Most people in any software project - including games - are just code monkeys. Most of them are not the guys making the creative decisions. In many cases, this is exactly the reason why software sucks. Because management refuses to listen even IF the developers know what they are talking about. Which leaves the developers with exactly two options: 1) Find another job 2) Knuckle under and implement the stupid decisions from above Of course, if this goes on for long enough, the company may find itself outmaneuvered by another, smarter company. Which will leave both managers and code monkeys out of work.
Re:Guess we'll finally see linux on the desktop
on
Intel Claims No DRM
·
· Score: 1
I'd expect at least one chipset/CPU maker to jump ship and cater to the FOSS crowd. The market may be only a few percent of the total market, but in absolute numbers it is still significant. If one vendor becomes the only supplier for that market slice, he might do quite well financially.
The real danger is in legislation that would mandate DRM technology, which makes the work of groups like EFF so important.
Obviously ISP's can't start making this mandatory right now. The Trust system doesn't really begin to roll out until the Longhorn release next summer. It would then take another few years for the majority of PCs to be replaced. PCs get replaced rather quickly through the normal obselecence and upgrade cycle. You can potentially see mandatory Trust compliance for internet access somewhere between 2010 and 2015.
Many PCs are not directly connected to the internet, but through a router with NAT. The PC is sitting somewhere on the LAN. Actually, I use this setup at home for a single computer because it is quite convenient. I think that will be another technical difficulty for enforcing Trust compliant computers (and if the $50 router has to be Trust compliant, so what? I could live with that and use it as it is)
While the principle has some merit, the authors of the article picked the worst possible example. AFAIK the Athlon 4000+ is still manufactured in 130 nm technology ("Clawhammer"), and it is not exactly cheap. For less money, you can get an Athlon 3800+ with the new Venice core (90nm technology) which uses MUCH less power than the 4000+. Unfortunately, the article does not give any numbers on the actual power consumption or ambient temperature, so we have to look elswhere: LostCircuits http://www.lostcircuits.com/cpu/amd_venice/ has some actual measurements of CPU power usage. The guys found out that the Venice/3800+ uses less than half the power of the Clawhammer/4000+. The actual clock frequency is the same for both processors, 2.4GHz. To top it off, they found that the 3800+ showed slightly better overall performance than the 4000+. It seems that the detail improvements that went into the Venice core do more than compensate the Venice's smaller cache.
From the article: Villasante also raised the potential of policy measures that could be taken nationally or internationally to encourage the use of open source software, such as in the areas of licensing and intellectual property rights... Now if he can only get the EU commission to listen, we might see an end of the attempts to establsh software patents in the EU.
A lot of redundancy seems the logical way to handle this. With the super-small circuits this technology implies, the increased usage of chip area should not be a problem. It might even improve the yield in manufacturing, because with enough redundancy a chip might be marketable despite some errors (that are overridden).
Expect the barely functional chips to go into consumer products and those with less errors at the time of packaging into server hardware.
In this case, fine by me. a) it seems to be a real invention, unlike a lot of the software patent crap we're seeing recently. b) by the time this hits the mass market, the 20 years of patent protection are probably over anyway.
Unlike typical internal combustion engines, which reach their max. torque at 2000 RPM or more, electric motors can be designed to reach their max. torque from zero RPM. The result is like driving at moderate RPMs and then hitting the accelerator pedal, only you can do it from a standstill. No need for a clutch.
The energy density of batteries remains the big problem, and if you want to know about the future of electric cars, follow the battery news rather than the electric motor news.
However in the software world, the employees are not hamstrung by monetary concerns. Any Joe Programmer can pick up a cheap $200 bare bones PC and a copy of Linux and be programming the next great thing. He doesn't need management to do this.
Well, add a good internet connection to that;-)
He also needs to be halfway good at marketing his stuff. For many people, including myself, this is the most important reason not to resign and start as a consultant.
I think Enderle does "miss the point" on purpose, in an attempt to paint the Linux community as dangerous criminals. Of course, his argumentation is weak enough to make him look like a moron. Which he probably is, otherwise he would act in a more subtle and difficult to disprove way.
Let's have a look at his technique:
1) He starts with the probably valid point that the emergence of a Software Labor Union is not unlikely, due to reasons he lists IMHO correctly.
2) "Linux: Critical Mass Requirement Met" Here, he jumps to talking about a union of Open Source developers, conveniently ignoring that the worst working conditions are reported from (closed source) game development studios. Which is where I would expect the first Software Labor Unions to pop up. He goes on to talk about "Linux and open source has penetrated most technical schools, government IT shops, and technology companies", in order to make it sound more dangerous. In order to buy this so far, you have to be uninformed or stupid. But this paragraph could still be excused as a slip in wording by a guy that needs a bit of pep in his article;-)
3) "Linux: Organization Requirement Met" Much talk about the trouble SCO is in, and vague accusations about criminal activities for which the Open Source is supposedly responsible. Again, he conveniently ignores that SCO made most of the trouble for itself. Sueing a company like IBM who can afford excellent lawyers itself, plus pissing off your own (ex-)business partners is hardly a good strategy. More talk about how "its [Microsofts] supporters often appear as an endangered species during a government-approved hunting season". By now, we are clearly in propaganda country.
3) "Linux: Power Requirement Met" Here, Enderle talks about how Maureen O'Gara wrote an article about PJ of Groklaw and the resulting backlash forced Sys-Con to pull the article. "Linux effectively made good on a threat that is beyond even Microsoft's reach, and often beyond the U.S. government's reach. That threat is putting your company out of business if the desired result is not achieved"
Wrong on two accounts: a) There were no "threats" in the way a mobster would make them. b) Microsoft has done similar things before. By other means, but they have put companies out of business that got in their way. Remember Netscape?
"What is even more amazing is the effort was so powerful it may have eliminated a sister publication as collateral damage. LinuxWorld may no longer be a viable publication after the voluntary departure of its entire editorial staff." And the editorial staff explicitly said that they were leaving due to lack of journalistic integrity on part of the publisher. Forgot to mention that, Mr.Enderle?
4) "Linux: Leadership Unmet" Here, he tries to pass of the Open Source community as fanatics and lunatics who use their power irresponsibly. To back up his argumentation, he gives a few links to media who have picked up the story about O'Gara and claims the effort backfired. But if you actually follow the links, O'Gara does NOT look like the good girl in this controversy.
Overall, Enderle comes off as a second-rate propaganda writer rather than as an analyst.
In theory, the EU does not support software patents (yet). In practice, the EU patent office has granted quite a few. While it might be possible to kill these in court, such ligitation is expensive and many small companies will settle rather than fight. Thus, having one's software patents declared invalid would still deny the patent holders a tool for extortion.
Additionally, the EU cannot take action against MS for pulling out of the market if MS pays the current fines. To declare MS's copyright void in such a case would set the kind of precedent that Washington DC politicians hunger for. If anything, the US government is proud of MS. It is a US company, and a real money maker. Whatever issues you think the government has with MS will disappear if the EU tries to pull anything.
Now this is the really interesting scenario. Suddenly being cut off from the most prevalent software might be considered the sort of national emergency that motivates the EU to take drastic measures.
If they do, we'll have major political fallout. Having "pirated" MS-ware legally available will both hurt Microsoft and the adaption of free software, with a temporary advantage to end users because they will get stuff for free. Of course, that stuff would be hopelessly obsolete some day and have to be replaced anyway.
If they don't, we will have something similar to your scenario. With the difference, however, that I think you underestimate the quality of free software. The transition would still hurt, but the EU would ultimately profit from getting rid of MS bloatware.
Now these people you speak of have real guns, and are using them to deprive MS executives of real liberties. Who is the real bully?
The MS executives are not deprived of personal liberties. The company is deprived of some liberties, but that is fine by me: Legal entities are not people who have human rights, so restraining large companies a bit in order to ensure a free market is perfectly OK. Note that I'm talking about the ethical side of the issue, not about existing laws. Existing US case law may grant more rights to corporations, but I don't care about that.
If it comes to that, I sincerely hope that your soldiers will have more courage to refuse orders than German soldiers did under Hitler.
Because your administration will certainly follow the Nazis' example of creating laws first that will make those orders at least formally legal. Actually, one might consider the PATRIOT act a first step in that direction. See also http://www.furnitureforthepeople.com/actpat.htm
So following the law about illegal orders will not be sufficient, because that law will not exist anymore when it is needed. Instead, your soldiers might have to ignore the laws outright in order to preserve liberty.
-Give the first player that touches the mob a significant preference when it comes to deciding who gets the kill. Weight his damage stronger than that of others
Fortunately, many companies who are serious about kill-stealing have gone on to code the rule into the game itself
Should be easy enough. -Give the first player that touches the mob a significant preference when it comes to deciding who gets the kill. Weight his damage -When said player leaves the mob alone for a minute or so, strike his bonus and give it to the next one who attacks that mob.
I see really no excuse for relying on EULAS in such things.
Well, voicechat would reduce your options in choosing a char somewhat. I like playing a girl now and then myself, which would obviously not work well with a male voice. On the other hand, faster communication per voice might be a big plus for fast-paced, FPS-ish MMORPGs. And a few of these are in development. Lets try it when a MMORPG with voicechat goes open beta - then we will know more.
Even so, the Linux community might prefer it over a non-documented graphics chip that only works with closed source drivers. Assuming there will be something like a Cell workstation, its Linux graphics drivers might do their calculations in the SPEs and use a cheap framebuffer card.
At a guess, I'd expect Linux to run on the normal-CPU PPC core, with use of the SPEs left as an exercise to the user. At least initially, because restructuring the whole OS to match the Cell structure would be a huge task.
As a casual player in Day Of Defeat, I usually score one kill for being killed 2-3 times. But it changes when I play a few evenings in a row, which I sometimes do on vacation. On these occasions, I usually catch up to the better players, with really nice rankings when the next map change comes up.
The terms of the license make a BIG difference here. From an end user's point of view, Open Source is almost as good as owning the copyrights. Because he is not in the business of selling the software anyway. A license that can be terminated at the whim of the vendor is the other extreme. If you are not careful what you agree to, you might be royally fucked when your application provider goes out of business. If you want to be sure that does not happen, you will need the following things at a minimum:
-an enforcable guarantee that you can get backups of your data when the service contract ends, for whatever reason. -the applications must store the data in a non-proprietary format, so you can take them to another application provider if necessary
The kernel is actually not very fat. At a former employer, we did some experimenting with XP Embedded. It needed around 3-4 Mbyte for itself to run the kernel and boot our application instead of the usual graphic shell. Now add a low-feature Explorer as in Windows 95 and you might get something that has memory consumption similar to Win98. Of course, you'd have to get rid of all services that are not necessary for a typical desktop. Otherwise, you would be back at a Win2000-like memory hog. Microsoft will have to compromise here, some applications might not run on the "XP light".
It might not happen by accident as often, but with pointers and a bit of explicit typecasting, you can create the same kind of havoc. It just takes a bit more work to tell Delphi that you want to write to some random memory address.
Slashdot Mirror
Most people in any software project - including games - are just code monkeys. Most of them are not the guys making the creative decisions.
In many cases, this is exactly the reason why software sucks. Because management refuses to listen even IF the developers know what they are talking about. Which leaves the developers with exactly two options:
1) Find another job
2) Knuckle under and implement the stupid decisions from above
Of course, if this goes on for long enough, the company may find itself outmaneuvered by another, smarter company. Which will leave both managers and code monkeys out of work.
I'd expect at least one chipset/CPU maker to jump ship and cater to the FOSS crowd. The market may be only a few percent of the total market, but in absolute numbers it is still significant. If one vendor becomes the only supplier for that market slice, he might do quite well financially.
The real danger is in legislation that would mandate DRM technology, which makes the work of groups like EFF so important.
Obviously ISP's can't start making this mandatory right now. The Trust system doesn't really begin to roll out until the Longhorn release next summer. It would then take another few years for the majority of PCs to be replaced. PCs get replaced rather quickly through the normal obselecence and upgrade cycle. You can potentially see mandatory Trust compliance for internet access somewhere between 2010 and 2015.
Many PCs are not directly connected to the internet, but through a router with NAT. The PC is sitting somewhere on the LAN. Actually, I use this setup at home for a single computer because it is quite convenient.
I think that will be another technical difficulty for enforcing Trust compliant computers (and if the $50 router has to be Trust compliant, so what? I could live with that and use it as it is)
It would also be the ultimate bandwidth saving method. But I'm afraid voice-to-text is not all that mature either...
While the principle has some merit, the authors of the article picked the worst possible example. AFAIK the Athlon 4000+ is still manufactured in 130 nm technology ("Clawhammer"), and it is not exactly cheap. For less money, you can get an Athlon 3800+ with the new Venice core (90nm technology) which uses MUCH less power than the 4000+.
Unfortunately, the article does not give any numbers on the actual power consumption or ambient temperature, so we have to look elswhere:
LostCircuits http://www.lostcircuits.com/cpu/amd_venice/ has some actual measurements of CPU power usage.
The guys found out that the Venice/3800+ uses less than half the power of the Clawhammer/4000+. The actual clock frequency is the same for both processors, 2.4GHz.
To top it off, they found that the 3800+ showed slightly better overall performance than the 4000+. It seems that the detail improvements that went into the Venice core do more than compensate the Venice's smaller cache.
From the article:
Villasante also raised the potential of policy measures that could be taken nationally or internationally to encourage the use of open source software, such as in the areas of licensing and intellectual property rights...
Now if he can only get the EU commission to listen, we might see an end of the attempts to establsh software patents in the EU.
Ooh, I'm sure the dirty, scurvey suffering swabbers cursed their jobs too. Even if it pays the rent, you do not necessarily enjoy that sort of work.
A lot of redundancy seems the logical way to handle this. With the super-small circuits this technology implies, the increased usage of chip area should not be a problem. It might even improve the yield in manufacturing, because with enough redundancy a chip might be marketable despite some errors (that are overridden).
Expect the barely functional chips to go into consumer products and those with less errors at the time of packaging into server hardware.
In this case, fine by me.
a) it seems to be a real invention, unlike a lot of the software patent crap we're seeing recently.
b) by the time this hits the mass market, the 20 years of patent protection are probably over anyway.
Unlike typical internal combustion engines, which reach their max. torque at 2000 RPM or more, electric motors can be designed to reach their max. torque from zero RPM. The result is like driving at moderate RPMs and then hitting the accelerator pedal, only you can do it from a standstill. No need for a clutch.
e ng.html
Also, electric motors that have an astounding power density already exist. An introduction to a particularly interesting design is here:
http://www.aerodesign.de/peter/2001/LRK350/index_
The energy density of batteries remains the big problem, and if you want to know about the future of electric cars, follow the battery news rather than the electric motor news.
However in the software world, the employees are not hamstrung by monetary concerns. Any Joe Programmer can pick up a cheap $200 bare bones PC and a copy of Linux and be programming the next great thing. He doesn't need management to do this.
;-)
Well, add a good internet connection to that
He also needs to be halfway good at marketing his stuff. For many people, including myself, this is the most important reason not to resign and start as a consultant.
I think Enderle does "miss the point" on purpose, in an attempt to paint the Linux community as dangerous criminals.
;-)
Of course, his argumentation is weak enough to make him look like a moron. Which he probably is, otherwise he would act in a more subtle and difficult to disprove way.
Let's have a look at his technique:
1) He starts with the probably valid point that the emergence of a Software Labor Union is not unlikely, due to reasons he lists IMHO correctly.
2) "Linux: Critical Mass Requirement Met"
Here, he jumps to talking about a union of Open Source developers, conveniently ignoring that the worst working conditions are reported from (closed source) game development studios. Which is where I would expect the first Software Labor Unions to pop up.
He goes on to talk about "Linux and open source has penetrated most technical schools, government IT shops, and technology companies", in order to make it sound more dangerous.
In order to buy this so far, you have to be uninformed or stupid. But this paragraph could still be excused as a slip in wording by a guy that needs a bit of pep in his article
3) "Linux: Organization Requirement Met"
Much talk about the trouble SCO is in, and vague accusations about criminal activities for which the Open Source is supposedly responsible.
Again, he conveniently ignores that SCO made most of the trouble for itself. Sueing a company like IBM who can afford excellent lawyers itself, plus pissing off your own (ex-)business partners is hardly a good strategy.
More talk about how "its [Microsofts] supporters often appear as an endangered species during a government-approved hunting season". By now, we are clearly in propaganda country.
3) "Linux: Power Requirement Met"
Here, Enderle talks about how Maureen O'Gara wrote an article about PJ of Groklaw and the resulting backlash forced Sys-Con to pull the article.
"Linux effectively made good on a threat that is beyond even Microsoft's reach, and often beyond the U.S. government's reach. That threat is putting your company out of business if the desired result is not achieved"
Wrong on two accounts:
a) There were no "threats" in the way a mobster would make them.
b) Microsoft has done similar things before. By other means, but they have put companies out of business that got in their way. Remember Netscape?
"What is even more amazing is the effort was so powerful it may have eliminated a sister publication as collateral damage. LinuxWorld may no longer be a viable publication after the voluntary departure of its entire editorial staff."
And the editorial staff explicitly said that they were leaving due to lack of journalistic integrity on part of the publisher. Forgot to mention that, Mr.Enderle?
4) "Linux: Leadership Unmet"
Here, he tries to pass of the Open Source community as fanatics and lunatics who use their power irresponsibly. To back up his argumentation, he gives a few links to media who have picked up the story about O'Gara and claims the effort backfired.
But if you actually follow the links, O'Gara does NOT look like the good girl in this controversy.
Overall, Enderle comes off as a second-rate propaganda writer rather than as an analyst.
In theory, the EU does not support software patents (yet).
In practice, the EU patent office has granted quite a few. While it might be possible to kill these in court, such ligitation is expensive and many small companies will settle rather than fight. Thus, having one's software patents declared invalid would still deny the patent holders a tool for extortion.
Additionally, the EU cannot take action against MS for pulling out of the market if MS pays the current fines. To declare MS's copyright void in such a case would set the kind of precedent that Washington DC politicians hunger for. If anything, the US government is proud of MS. It is a US company, and a real money maker. Whatever issues you think the government has with MS will disappear if the EU tries to pull anything.
Now this is the really interesting scenario. Suddenly being cut off from the most prevalent software might be considered the sort of national emergency that motivates the EU to take drastic measures.
If they do, we'll have major political fallout. Having "pirated" MS-ware legally available will both hurt Microsoft and the adaption of free software, with a temporary advantage to end users because they will get stuff for free. Of course, that stuff would be hopelessly obsolete some day and have to be replaced anyway.
If they don't, we will have something similar to your scenario. With the difference, however, that I think you underestimate the quality of free software. The transition would still hurt, but the EU would ultimately profit from getting rid of MS bloatware.
Now these people you speak of have real guns, and are using them to deprive MS executives of real liberties. Who is the real bully?
The MS executives are not deprived of personal liberties. The company is deprived of some liberties, but that is fine by me:
Legal entities are not people who have human rights, so restraining large companies a bit in order to ensure a free market is perfectly OK. Note that I'm talking about the ethical side of the issue, not about existing laws. Existing US case law may grant more rights to corporations, but I don't care about that.
If it comes to that, I sincerely hope that your soldiers will have more courage to refuse orders than German soldiers did under Hitler.
Because your administration will certainly follow the Nazis' example of creating laws first that will make those orders at least formally legal. Actually, one might consider the PATRIOT act a first step in that direction. See also
http://www.furnitureforthepeople.com/actpat.htm
So following the law about illegal orders will not be sufficient, because that law will not exist anymore when it is needed. Instead, your soldiers might have to ignore the laws outright in order to preserve liberty.
-Give the first player that touches the mob a significant preference when it comes to deciding who gets the kill. Weight his damage stronger than that of others
Fortunately, many companies who are serious about kill-stealing have gone on to code the rule into the game itself
Should be easy enough.
-Give the first player that touches the mob a significant preference when it comes to deciding who gets the kill. Weight his damage
-When said player leaves the mob alone for a minute or so, strike his bonus and give it to the next one who attacks that mob.
I see really no excuse for relying on EULAS in such things.
Well, voicechat would reduce your options in choosing a char somewhat. I like playing a girl now and then myself, which would obviously not work well with a male voice.
On the other hand, faster communication per voice might be a big plus for fast-paced, FPS-ish MMORPGs. And a few of these are in development. Lets try it when a MMORPG with voicechat goes open beta - then we will know more.
Even so, the Linux community might prefer it over a non-documented graphics chip that only works with closed source drivers.
Assuming there will be something like a Cell workstation, its Linux graphics drivers might do their calculations in the SPEs and use a cheap framebuffer card.
At a guess, I'd expect Linux to run on the normal-CPU PPC core, with use of the SPEs left as an exercise to the user. At least initially, because restructuring the whole OS to match the Cell structure would be a huge task.
t ion_name=dev&aid=3451
But even so, it would be a start for using things like Cell-based workstations, see also:
http://www.gamesindustry.biz/content_page.php?sec
The more you play, definitely.
As a casual player in Day Of Defeat, I usually score one kill for being killed 2-3 times. But it changes when I play a few evenings in a row, which I sometimes do on vacation. On these occasions, I usually catch up to the better players, with really nice rankings when the next map change comes up.
The terms of the license make a BIG difference here. From an end user's point of view, Open Source is almost as good as owning the copyrights. Because he is not in the business of selling the software anyway.
A license that can be terminated at the whim of the vendor is the other extreme. If you are not careful what you agree to, you might be royally fucked when your application provider goes out of business. If you want to be sure that does not happen, you will need the following things at a minimum:
-an enforcable guarantee that you can get backups of your data when the service contract ends, for whatever reason.
-the applications must store the data in a non-proprietary format, so you can take them to another application provider if necessary
Good luck getting this cheaper than your own PC.
The kernel is actually not very fat. At a former employer, we did some experimenting with XP Embedded. It needed around 3-4 Mbyte for itself to run the kernel and boot our application instead of the usual graphic shell.
Now add a low-feature Explorer as in Windows 95 and you might get something that has memory consumption similar to Win98. Of course, you'd have to get rid of all services that are not necessary for a typical desktop. Otherwise, you would be back at a Win2000-like memory hog.
Microsoft will have to compromise here, some applications might not run on the "XP light".
More like:
Dateline 2009 Google buys Red Hat.
It might not happen by accident as often, but with pointers and a bit of explicit typecasting, you can create the same kind of havoc.
It just takes a bit more work to tell Delphi that you want to write to some random memory address.