking about overthrowing Linus (and Alan Cox seems to be among them) This will not solve the problem. It is not that **Linus** doesn't scale, but rather that any single maintainer doesn't scale. Any revolution will just be "same shit, different name"
CVS doesn't solve the problem either, as anyone with commit privilages can jsut dump anything they want in there, and it becomes canon. Linus' function as a shit filter is very very important. every patch *must* be audited by someone with very high standards before it goes in.
I hope the crew rallies around Linus - as usual, he's right.
DG
Maybe it's just me, but I don't understand the concept of streaming video to a phone.
For me, the ultimate cellphone would have the following:
1) A decent phone, with decent battery life, reception, and audio clarity (how many convergence devices overlook the fact that the primary purpose of the damn thing is a telephone?)
2) Palm Pilot-like functionality, with emphasis on providing space/ability for uploading apps of my own as well as the canned apps.
3) A GPS, with detailed street maps and wayfinding ability built in.
4) Integrate this stuff as tightly as possible, and keep as much of the data local as possible.
I can see, for example, having the complete North American phone directory on the phone, so I can look up numbers without hitting the network. Tie this into the GPS, and now I can do stuff like "get me the phone number for the house I'm standing in front of right now" or "Let me search the yellow pages for [whatever] and now that I've made a selection, give me driving directions to get there"
Or allow phones to transport GPS data on voice connections, and now I can get a map of where whoever it is I'm talking to is - geographic caller ID.
Network access is all well and good, but phones are phones first and foremost. Build in apps that support the "phone" part (things like searchable directories) and the "mobile" part (with the GPS) and now you're talking!
The Kyocera Smartphone (which is a Palm) seems close, and will probably be my next phone, but I'm still looking forward to a well-done phone+GPS combo.
DG
There seems to be a fair number of people actually employed in these fields responding to this article, so this seems like a good place to ask this question.
It appears to me that we're pretty far along when it comes to the biology of sickness-by-infection, where an illness is caused by being attacked by other organisms. There's a long way to go, but it seems to an outsider that most of the fundimental processes are understood, and the lion's share of what remains is of the nature of "find germ, study germ, develop treatment that kills germ without killing host"
But it also seems that we're not very far along when it comes to understanding sickness-though-internal-breakdown, where actual body processes either fail to function or function abnormally.
It strikes me that understanding how human genetics really work is the key to all survival. If we knew how every gene and every internal process functioned, then we cound re-engineer our own genome to fix problems. Eliminate cancer, eliminate AGING, and so forth.
It would thus suggest to me that working on deciphering the human genome is the most important problem in human biology in history, and perhaps even the most important problem EVER.
We should have huge amounts of public money poured into this problem, with all results made public, and all information shared.
Would you agree? Have I made any erronious assumptions?
DG
I've never been much of a fan of the concept of software for sale, as product. To my mind, writing software is a SERVICE, with the end result free for all to use.
So Loki was providing the service of selecting games of good quality, and porting them to Linux where I could use them with my OS of choice.
As such, I wasn't a Loki *customer* so much as I was a Loki *sponsor* or *patron*.
And accordingly, I paid for every Loki game I have. RT and M2 I bought direct from Loki. D3, SoF, and Q3 I bought from the local EB.
Myth 2 I bought yesterday, when I heard they were closing. If I had more ready cash, I would have picked up the rest of their ouvre.
Oh well. Farewell Loki, you did great work. I did what little I could to support your efforts.
Your patron,
DG
I've got a Windoze 95 laptop that I need to use for programing my race car - every other machine I own/touch is Unix/Linux.
It would be good to have some sort of anti-theft reporting tool on it.
Where'd you find this ipconfig program? I had a quick look at the laptop, and didn't see it there. Is it a Win95 deal?
Does Win95 have the ability to run a script at network start?
Thanks
DG
The Motec ECU is a pretty serious system. It gets used in LeMans LMP cars, WRC cars, and the like.
http://www.motec.com to see. You can download evaluation copies of the programs to play with, if you like.
I have a similar system in my race car, made by GEMS (http://www.gems.co.uk) and it's awesome. Computerized fuel injection has revolutionized racing. You wouldn't believe the crap that thing can do.
Of course, the downside to having complete and total control over the engine is that you have to do all the calibration yourself. After having programmed the idle control/antistall stepper motor, and the accursed warmup maps, I have a huge amount of resepct for the OEM calibration guys.
Getting a car to idle smoothly under all weather conditions is tough work.
In any case, if Romero's Ferrari has a Motec system in it, then that ain't your garden variety Ferarri.
DG
I'm kinda-sorta thinking about picking up a new processor/MB/case, and I know I want an AMD chip, but I've yet to find any easy to understand table that clearly deliniates what the differences are between the various Athlons.
Anybody got a source for such a beast? Or even feel like explaining it here?
DG
I live in a Windows-free world, so I guess I'm a Linux gamer.
The downside is that there's not much in the way of games. The upside is that those that are there are pretty good.
My "game machine" is a P1-233 with a PCI GF-MX200 in it. Plays Q3 just fine. And I've been making my way through the Loki ouvre and having a ball with it.
I'm almost finished SOF, and then there's Descent 3 waiting in the wings.
Am I typical of the bleeding-edge, overclocked, 3000 FPS gamer that Windows seems to attract? Hell no. But I AM using Linux as my gaming OS, and I AM having fun with it, and I'm using a machine that's over 5 years old.
Which is pretty cool, as far as I'm concerned.
DG
You're quite right that a BSA "audit" causes nothing but fear and resentment.
Who else does audits? The IRS. And when was the last time you heard anyone preaching their love of the IRS?
But with the IRS, you have no choice. You must pay your taxes, and they'll do what they have to get them, public relations be dammed.
Monopolies don't have to worry about customer service. Where else are you going to go?
Now hand over the money, slave.
DG
Well, there's a major difference between this technology and a lie detector.
Humans are not very good at detecting lies, and can in fact get very good at certain, specialized forms of deception, even in the face of equally speciallized deception-detectors. (I'm thinking here of professional poker players, who make their living on deceiving and detecting deception, in a very specialized manner)
So a lie detector is an attempt to augment an ability.
But humans are VERY good at facial recognition, much better than any machine is. The limitation here that the machine is trying to address is one of storage capacity (memory) and speed, not ability per sae.
I make the assumption here that the interface would display the picture of the person that it thinks the person in question matches alongside the picture from its camera (or print one for the officer to carry with him for an in-person comparison) so that the vastly superior human facial recognition abilities could be brought into play.
So, really, I don't see any reason to get upset. It's no different than if an officer saw your photo on a "wanted" poster in the post office, and then made the match from memory, only now the "memory" is much larger.
When an arrest warrent is cut, your driver's licence is flagged in the DMV computers. If you get pulled over (for any reason) and your licence is checked against this database, they get you. How is this any different?
DG
As I understood it, the issue with facial recognition was the possibility of false positives; ie, I'm just trying to watch the SuperBowl. the FR system tags me as a known terrorist (incorrectly :) and the next thing I know I'm being dragged off to the can for some serious interregation (and not only unjustly tramatized, but I miss the game too)
But from the ACLU's press release, there was always a human step in the process, where a real live human being would examine each purported match before anybody got dragged off anywhere.
As such, all the face recognition software is is a _filter_, cutting down on the number of people a human agent must examine. Where's the problem?
After all, law enforcement officers have placed themselves in public places, looking for people they knew, for probably as long as there have been law enforcement officers.
A friend of mine was a sergent in the British Army, and he did a few tours in Northern Ireland. Part of his training was memorizing the faces of a large number of known IRA "players" (and apparently the IRA did the same thing with British soldiers' faces)
How is this any different?
I guess I don't understand the ACLU's beef here.
DG
Keep track of, list on the user page, and possibly use as the default rating of a user's posts, the "average rating per post".
:)
IE, if user FOO has made 100 posts, and the average rating of those posts is +3, then display it on ~FOO and let FOO post at +3, if he chooses.
This doesn't (at first blush) seem to be vulnerable to abuse, and it seems like it would self-manage trolls (who I assume would have an average of -1)
Oh, and a spellcheque facility would be nice
DG
Aside from the usual privacy rants, there's another chilling effect here.
What if what I'm looking for is the info that is not normally availible to me through my local-access media.
Maybe I WANT info in not-my-language. Maybe I WANT the teasers from the other market. Maybe I WANT the spoilers. Etc. Etc Etc.
Something a lot of old-media people haven't seemed to grasp yet is that the Internet-as-media is an INDIVIDUAL DEMAND system, or if you prefer, a "pull" instead of a "push"
The more you try to force push-style paradiems on me, the less useful your outlet becomes.
The overriding consideration for any Internet media is "make it as easy to find what I'm looking for as possible" with things like clear indicies, functional and accurate searches, and so on. Don't make assumptions about what it is I want, and push it at me.
Here's the best analogy I can think of: A "push" media, like TV or Radio, can be used by a group. Assuming the audience has similar interests (and that the TV/speakers are large enough) a TV show can be watched alone, in a small group, or even a football stadium full of people.
Can you imagine a football stadium of people, watching the screen as one guy surfs the site?
Using the Internet is a personal and individual experience. It's a library, not a program. Trying to force broadcast-media concepts on it just doesn't work.
DG
Part of the problem with revolutionary ideas is that there will forever be people who just don't understand; who cannot grasp the new concept, and who will attempt to recast it in terms they _do_ understand - only to miss the whole point all over again.
Such is Linux and Windows.
Windows is a PRODUCT. It is for sale, complete with sales reps, marketing budgets, and an army of lawyers to try and enforce the alien concept of "product scarcity" on a digital entity.
As a "product", it is subject to the rules of the market; the ebb and tide of economics.
Linux is NOT A PRODUCT, it is something else entirely. It's part common property, part social movement, part fun little hobby, and part irresistable juggernaut. In fact, I don't yet think there exists an English word that adequately expresses what Linux is. What do you call a tool that is owned by nobody, is constructed and maintained by many, and freely availible to all?
There are companies that produce products BASED on Linux, and these companies often subsidize contributions back to the greater whole, but these companies are no more "Linux" than Frito Lay or Doritos are "corn".
As long as the source code remains availible, and as long as it continues to function on existing hardware. Linux cannot "fail".
This is what the article author does not understand, and why Linux is so dangerous to Microsoft's monopoly. Linux, in some form, will _always_ be there. It will _never_ go away. It cannot be bought, swept under the rug, supressed, or otherwise made to go away.
The best you can do is to write code that does the same job, better - but we're seeing that Linux can develop every bit as fast (and oftentimes faster) as any proprietary product. No company, no matter how big, can muster a workforce as large as that actively working on Linux. Given enough time, Linux will eventually catch you and beat you on quality.
Bill Gates is often given credit for "inventing" the concept of software-for-sale, where previously, software was shared amongst users and developers free of cost. Well then, Bill has made his own bed. Linux is the ultimate competitor; the anti-Microsoft incarnate.
And a welcome CORRECTION, bringing software back from the artificial world of "product", to the real world of "service" where it originated and BELONGS.
.
Let's try this again.
Wouldn't clustering be a way to circumvent the law in the first place?
The problems that the law was intended to make difficult to solve (nuclear weapons simulation, aero flow analysis, cryptography, and so on) are, as far as I can tell, problems that can can be attacked in parallel, and so are good applications for clusters to tackle.
Well then, if the restriction prevented the export of any computer faster than x, couldn't a cluster of n export-legal computers of speed y (y less-than x ) produce a total throughput power Y (Y greater-than x)?
And for smaller values of y, substitute larger values of n to gain the same net power Y.
So really, I would think that clustering technology rendered (heh) the restriction moot a long time ago.
.
Wouldn't clustering be a way to circumvent the law in the first place?
The problems that the law was intended to make difficult to solve (nuclear weapons simulation, aero flow analysis, cryptography, and so on) are, as far as I can tell, problems that can can be attacked in parallel, and so are good applications for clusters to tackle.
Well then, if the restriction prevented the export of any computer faster than x, couldn't a cluster of n export-legal computers of speed y (y x ?
And for smaller values of y, substitute larger values of n to gain the same net power Y.
So really, I would think that clustering technology rendered (heh) the restriction moot a long time ago.
.