Recently, I was inspired to look up an old company I use to work for. They employed about 12 people total.
They had three sales people, three support people, on tester, one secretary, three programmers. One of the programmers doubled as their sysadmin. The support staff had to work on bugs for Q&A in their time between calls. They literally had clients that were some of the biggest lawfirms around.
They made a product. They sold a product. They made money.
The guys who started the thing took out personal loans to keep it going for awhile. He passed out profits back to the employees when times were good. Honestly, if there was a place to be promoted to or a position open when I was ready to go on I probably would have never left.
Small companies can survive in the IT world. They just have to have half a clue in their heads to do it.
Fill a niche, concetrate and expand along the niche not outside it, keep employee and overhead costs low (their building was nothing grand but I had my own office).
This is basic business stuff that many companies still have no concept of.
It's interesting that Sony is working on a series of "convergence" devices like the "multipurpose" PSX. That would be the caused by the many sales of xboxes related to people wanting to use it as a multifunction entertainment device.
Of course, people who use the xbox as a multifunction device tend to hack it and use linux or whatever to drive it. Not Microsoft's intention I'm sure!
The consumer electronics industry has pushed "convergence" for a long time, and the market has continued to support the modularity of separate devices over all-in-one machines.
Of course, that doesn't mean that some multifunctionality isn't desired. Personally, I love the fact that my game consoles (minus the Cube) play DVD movies. That sort of convergence wasn't forced, but was just a natural choice, given the fact that Sony and Microsoft opted to use the DVD format for their game discs. Now, adding DVD to a console that doesn't use DVD for the game functionality would be a different story.
At this point, though, I don't see shoving TiVO-like capabilities into game machines. In the future, when the standard hard drives are much larger, maybe it will be just a natural feature to add in. But I think it is ill-advised to add features that require equipment above and beyond what you're using for the device's primary functionality.
Simultaneous Multithreading (SMT) is not a new idea, although no one to my knowledge has implemented it yet. Intel just calls it "Hyperthreading"...it is essentially SMT.
And yes, this is a very good idea. A modern superscaler out-of-order processor, like the Athlon and Pentium Pro (and later), can issue and retire multiple instructions per clock cycle. However, it can *only* do this if there is enough instruction-level parallelism (ILP). Turns out, there is not enough ILP in current programs to take full advantage of the chips processing capabilities. Issue slots and function units go unused due to dependencies in the program and cache misses that stall the processing. A typical processor can only look at about 32 instructions at a time. This is not a large enough window to execute future instructions out-of-order when such a stall occurs.
However, 2 threads of execution will likely fill all of the issue slots. They are also independent threads of execution, so dependencies don't exist between them. This means that when the pipeline stalls due to a cache miss, the other thread can keep on retiring instructions.
To all those saying that this is dumb, I suggest you study some modern architecture (I'm not talking about your undergrad architecture course either). A paper I read recently studied the affects of SMT on a simulated Alpha processor. The results were astounding with very little changes to the processor core. I heard that the next Alpha was slated to include SMT before Intel killed it.
I agree. I really don't understand most Slashdot readers. In every news about KDE or Gnome people start fighting on what is better, Qt or Gtk, C or C++, Gnome or KDE, with theory on how SuSE buyout will end KDE, why Qt isn't free, that Linus uses KDE, Trolltech is owned by SCO, etc. People who keeps posting things like this must be new to free software, or just don't understand it at all. The goal is *NOT* to kill Microsoft Windows and every OS and have just GNU/Linux with one desktop installed on all computers. The goal is freedom, is choice. I don't want to be like 10 years ago, when I thought DOS/Windows were the only operationg system available. Also, most free software projects are coded for fun. I can assure you, even if the whole world start using Gnome and KDE is just used by it's own developers, KDE will keep existing! There's no desktop war, so there is not going to be a winner. So, understand the community, and stop flames. We should be discussing how great it is that someone is trying to make GTK apps integrate better to the KDE environment, and hoping a GTK coder will start doing the same. I use KDE and I get really happy when I see a new feature on Gnome, cause probably KDE will have it too soon, Gnome users should feel the same way when KDE gets a new feature. And, while we're still talking about this, please, when a project is posted here, let's not comment on how there was already a project with the same goal and how duplicate effort is lame: if someone think it'll be fun to code another mp3 player, let him do it, *For coders, projects are mostly about fun!*
I hope they get it right, already. I bet it's gonna be some bloated kitchen sink that resembles Nautilus in complexity, complete with all kinds of previews and bells and whistles, and that it still won't be able to remember the last used directory.
It also looks like they are trying to catch up with KDE's file selector. No matter what people say, that's not my ideal one. I'm much more fond of the one Mozilla [Firebird] has -- that one is the embodiment of the KISS principle to such extent I'd venture to call it perfect. That's if you agree on the definition of perfect as being "not nothing to add, but nothing left to take away".
I believe there was a Ask Slashdot a few weeks ago regarding building your own PVR. The majority of the comments seemed say "Why bother, just buy a TIVO/Replay TV, its already done for."
Well, this is why you roll your own. Yes, its a little more work, the cost is pretty much the same, but there is no monthly fee, and features don't get yanked out from under you.
MythTV is absolutely amazing, and its evolving incredibly fast. If your lookinng for a PVR, I recommend giving it a shot.
Not to mention that Google is basically being forced to IPO and doesn't have much of a choice in the matter.
Google has remained private as long as possible. If their VCs were looking to cash out, they could have done it before the crash. And everyone has been asking them to IPO for the last 2 years to kickstart the stock market. It was smart of them to wait until the DJ was above 10,000, but probably unnecessary.
The reason they are going public is because SEC rules force companies with a certain number of owners to go public. The companies have to file all the costly paperwork as if they were a public company, and they lose most of the advantages of staying private, such as not releasing all that information about their activities. There is little reason to stay private, and the extra cash from the IPO is handy for paying for all that paperwork.
The famous case of this happening was Microsoft. Too many employees were exchanging shares privately, and the SEC forced them to go public. They did really well, and you cannot blame their decline on being a public company since the prior management is still running things. OTOH, because MSFT is public, the shareholders can insist on new management, but they will probably wait until the stock goes under $10, and that will be too late to save the company, if it isn't already.
Maybe Yahoo is changing for a reason
on
Yahoo to Dump Google
·
· Score: 1, Insightful
The problem with Google now is that it has almost entirely been taken over by commercial entities. When I was recently shopping for a digital camera, I did the usual internet searches. A few years back, similar searches would have found lots and lots of sites ABOUT the product in question (fan sites, discussion forums, reviews). Now I have to sort through page upon page of sites wanting to sell me said item, most of which aren't even actual store-fronts but instead just referral pages which have manipulated the Google ranking system to get on top. I recenlty hit the same problem when doing vacation planning. It used to be that I could easily find hundreds of pages ABOUT the destination, now I just find sites wanting to sell me airfare, book me into a hotel, and rent me a car. It's become extremely frustrating and has made Google far less useful than it once was. In fact, most of the big search engines are far less useful than they once were.
Yahoo used to be THE place to get organized info on any subject. Maybe they are switching to a better search engine, like DMOZ or Vivisimo?
By sending out these clearly fraudulent legal notices-- which at best claim copyright over something which is uncopyrightable, and at worst is an attempt by a third party to claim that it is illegal for people to use the materials owned by the BSD regents under the BSD license in the manner in which the BSD regents intended the BSD license to be used-- has SCO opened itself up to legal action?
SCO has in the past managed to sidestep most allegations of fraud by being horrendously vague. They said that they were owned money but never sent any invoices, sidestepping mail fraud. They tried to present things as if you needed an SCO license to use linux, but if you tried to talk to talk to them, they were actually selling UnixWare licenses and not in the process actually distributing linux to you, sidestepping GPL violations. However, this is entirely non-vague. It seems to me that SCO has stepped over some sort of line here and this is actionable.
I know that the law does not seem to have many consequences for people who send out bad takedown notices, but surely there must be something preventing company A from finding lists of competitor B's customers and sending them takedown notices for using some portion of competitor B's product that company A does not, in fact, own.
At the least, can this be added to the lanham act/ restraint of trade/ libel or whatever countersuits that Redhat and IBM have going? What are the options from here, and what will actually happen?
FCC's chairman Michael K. Powell issued the following statement:
"We will not rest until consumers find peace from unwanted and unlawful intrusions - whether from telemarketing calls or junk faxes."
Uhm... wow, all hail Michael, here to save us from junk-faxes. Is this guy for real? Is he running for office, and/or trying to cover up the fact that they really can't do much about junk mail?
Fax.com might not have been doing something that made people feel all warm and gooey inside, but it was contributing to the economy and giving people valuable employment. Like all government intervention in business, this will lead to inefficiency and stagnation. The free market should be left to decide these matters, not lawyers and government departments.
The scheduler is an important part of the kernel, so this is an important thing to understand. This article describes it well, but just to summarize a bit more technically, the benefits of the O(1) scheduler are:
Lean and mean (low overhead).
Scales well with the number of tasks (O(1)).
Scales well with the number of processors (O(1) for scheduling, O(N) for balancing).
Strong affinity avoids tasks bouncing needlessly between CPUs.
Initial affinity makes it likely that request/response-type tasks stay on the same CPU (i.e., good for LMbench lat_udp et al)
BTW, It's good to see that the starvation and affinity problems that plagued the early versions of the O(1) scheduler have been ironed out.
But it's not like successful anti-trust lawsuits ever punish infringing companies enough. For example, Microsoft has been found to be an illegal monopoly time after time. But no serious punishment or solution, such as splitting up the company, has even been considered. Meanwhile, Microsoft continues to unfairly exploit it's desktop monopoly and crush any competition.
I am doing some graduate studies in this field, so here's a quick breakdown.
Power electronic circuits have traditionally been based on magnetic technology, and until recently, have not been part of the tide of miniaturization and integration advances from which signal-processing integrated circuits have benefited. In an effort to miniaturize power components, acoustic rather than inductive coupling can be used as the basis for a transformer. Note that acoustic coupling can be achieved through piezoelectric *or* magnetostrictive means.
In a piezoelectric transformer, the direct and converse piezoelectric effects are used to acoustically transform power from one voltage and current level to another through a vibrating structure. The converse piezoelectric effect, in which an applied electric field produces a resulting strain in a body, is used to convert an oscillating electric field applied to the left half of a structure, such as a bar, into a vibrational mode of the entire bar. If driven at resonance, standing-wave distributions of large amplitudes of stress and strain result. The resonantly amplified strain in the right half of each bar is converted to a voltage across the output terminals by the direct piezoelectric effect. Depending upon geometry and materials parameters, you can obtain voltage amplifications of various magnitudes, with associated step-downs in current level.
The unique nature of the piezoelectric transformer offers the opportunity for innovative circuit design such as operating above resonance for inductive behavior to achieve soft-switching without compensating inductors.
I believe there was a Ask Slashdot a few weeks ago regarding building your own PVR. The majority of the comments seemed say "Why bother, just buy a TIVO/Replay TV, its already done for."
Well, this is why you roll your own. Yes, its a little more work, the cost is pretty much the same, but there is no monthly fee, and features don't get yanked out from under you.
MythTV is absolutely amazing, and its evolving incredibly fast. If your lookinng for a PVR, I recommend giving it a shot.
Slashdot Mirror
Recently, I was inspired to look up an old company I use to work for. They employed about 12 people total.
They had three sales people, three support people, on tester, one secretary, three programmers. One of the programmers doubled as their sysadmin. The support staff had to work on bugs for Q&A in their time between calls. They literally had clients that were some of the biggest lawfirms around.
They made a product. They sold a product. They made money.
The guys who started the thing took out personal loans to keep it going for awhile. He passed out profits back to the employees when times were good. Honestly, if there was a place to be promoted to or a position open when I was ready to go on I probably would have never left.
Small companies can survive in the IT world. They just have to have half a clue in their heads to do it.
Fill a niche, concetrate and expand along the niche not outside it, keep employee and overhead costs low (their building was nothing grand but I had my own office).
This is basic business stuff that many companies still have no concept of.
It's interesting that Sony is working on a series of "convergence" devices like the "multipurpose" PSX. That would be the caused by the many sales of xboxes related to people wanting to use it as a multifunction entertainment device.
Of course, people who use the xbox as a multifunction device tend to hack it and use linux or whatever to drive it. Not Microsoft's intention I'm sure!
The consumer electronics industry has pushed "convergence" for a long time, and the market has continued to support the modularity of separate devices over all-in-one machines.
Of course, that doesn't mean that some multifunctionality isn't desired. Personally, I love the fact that my game consoles (minus the Cube) play DVD movies. That sort of convergence wasn't forced, but was just a natural choice, given the fact that Sony and Microsoft opted to use the DVD format for their game discs. Now, adding DVD to a console that doesn't use DVD for the game functionality would be a different story.
At this point, though, I don't see shoving TiVO-like capabilities into game machines. In the future, when the standard hard drives are much larger, maybe it will be just a natural feature to add in. But I think it is ill-advised to add features that require equipment above and beyond what you're using for the device's primary functionality.
Simultaneous Multithreading (SMT) is not a new idea, although no one to my knowledge has implemented it yet. Intel just calls it "Hyperthreading"...it is essentially SMT.
And yes, this is a very good idea. A modern superscaler out-of-order processor, like the Athlon and Pentium Pro (and later), can issue and retire multiple instructions per clock cycle. However, it can *only* do this if there is enough instruction-level parallelism (ILP). Turns out, there is not enough ILP in current programs to take full advantage of the chips processing capabilities. Issue slots and function units go unused due to dependencies in the program and cache misses that stall the processing. A typical processor can only look at about 32 instructions at a time. This is not a large enough window to execute future instructions out-of-order when such a stall occurs.
However, 2 threads of execution will likely fill all of the issue slots. They are also independent threads of execution, so dependencies don't exist between them. This means that when the pipeline stalls due to a cache miss, the other thread can keep on retiring instructions.
To all those saying that this is dumb, I suggest you study some modern architecture (I'm not talking about your undergrad architecture course either). A paper I read recently studied the affects of SMT on a simulated Alpha processor. The results were astounding with very little changes to the processor core. I heard that the next Alpha was slated to include SMT before Intel killed it.
I agree. I really don't understand most Slashdot readers. In every news about KDE or Gnome people start fighting on what is better, Qt or Gtk, C or C++, Gnome or KDE, with theory on how SuSE buyout will end KDE, why Qt isn't free, that Linus uses KDE, Trolltech is owned by SCO, etc. People who keeps posting things like this must be new to free software, or just don't understand it at all. The goal is *NOT* to kill Microsoft Windows and every OS and have just GNU/Linux with one desktop installed on all computers. The goal is freedom, is choice. I don't want to be like 10 years ago, when I thought DOS/Windows were the only operationg system available. Also, most free software projects are coded for fun. I can assure you, even if the whole world start using Gnome and KDE is just used by it's own developers, KDE will keep existing! There's no desktop war, so there is not going to be a winner. So, understand the community, and stop flames. We should be discussing how great it is that someone is trying to make GTK apps integrate better to the KDE environment, and hoping a GTK coder will start doing the same. I use KDE and I get really happy when I see a new feature on Gnome, cause probably KDE will have it too soon, Gnome users should feel the same way when KDE gets a new feature. And, while we're still talking about this, please, when a project is posted here, let's not comment on how there was already a project with the same goal and how duplicate effort is lame: if someone think it'll be fun to code another mp3 player, let him do it, *For coders, projects are mostly about fun!*
I hope they get it right, already. I bet it's gonna be some bloated kitchen sink that resembles Nautilus in complexity, complete with all kinds of previews and bells and whistles, and that it still won't be able to remember the last used directory.
It also looks like they are trying to catch up with KDE's file selector. No matter what people say, that's not my ideal one. I'm much more fond of the one Mozilla [Firebird] has -- that one is the embodiment of the KISS principle to such extent I'd venture to call it perfect. That's if you agree on the definition of perfect as being "not nothing to add, but nothing left to take away".
I believe there was a Ask Slashdot a few weeks ago regarding building your own PVR. The majority of the comments seemed say "Why bother, just buy a TIVO/Replay TV, its already done for."
Well, this is why you roll your own. Yes, its a little more work, the cost is pretty much the same, but there is no monthly fee, and features don't get yanked out from under you.
MythTV is absolutely amazing, and its evolving incredibly fast. If your lookinng for a PVR, I recommend giving it a shot.
Not to mention that Google is basically being forced to IPO and doesn't have much of a choice in the matter.
Google has remained private as long as possible. If their VCs were looking to cash out, they could have done it before the crash. And everyone has been asking them to IPO for the last 2 years to kickstart the stock market. It was smart of them to wait until the DJ was above 10,000, but probably unnecessary.
The reason they are going public is because SEC rules force companies with a certain number of owners to go public. The companies have to file all the costly paperwork as if they were a public company, and they lose most of the advantages of staying private, such as not releasing all that information about their activities. There is little reason to stay private, and the extra cash from the IPO is handy for paying for all that paperwork.
The famous case of this happening was Microsoft. Too many employees were exchanging shares privately, and the SEC forced them to go public. They did really well, and you cannot blame their decline on being a public company since the prior management is still running things. OTOH, because MSFT is public, the shareholders can insist on new management, but they will probably wait until the stock goes under $10, and that will be too late to save the company, if it isn't already.
The problem with Google now is that it has almost entirely been taken over by commercial entities. When I was recently shopping for a digital camera, I did the usual internet searches. A few years back, similar searches would have found lots and lots of sites ABOUT the product in question (fan sites, discussion forums, reviews). Now I have to sort through page upon page of sites wanting to sell me said item, most of which aren't even actual store-fronts but instead just referral pages which have manipulated the Google ranking system to get on top. I recenlty hit the same problem when doing vacation planning. It used to be that I could easily find hundreds of pages ABOUT the destination, now I just find sites wanting to sell me airfare, book me into a hotel, and rent me a car. It's become extremely frustrating and has made Google far less useful than it once was. In fact, most of the big search engines are far less useful than they once were.
Yahoo used to be THE place to get organized info on any subject. Maybe they are switching to a better search engine, like DMOZ or Vivisimo?
By sending out these clearly fraudulent legal notices-- which at best claim copyright over something which is uncopyrightable, and at worst is an attempt by a third party to claim that it is illegal for people to use the materials owned by the BSD regents under the BSD license in the manner in which the BSD regents intended the BSD license to be used-- has SCO opened itself up to legal action?
SCO has in the past managed to sidestep most allegations of fraud by being horrendously vague. They said that they were owned money but never sent any invoices, sidestepping mail fraud. They tried to present things as if you needed an SCO license to use linux, but if you tried to talk to talk to them, they were actually selling UnixWare licenses and not in the process actually distributing linux to you, sidestepping GPL violations. However, this is entirely non-vague. It seems to me that SCO has stepped over some sort of line here and this is actionable.
I know that the law does not seem to have many consequences for people who send out bad takedown notices, but surely there must be something preventing company A from finding lists of competitor B's customers and sending them takedown notices for using some portion of competitor B's product that company A does not, in fact, own.
At the least, can this be added to the lanham act/ restraint of trade/ libel or whatever countersuits that Redhat and IBM have going? What are the options from here, and what will actually happen?
Uhm... wow, all hail Michael, here to save us from junk-faxes. Is this guy for real? Is he running for office, and/or trying to cover up the fact that they really can't do much about junk mail?
the State hits down hard on free enterprise.
Fax.com might not have been doing something that made people feel all warm and gooey inside, but it was contributing to the economy and giving people valuable employment. Like all government intervention in business, this will lead to inefficiency and stagnation. The free market should be left to decide these matters, not lawyers and government departments.
- Lean and mean (low overhead).
- Scales well with the number of tasks (O(1)).
- Scales well with the number of processors (O(1) for scheduling, O(N) for balancing).
- Strong affinity avoids tasks bouncing needlessly between CPUs.
- Initial affinity makes it likely that request/response-type tasks stay on the same CPU (i.e., good for LMbench lat_udp et al)
BTW, It's good to see that the starvation and affinity problems that plagued the early versions of the O(1) scheduler have been ironed out.But it's not like successful anti-trust lawsuits ever punish infringing companies enough. For example, Microsoft has been found to be an illegal monopoly time after time. But no serious punishment or solution, such as splitting up the company, has even been considered. Meanwhile, Microsoft continues to unfairly exploit it's desktop monopoly and crush any competition.
Now all it needs is (proper!) transactions, rollbacks and subselects, and then it might be suitable for a production environment.
I am doing some graduate studies in this field, so here's a quick breakdown.
Power electronic circuits have traditionally been based on magnetic technology, and until recently, have not been part of the tide of miniaturization and integration advances from which signal-processing integrated circuits have benefited. In an effort to miniaturize power components, acoustic rather than inductive coupling can be used as the basis for a transformer. Note that acoustic coupling can be achieved through piezoelectric *or* magnetostrictive means.
In a piezoelectric transformer, the direct and converse piezoelectric effects are used to acoustically transform power from one voltage and current level to another through a vibrating structure. The converse piezoelectric effect, in which an applied electric field produces a resulting strain in a body, is used to convert an oscillating electric field applied to the left half of a structure, such as a bar, into a vibrational mode of the entire bar. If driven at resonance, standing-wave distributions of large amplitudes of stress and strain result. The resonantly amplified strain in the right half of each bar is converted to a voltage across the output terminals by the direct piezoelectric effect. Depending upon geometry and materials parameters, you can obtain voltage amplifications of various magnitudes, with associated step-downs in current level.
The unique nature of the piezoelectric transformer offers the opportunity for innovative circuit design such as operating above resonance for inductive behavior to achieve soft-switching without compensating inductors.
I believe there was a Ask Slashdot a few weeks ago regarding building your own PVR. The majority of the comments seemed say "Why bother, just buy a TIVO/Replay TV, its already done for."
Well, this is why you roll your own. Yes, its a little more work, the cost is pretty much the same, but there is no monthly fee, and features don't get yanked out from under you.
MythTV is absolutely amazing, and its evolving incredibly fast. If your lookinng for a PVR, I recommend giving it a shot.