I imagine that this portion of the document might have something to do with it not being adopted, since it seems to contravene Roe v. Wade:
Bearing in mind that, as indicated in the Declaration of the Rights of the Child, "the child, by reason of his physical and mental immaturity, needs special safeguards and care, including appropriate legal protection, before as well as after birth",
Of course, if you're the type who thinks that women should not have rights over their own bodies, and would force them into slavery as incubators, your view might differ from that of the U.S. Supreme Court. http://en.wikipedia.org/wiki/Roe_v._Wade.
China has signed onto that and onto other human rights conventions, yet still practices coercive abortion and sterilization. Peru also carried out a similar program under Alberto Fujimori's administration, which is why they've been trying to extradite him from Japan; Mexico continues a similar policy: http://www.libertadlatina.org/Crisis_Forced_Sterilization.htm
It's not politically correct to point out that this type of document isn't worth spit if it's adopted but not enforced, but it's nevertheless true.
"It makes sense that people who use more should pay more. Why shouldn't the people who use more, pay more?"
When I buy a day pass to Disneyland, I expect that I can go on the Haunted Mansion ride 15 times in a row if I want to. That's why I bought a day pass. If they want me to go on it a maximum of three times, they shouldn't call it a day pass, they should give me a punch-pass instead, and when my 3 punches have been punched, I'm done (unless I buy another punch pass, of course...).
Also, be aware, that every one of those plans that the carriers sell is subject to tariff agreements between them and the PUC (Public Utilities Commission), which gives them the right to charge certain amounts for certain things, and have certain things and not others included in bundles, as part of their regulated monopoly on providing the utility in a given area. This is why your phone bill makes such a big deal about caling out the line items for lifeline service, rural service subsidies, and so on: in exchange for the monopoly in the lucrative markets, they have to serve the non-lucrative markets as well, or they lose the monopoly (and with it, the public rights of way for their wires, where needed, or frequency spectrum, where wires are not in use).
What they are crying about here is the death of their circuit-centric business model, where they used to get to charge for call competition, and for distance between end points (if you ever wondered why there are "services" such as custom ring indication based on the caller and "free" voice mail, wonder no longer). They are deathly afraid that higher data rates and VOIP are going to kill their current cash cows. And they're right: their current cash cows are going to die.
The cell phone boom saved them from becoming nothing more than dumb pipes, where the get to bill based on the diameter, but it only saved them for a short time, and now it's time to pay the piper.
"It is more of a social engineering thing - trying to trick unsuspecting users to install a malicious script by hiding it as a theme or screensaver."
Not all desktops are themeable. Apart from security, this is also a good thing when a user calls into technical support, and the person answering the phone needs to be able to tell them which icon they need to click on and where it is/should be located on their desktop. Only desktops which support/permit themes need to worry about trojans in theme bundles.
You are also not including transportation reinbursement, offset contribution to public transport (mandated for businesses with more than a small number of employees in California), and other costs, such as RSUs, stock options, employer contributions to benefits, matching on 401K plans, amortized costs for a 3 year lifetime on multiple computers per employee (I cost ~$4,000/year there) and other costs (mileage reimbursement for someone who lives in SFO and commutes to SJO can go as high as ~$10,000/year).
Managers are on call to make up for employee slack.
They get paid a salary, and they end up on call in case one of their employees babies is stolen by a dingo, or something else happens which results in the employee unavailable. The concept is called coverage, and applies not only to sales people, but to anyone who needs to answer to a pager, a cell phone, a BlackBerry, or is otherwise critical to the business because someone has to be ultimately responsible and fix things when problems arise.
You also have to consider that the costs of an additional person are way out of whack with the TCO for an employee. Your salary accounts for ~1/3 of the cost to the employer for having you around, if you count facilities and energy costs and taxes. This is only going to get worse as taxes go up to cover the costs of government spending on things like the war in Afghanistan and on universal healthcare. So it's a lot easier to increase responsibility at the cost of a small increase in salary and some title inflation than it is to hire more people.
I also expect that other posts in this thread are correct, and that you've been selected as someone to keep on the payroll prior to a pending round of cost-cutting and/or outsourcing.
A "best case" scenario is this is to put you on the hook for the end of the year holiday season.
In situations like this it's generally best to see which way the wind is blowing, and keep your options open, including the option of alternative employment.
Yes, I know people who have quit over being told their vacation was cancelled.
I attended a number of conferences with Bob (Robert) Wallace, the author of PC-Write, Andrew Fluegelman, the author of PC-Talk, and Jim Button (Knopf), the author of PC-File.
These are largely acknowledged as the first successful shareware programs, and all the authors had basically the same philosophy, but I wlil give you the nutshell version given to me by Bob Wallace:
"I don't sell software; software isn't real; it's all up here", at which point he waved his hands on both sides of his head, and then continued, "I sell manuals".
This paradigm, where the revenue is tied to the documentation, which would be relatively expensive to reproduce, whereas software was relatively easy to copy, is what drove shareware sales for Buttonware, Quicksoft, and others, who depended on their software being copied, and then people coming to them for documentation when the software was hard to use, due to it being intentionally non-intuitive.
As software designers, we've never gotten out of this rut (for the most part), and it's seen as useful to have online help and other extensive documentation for something that should be actually simple, as well as conceptually simple. Ironically, we've even gone to online documentation, either supplied on the distribution media, or on the Internet, or some combination of both (like Microsoft does), in an attempt to reduce publication costs without having to change our design paradigm.
I even have a friend who is actively *proud* of this; he says things like "If it was hard to write, it should be hard to use", and similar gems (he's my friend, but that doesn't stop him from being an idiot).
For tools intended for use by programmers, yes, there should be documentation; there's only so much complexity you can abstract there. But for finished products, if you can't start them up and just start using them for their intended purpose, then there's something wrong with the design.
"if you need a hospital or support for some physical limitation, you won't have to pay through your nose to get it, as this is seen as a basic human right"
Medical but not food as a basic right is amusing; same goes for breathing, water, sex, sleep, homeostasis, and excretion.
Only after you take care of these physiological needs do you get to the next tier of Maslow's hierarchy of needs:
where "health" is located... and that's considered a safety issue, after security of: body, employment, resources, morality, and the family, only just edging out security of property.
He declared a project, didn't put much in beyond the idea for the project, and then expected people to flock to the project and build it for him. Very much the same way people declare projects on SourceForge, and then expect volunteer programmers to come out of the woodwork and build whatever their pet project happens to be.
The only difference here is that it involved both hardware and software, and not just software.
I went to the Internet WayBack Machine and read all of the blogs from Fusion Garage (the actual site is currently down, probably intentionally), and it looks like all of the software was done by them, including the OS, and it looks (from Arrington's blogs) like a lot of the hardware was done by Intel.
I don't mean to discount the value of vision or publicity, but really, he intended to Open Source everything about the tablet when he declared the project, and I don't really see a lot of value being taken from him in that case, since he wouldn't be building hardware anyway. The only money would be in margin on the hardware if the software was all out there. A lot of people have posted similar specifications for what they'd like to see in a tablet computer, and the only difference between them and Arrington is Arrington has a lot of self-publicity and got a startup to bite on the bait to actually build the thing.
Arrington might have some arguments with regard to industrial design, but the prototype hardware was not built by him, and the software that made that hardware live and breathe was definitely not his.
I've worked at combo hardware/software startups, and I've worked at software-only startups (including my own), and universally, the hardware in the hardware/software startups was all about minimal COGS and industrial design (being at Apple now, that's pretty much all there is). The value-add over commodity hardware is that it isn't "cheapest vendor of the part of the day" (so the hardware is reliable and not crap because of constantly retooled assembly lines), and it's the software. When Apple builds a laptop, it doesn't build hardware, and it doesn't build software, it builds systems. The people who don't get that and churn out 1.5%-4% margin crap do so at their peril.
My reading of things is that Arrington is no Jonathan Ive, and he's no Steve Jobs when it comes to design of hardware or software.
Fusion Garage may have taken his idea and run with it.. and they want to cut him in on profits from something, the intrinsic value of which he intended to give away for no profit, but they don't seem to be ripping him off to do it, although they do seem to be leveraging as much as they can to get him to accept a minority role with regard to what he brings to the table (which, per the above, by my reckoning, isn't much; sorry, Arrington).
As more than one V.C. has told me in the past, the point is not the idea; there are millions of good ideas that go unfunded all the time (I'll point at SourceForge again, where "funding" equates to "provision of manpower necessary to complete a project"); what a V.C. funds is the ability to execute on a vision, no matter whose vision it is, and the team behind that ability to execute on the vision and bring a product to market. 1 in 10 entrepreneurs get funded; 1 in 10 of those fail in the first year. That's only a 1 in 100 chance of being around after a year.
Arrington's failure is no less spectacular than anyone else's in that 99 out of 100 failures, he's just made it more public by ranting about it.
Ironically, the idea may still not be a failure, merely a failure on his part to control the thing which was built on his (and a lot of other peoples similar) idea, if Fusion Garage or someone else simply continues on and executes on it.
Good luck in your future endeavors, but don't think that by declaring an idea publically that you've built or created anything.
He's being taxed by the amount it is costing him to record keep, collect, and forward the tax to the state in question.
Brick and mortar stores are taxed in the same amount, but it's very easy to record-keep, as you program up the cash registers and load the (small) exception table for the items you have to deal with. In some cases, as in prepared food at supermarkets, this adds a requirement that the person making the purchase be asked whether or not they intend to eat at tables set aside at the store, or whether they intend to take their purchase home. The answer to this question changes the tax rate in some places.
For an Internet retailer, where it's still not legally clear whether the transaction is taking place in the state of the purchaser or the seller, this is onerous in the extreme. Unlike a brick and mortar company, the transaction is either at the sellers location or the drop-ship warehouse (where consideration is finally exchanged resulting in a sales contract; this is supposedly one of the "outs" Amazon is "exploiting"), or it's at every sellers location at the time of the purchase. For example, on an airplane at 36,000 feet somewhere over the midwest.
The part which is truly onerous to the retailer is that there are, at least on the last database update I'm aware of, 160,000+ sales tax rates at various locations in the U.S., and it is nearly impossible to correlate the location with the GIS information such that you can pick the right sales tax(es) to collect for a given location.
Mail order purchases (before the Internet was such an avenue of commerce) are the reason there are state "use tax" laws, and it's the responsibility of the purchaser to pay the tax in the locality that the purchase is intended to be used.
Amazon is being given a bum rap here. Because they are a single "one stop shopping" target for the government to use as enforcement proxy via unfunded mandate ("get Amazon and you get all the tax on all the traffic through Amazon, and get them to pay for the collection process"), they have a huge bullseye painted on their back. Although the idea of normalizing sales tax across all venues, as was suggested in the article, is very attractive on the face of it, it's unlikely to ever happen.
Hogwash: Building codes are regulatory, just like FCC and FAA rules, or public utilities commission rules. The only laws involved are usually rather simple and to the point in delegating the authority to an administrative agency generally controlled by the executive branch of the appropriate government.
As far as tax law, it's only necessary to not have a graduated flat tax (e.g. taxed on what you earn above minimum was times 2080 hours + $1) if you are intent on hiding your legislative cronyism, malfeasance, kickbacks, and unfunded mandates in the tax code. If you want to legislate social policy, then be honest and legislate social policy, and if what you do is unpopular, you don't get reelected.
Also, I remember a debate from my college days when it was suggested that the best form of government was in fact a benevolent dictatorship. No thank you.
P.S.: I'd still like someone to explain to me why the disincentive for second degree murder should be less than the disincentive for first degree murder; the victim is still just as dead, right?
Most of the places they say "NO_SYSCALL_STUB" are actually decorated version of system calls for binary backward compatibility. If they ever include Framework support for certain existing frameworks, they are going to find that their use of 32 bit libraries may result in unexpected-by-the-framework cancellation points that need to not be there.
The kill system call changed from older versions of the OS to ensure POSIX compliance for one compilation environment, and historical behaviour; in fact, for the 32 bit version of libc, there are 189 decorated symbols for doing everything from translating error codes to getting variant behaviour to enforcing POSIX semantics in user space in libSystem itself, and there are 74 variants for cancellation points in both 32 and 64 bits.
Changing the contract between libSystem and the kernel under the covers to fix bugs happens often. The posix_spawn() system call is a good example of that. Apple has extended spawn several times by modifying the _posix_spawn_args_desc. The code itself uses accessors and mutators on an allocate descriptor. It can change (and does) size when new extensions are added.
Basically by providing their own direct system call interface, they've guaranteed that they will constantly be chasing a moving target, and they've lost a lot of the default POSIX semantics that are implemented at the libSystem level. For kill and similar calls, they've lost the go code portability between OSs by doing this.
Finally, they need a type mechanism so that the unsupported interfaces can be conditionalized based on the UNIX standard permitting non-implementation of certain interfaces; otherwise they will run into things like unnamed semaphores which will throw a SIGSYS, which will abort their programs.
The problem with fixating on something like ELF for fatness is that it's just the tip of the iceberg.
There's much more to the question of whether or not something will run on an arbitrary copy of Linux than the CPU arch.
Why do you think one of the discriminators in a fat binary can't be a distribution identifier, such that there are fat slices for supporting Debian, RedHat, Ubuntu, etc., all from the dame binary file?
Or that they can't have different slices in the fat binary for Gnome vs. KDE, or desktop vs. Android, and so on?
Also, the arguments about disk space are specious; at least in the Mac OS X world, there is a utility called "lipo" which will pull apart a fat bnary into only the pieces you need/want to install. Typically you only install the actually fat binary on server systems, where the software has to be able to run on multiple client machines, and otherwise you run scripts to reclaim disk space (or in the case of an embedded device, you run them over the install image before you install them).
Same goes for an Apple fat binary really...
Obligatory disclaimer: I am the person who maintains the fat binary loader in the Mac OS X kernel.
In any unregulated market, natural monopolies will arise as bigger players buy out the smaller players, and they will go after smaller and smaller players as their marginal ability to increase their business is eroded by their own success in controlling the market.
Unless you are suggesting regulating ownership of ISPs in a given area in the same way that newspaper and media ownership was regulated by market so that there was not a single monopoly news source, I don't see this changing in such a way that your "everyone should have a choice of providers" utopia will ever come about.
Too bad, the only choice you have is that provider or no provider.
Why is that I wonder? It wouldn't have anything to do with the practice of local governments granting monopolies, would it?
...which they do, because they are not willing to finance the infrastructure costs so that they own the pipes themselves and only lease space on them to the ISPs.
There are such things a municipal telephone companies in the U.S., but they are mostly concentrated in small population-dense areas. The best thing that could ever happen to telecommunications in the U.S. would be for the government to nationalize the infrastructure and then contract out maintenance and operation. There's a reason that Japan and Europe tend to have better net connectivity than most areas of the U.S., and it's not entirely due to the sparse population of most of the U.S. land mass (50% of the U.S. population lives within 50 miles of a coastline).
Dual licenses don't work for open source... without an assignment of rights.
Specifically, if I license something under the "GPL or the Artistic License", and someone takes it under the terms of the GPL, makes modifications to it, and donates those changes back to me, those changes are a derivative of a GPL licensed work, and therefore must be under the GPL. Only if in their donation back to me there is an assignment of rights to me, am I free to relicense the changes out under the Artistic License.
This is what bit SGI, when they thought they were gong to get a boatload of developers jumping on their attempt to step out in front of the open source parade with a baton and lead the parade; almost none of the changes that were made by outside developers were usable to them, due to them lacking an assignment of rights.
It has been burnt-in and certified to operate within manufacturer specifications. The RAM that fails certification for things like amount of time needed to switch frequencies up immediately after switching down in order to save power, etc., is what people buy when they buy "cheaper than Apple RAM".
One of the reasons a Mac Book Pro can operate within the battery budget that it does is that the OS knows what the hardware is capable of, and runs it right at the edge of that capability. Windows, on the other hand, assumes lowest common denominator. This shows in the relative difference in battery life on Windows on a Mac Book Pro vs. Mac OS X on a Mac Book Pro.
"You waste 45 minutes driving while your boss could be working while he is sitting on the train, because he isn't driving."
Cool. Call me back with train tickets as soon as the time I spend working while on the train there and back counts as work hours instead of personal time.
The Linux scheduler historically isn't thread-fair
Specifically, as long as there is work to do in the thread, it doesn't force a context switch; this attribute is why the GIF renderer on Netscape would crash on FreeBSD but not on Linux, because after an involuntary preemption, FreeBSD did not guarantee that the preempted thread would be the one reselected to run; Linux did, but in the process eliminated scheduling fairness for the other threads. So FreeBSD defaults to better thread fairness. This is also why the SQL benchmarks from last year were better on Mac OS X 10.5.5 (Leopard) than on Ubuntu.
If Linux actually honors the SCHED_RR policy, and were to explicitly set it, it would probably do better on the SQL benchmarks; however, it would likely do worse than with SCHED_OTHER or SCHED_FIFO (not sure which is the implementation default in the current Linux scheduler), but certain workloads like to serialize their operations and get better cache line efficiencies from doing so. Oh, and if the default scoping is PTHREAD_SCOPE_SYSTEM vs. PTHREAD_SCOPE_PROCESS will also matter here (again, assuming it's supported).
A decent benchmark would explicitly set the policy and scope to the best value for the system it was running on. The FreeBSD numbers could have easily been better as well, by adjusting these values to explicit values, rather than taking the system defaults.
Slashdot Mirror
It's the Kemp-Kasten Ammendment:
http://www.nchla.org/datasource/idocuments/KempK8503.pdf
I imagine that this portion of the document might have something to do with it not being adopted, since it seems to contravene Roe v. Wade:
Of course, if you're the type who thinks that women should not have rights over their own bodies, and would force them into slavery as incubators, your view might differ from that of the U.S. Supreme Court. http://en.wikipedia.org/wiki/Roe_v._Wade.
China has signed onto that and onto other human rights conventions, yet still practices coercive abortion and sterilization. Peru also carried out a similar program under Alberto Fujimori's administration, which is why they've been trying to extradite him from Japan; Mexico continues a similar policy: http://www.libertadlatina.org/Crisis_Forced_Sterilization.htm
It's not politically correct to point out that this type of document isn't worth spit if it's adopted but not enforced, but it's nevertheless true.
-- Terry
"It makes sense that people who use more should pay more. Why shouldn't the people who use more, pay more?"
When I buy a day pass to Disneyland, I expect that I can go on the Haunted Mansion ride 15 times in a row if I want to. That's why I bought a day pass. If they want me to go on it a maximum of three times, they shouldn't call it a day pass, they should give me a punch-pass instead, and when my 3 punches have been punched, I'm done (unless I buy another punch pass, of course...).
Also, be aware, that every one of those plans that the carriers sell is subject to tariff agreements between them and the PUC (Public Utilities Commission), which gives them the right to charge certain amounts for certain things, and have certain things and not others included in bundles, as part of their regulated monopoly on providing the utility in a given area. This is why your phone bill makes such a big deal about caling out the line items for lifeline service, rural service subsidies, and so on: in exchange for the monopoly in the lucrative markets, they have to serve the non-lucrative markets as well, or they lose the monopoly (and with it, the public rights of way for their wires, where needed, or frequency spectrum, where wires are not in use).
What they are crying about here is the death of their circuit-centric business model, where they used to get to charge for call competition, and for distance between end points (if you ever wondered why there are "services" such as custom ring indication based on the caller and "free" voice mail, wonder no longer). They are deathly afraid that higher data rates and VOIP are going to kill their current cash cows. And they're right: their current cash cows are going to die.
The cell phone boom saved them from becoming nothing more than dumb pipes, where the get to bill based on the diameter, but it only saved them for a short time, and now it's time to pay the piper.
-- Terry
This assumes your desktop is themeable.
"It is more of a social engineering thing - trying to trick unsuspecting users to install a malicious script by hiding it as a theme or screensaver."
Not all desktops are themeable. Apart from security, this is also a good thing when a user calls into technical support, and the person answering the phone needs to be able to tell them which icon they need to click on and where it is/should be located on their desktop. Only desktops which support/permit themes need to worry about trojans in theme bundles.
-- Terry
I am going to guess you are not in the valley.
Here is a 2003 report that indicates non-green buildings put the TCOE (total cost of employee) at about 200% of Salary in non-green buildings:
http://www.usgbc.org/Docs/News/News477.pdf
You are also not including transportation reinbursement, offset contribution to public transport (mandated for businesses with more than a small number of employees in California), and other costs, such as RSUs, stock options, employer contributions to benefits, matching on 401K plans, amortized costs for a 3 year lifetime on multiple computers per employee (I cost ~$4,000/year there) and other costs (mileage reimbursement for someone who lives in SFO and commutes to SJO can go as high as ~$10,000/year).
-- Terry
Managers are on call to make up for employee slack.
They get paid a salary, and they end up on call in case one of their employees babies is stolen by a dingo, or something else happens which results in the employee unavailable. The concept is called coverage, and applies not only to sales people, but to anyone who needs to answer to a pager, a cell phone, a BlackBerry, or is otherwise critical to the business because someone has to be ultimately responsible and fix things when problems arise.
You also have to consider that the costs of an additional person are way out of whack with the TCO for an employee. Your salary accounts for ~1/3 of the cost to the employer for having you around, if you count facilities and energy costs and taxes. This is only going to get worse as taxes go up to cover the costs of government spending on things like the war in Afghanistan and on universal healthcare. So it's a lot easier to increase responsibility at the cost of a small increase in salary and some title inflation than it is to hire more people.
I also expect that other posts in this thread are correct, and that you've been selected as someone to keep on the payroll prior to a pending round of cost-cutting and/or outsourcing.
A "best case" scenario is this is to put you on the hook for the end of the year holiday season.
In situations like this it's generally best to see which way the wind is blowing, and keep your options open, including the option of alternative employment.
Yes, I know people who have quit over being told their vacation was cancelled.
-- Terry
THIS NEEDS TO BE SAID:
Good software does not require documentation.
I attended a number of conferences with Bob (Robert) Wallace, the author of PC-Write, Andrew Fluegelman, the author of PC-Talk, and Jim Button (Knopf), the author of PC-File.
These are largely acknowledged as the first successful shareware programs, and all the authors had basically the same philosophy, but I wlil give you the nutshell version given to me by Bob Wallace:
"I don't sell software; software isn't real; it's all up here", at which point he waved his hands on both sides of his head, and then continued, "I sell manuals".
This paradigm, where the revenue is tied to the documentation, which would be relatively expensive to reproduce, whereas software was relatively easy to copy, is what drove shareware sales for Buttonware, Quicksoft, and others, who depended on their software being copied, and then people coming to them for documentation when the software was hard to use, due to it being intentionally non-intuitive.
As software designers, we've never gotten out of this rut (for the most part), and it's seen as useful to have online help and other extensive documentation for something that should be actually simple, as well as conceptually simple. Ironically, we've even gone to online documentation, either supplied on the distribution media, or on the Internet, or some combination of both (like Microsoft does), in an attempt to reduce publication costs without having to change our design paradigm.
I even have a friend who is actively *proud* of this; he says things like "If it was hard to write, it should be hard to use", and similar gems (he's my friend, but that doesn't stop him from being an idiot).
For tools intended for use by programmers, yes, there should be documentation; there's only so much complexity you can abstract there. But for finished products, if you can't start them up and just start using them for their intended purpose, then there's something wrong with the design.
-- Terry
"if you need a hospital or support for some physical limitation, you won't have to pay through your nose to get it, as this is seen as a basic human right"
Medical but not food as a basic right is amusing; same goes for breathing, water, sex, sleep, homeostasis, and excretion.
Only after you take care of these physiological needs do you get to the next tier of Maslow's hierarchy of needs:
http://en.wikipedia.org/wiki/Maslow's_hierarchy_of_needs
where "health" is located... and that's considered a safety issue, after security of: body, employment, resources, morality, and the family, only just edging out security of property.
-- Terry
The broad leaf plant is Ransoms (Bear) Garlic, sometimes also called "wild garlic".
Here is an almost identical modern illustration:
http://www.kitchengardenseeds.com/cgi-bin/catview.cgi?_fn=Product&_category=145
It was commonly used as a folk medicine in da Vinci's time, which would have made it of interest.
-- Terry
Arrington pulled a SourceForge.
http://www.techcrunch.com/2008/07/21/we-want-a-dead-simple-web-tablet-help-us-build-it/
He declared a project, didn't put much in beyond the idea for the project, and then expected people to flock to the project and build it for him. Very much the same way people declare projects on SourceForge, and then expect volunteer programmers to come out of the woodwork and build whatever their pet project happens to be.
The only difference here is that it involved both hardware and software, and not just software.
I went to the Internet WayBack Machine and read all of the blogs from Fusion Garage (the actual site is currently down, probably intentionally), and it looks like all of the software was done by them, including the OS, and it looks (from Arrington's blogs) like a lot of the hardware was done by Intel.
I don't mean to discount the value of vision or publicity, but really, he intended to Open Source everything about the tablet when he declared the project, and I don't really see a lot of value being taken from him in that case, since he wouldn't be building hardware anyway. The only money would be in margin on the hardware if the software was all out there. A lot of people have posted similar specifications for what they'd like to see in a tablet computer, and the only difference between them and Arrington is Arrington has a lot of self-publicity and got a startup to bite on the bait to actually build the thing.
Arrington might have some arguments with regard to industrial design, but the prototype hardware was not built by him, and the software that made that hardware live and breathe was definitely not his.
I've worked at combo hardware/software startups, and I've worked at software-only startups (including my own), and universally, the hardware in the hardware/software startups was all about minimal COGS and industrial design (being at Apple now, that's pretty much all there is). The value-add over commodity hardware is that it isn't "cheapest vendor of the part of the day" (so the hardware is reliable and not crap because of constantly retooled assembly lines), and it's the software. When Apple builds a laptop, it doesn't build hardware, and it doesn't build software, it builds systems. The people who don't get that and churn out 1.5%-4% margin crap do so at their peril.
My reading of things is that Arrington is no Jonathan Ive, and he's no Steve Jobs when it comes to design of hardware or software.
Fusion Garage may have taken his idea and run with it.. and they want to cut him in on profits from something, the intrinsic value of which he intended to give away for no profit, but they don't seem to be ripping him off to do it, although they do seem to be leveraging as much as they can to get him to accept a minority role with regard to what he brings to the table (which, per the above, by my reckoning, isn't much; sorry, Arrington).
As more than one V.C. has told me in the past, the point is not the idea; there are millions of good ideas that go unfunded all the time (I'll point at SourceForge again, where "funding" equates to "provision of manpower necessary to complete a project"); what a V.C. funds is the ability to execute on a vision, no matter whose vision it is, and the team behind that ability to execute on the vision and bring a product to market. 1 in 10 entrepreneurs get funded; 1 in 10 of those fail in the first year. That's only a 1 in 100 chance of being around after a year.
Arrington's failure is no less spectacular than anyone else's in that 99 out of 100 failures, he's just made it more public by ranting about it.
Ironically, the idea may still not be a failure, merely a failure on his part to control the thing which was built on his (and a lot of other peoples similar) idea, if Fusion Garage or someone else simply continues on and executes on it.
Good luck in your future endeavors, but don't think that by declaring an idea publically that you've built or created anything.
-- Terry
Actually he IS being taxed..
He's being taxed by the amount it is costing him to record keep, collect, and forward the tax to the state in question.
Brick and mortar stores are taxed in the same amount, but it's very easy to record-keep, as you program up the cash registers and load the (small) exception table for the items you have to deal with. In some cases, as in prepared food at supermarkets, this adds a requirement that the person making the purchase be asked whether or not they intend to eat at tables set aside at the store, or whether they intend to take their purchase home. The answer to this question changes the tax rate in some places.
For an Internet retailer, where it's still not legally clear whether the transaction is taking place in the state of the purchaser or the seller, this is onerous in the extreme. Unlike a brick and mortar company, the transaction is either at the sellers location or the drop-ship warehouse (where consideration is finally exchanged resulting in a sales contract; this is supposedly one of the "outs" Amazon is "exploiting"), or it's at every sellers location at the time of the purchase. For example, on an airplane at 36,000 feet somewhere over the midwest.
The part which is truly onerous to the retailer is that there are, at least on the last database update I'm aware of, 160,000+ sales tax rates at various locations in the U.S., and it is nearly impossible to correlate the location with the GIS information such that you can pick the right sales tax(es) to collect for a given location.
Mail order purchases (before the Internet was such an avenue of commerce) are the reason there are state "use tax" laws, and it's the responsibility of the purchaser to pay the tax in the locality that the purchase is intended to be used.
Amazon is being given a bum rap here. Because they are a single "one stop shopping" target for the government to use as enforcement proxy via unfunded mandate ("get Amazon and you get all the tax on all the traffic through Amazon, and get them to pay for the collection process"), they have a huge bullseye painted on their back. Although the idea of normalizing sales tax across all venues, as was suggested in the article, is very attractive on the face of it, it's unlikely to ever happen.
-- Terry
"Fly" programming language ...and just like the original thing, it has garbage collection!
-- Terry
Hogwash: Building codes are regulatory, just like FCC and FAA rules, or public utilities commission rules. The only laws involved are usually rather simple and to the point in delegating the authority to an administrative agency generally controlled by the executive branch of the appropriate government.
As far as tax law, it's only necessary to not have a graduated flat tax (e.g. taxed on what you earn above minimum was times 2080 hours + $1) if you are intent on hiding your legislative cronyism, malfeasance, kickbacks, and unfunded mandates in the tax code. If you want to legislate social policy, then be honest and legislate social policy, and if what you do is unpopular, you don't get reelected.
Also, I remember a debate from my college days when it was suggested that the best form of government was in fact a benevolent dictatorship. No thank you.
P.S.: I'd still like someone to explain to me why the disincentive for second degree murder should be less than the disincentive for first degree murder; the victim is still just as dead, right?
-- Terry
Unfortunately they blew the System call interface on Mac OS X.
Mac OS X defines the system call API as the top of Libc, not the bottom at the user/kernel boundary.
You can see this in their implementation of the "kill" and the "posix_spawn" system calls, but there are tons of others:
SYS_KILL = 37; // { int kill(int pid, int signum, int posix); }
SYS_POSIX_SPAWN = 244; // { int posix_spawn(pid_t *pid, const char *path, const struct _posix_spawn_args_desc *adesc, char **argv, char **envp); }
Most of the places they say "NO_SYSCALL_STUB" are actually decorated version of system calls for binary backward compatibility. If they ever include Framework support for certain existing frameworks, they are going to find that their use of 32 bit libraries may result in unexpected-by-the-framework cancellation points that need to not be there.
The kill system call changed from older versions of the OS to ensure POSIX compliance for one compilation environment, and historical behaviour; in fact, for the 32 bit version of libc, there are 189 decorated symbols for doing everything from translating error codes to getting variant behaviour to enforcing POSIX semantics in user space in libSystem itself, and there are 74 variants for cancellation points in both 32 and 64 bits.
Changing the contract between libSystem and the kernel under the covers to fix bugs happens often. The posix_spawn() system call is a good example of that. Apple has extended spawn several times by modifying the _posix_spawn_args_desc. The code itself uses accessors and mutators on an allocate descriptor. It can change (and does) size when new extensions are added.
Basically by providing their own direct system call interface, they've guaranteed that they will constantly be chasing a moving target, and they've lost a lot of the default POSIX semantics that are implemented at the libSystem level. For kill and similar calls, they've lost the go code portability between OSs by doing this.
Finally, they need a type mechanism so that the unsupported interfaces can be conditionalized based on the UNIX standard permitting non-implementation of certain interfaces; otherwise they will run into things like unnamed semaphores which will throw a SIGSYS, which will abort their programs.
-- Terry
So, remind me again: why exactly is it not possible to implement all that in a package manager and we need to have a Really Fat ELF?
Because Linux distributions can not agree on a single GUI technology, let alone a package manager.
-- Terry
The problem with fixating on something like ELF for fatness is that it's just the tip of the iceberg.
There's much more to the question of whether or not something will run on an arbitrary copy of Linux than the CPU arch.
Why do you think one of the discriminators in a fat binary can't be a distribution identifier, such that there are fat slices for supporting Debian, RedHat, Ubuntu, etc., all from the dame binary file?
Or that they can't have different slices in the fat binary for Gnome vs. KDE, or desktop vs. Android, and so on?
Also, the arguments about disk space are specious; at least in the Mac OS X world, there is a utility called "lipo" which will pull apart a fat bnary into only the pieces you need/want to install. Typically you only install the actually fat binary on server systems, where the software has to be able to run on multiple client machines, and otherwise you run scripts to reclaim disk space (or in the case of an embedded device, you run them over the install image before you install them).
Same goes for an Apple fat binary really...
Obligatory disclaimer: I am the person who maintains the fat binary loader in the Mac OS X kernel.
-- Terry
"mankind's first permanent space colony"
Someone needs to tell Mark Harris that the ISS is scheduled to be deorbited 1Q 2016 before he moves in to his condo there.
http://en.wikipedia.org/wiki/International_Space_Station
-- Terry
I had hoped that the best scientist and engineers would be motivated by something more than just money.
I bet a lot of management types are hoping the same thing.
-- Terry
So say we find a bug...
Do we disclose it, or do we sell it to the highest bidder?
I mean this assumes the bug will be discovered by at least one honest person who chooses to disclose, right?
-- Terry
I had a nice ISP...
They were bought by EarthLink.
So I changed ISPs to another nice ISP.
They were bought by a different company.
That company was then bought by EarthLink.
I changed to a third ISP.
A while later, they were bought by EarthLink.
In any unregulated market, natural monopolies will arise as bigger players buy out the smaller players, and they will go after smaller and smaller players as their marginal ability to increase their business is eroded by their own success in controlling the market.
Unless you are suggesting regulating ownership of ISPs in a given area in the same way that newspaper and media ownership was regulated by market so that there was not a single monopoly news source, I don't see this changing in such a way that your "everyone should have a choice of providers" utopia will ever come about.
-- Terry
Too bad, the only choice you have is that provider or no provider.
Why is that I wonder? It wouldn't have anything to do with the practice of local governments granting monopolies, would it?
...which they do, because they are not willing to finance the infrastructure costs so that they own the pipes themselves and only lease space on them to the ISPs.
There are such things a municipal telephone companies in the U.S., but they are mostly concentrated in small population-dense areas. The best thing that could ever happen to telecommunications in the U.S. would be for the government to nationalize the infrastructure and then contract out maintenance and operation. There's a reason that Japan and Europe tend to have better net connectivity than most areas of the U.S., and it's not entirely due to the sparse population of most of the U.S. land mass (50% of the U.S. population lives within 50 miles of a coastline).
-- Terry
"You can dual-license something all you want."
Dual licenses don't work for open source ... without an assignment of rights.
Specifically, if I license something under the "GPL or the Artistic License", and someone takes it under the terms of the GPL, makes modifications to it, and donates those changes back to me, those changes are a derivative of a GPL licensed work, and therefore must be under the GPL. Only if in their donation back to me there is an assignment of rights to me, am I free to relicense the changes out under the Artistic License.
This is what bit SGI, when they thought they were gong to get a boatload of developers jumping on their attempt to step out in front of the open source parade with a baton and lead the parade; almost none of the changes that were made by outside developers were usable to them, due to them lacking an assignment of rights.
-- Terry
The RAM is definitely not the same
It has been burnt-in and certified to operate within manufacturer specifications. The RAM that fails certification for things like amount of time needed to switch frequencies up immediately after switching down in order to save power, etc., is what people buy when they buy "cheaper than Apple RAM".
One of the reasons a Mac Book Pro can operate within the battery budget that it does is that the OS knows what the hardware is capable of, and runs it right at the edge of that capability. Windows, on the other hand, assumes lowest common denominator. This shows in the relative difference in battery life on Windows on a Mac Book Pro vs. Mac OS X on a Mac Book Pro.
-- Terry
Consider getting Tom Hall to publish the source
Otherwise no BioEdit fixes unless Tom can get around to them.
-- Terry
Working while riding...
"You waste 45 minutes driving while your boss could be working while he is sitting on the train, because he isn't driving."
Cool. Call me back with train tickets as soon as the time I spend working while on the train there and back counts as work hours instead of personal time.
-- Terry
The Linux scheduler historically isn't thread-fair
Specifically, as long as there is work to do in the thread, it doesn't force a context switch; this attribute is why the GIF renderer on Netscape would crash on FreeBSD but not on Linux, because after an involuntary preemption, FreeBSD did not guarantee that the preempted thread would be the one reselected to run; Linux did, but in the process eliminated scheduling fairness for the other threads. So FreeBSD defaults to better thread fairness. This is also why the SQL benchmarks from last year were better on Mac OS X 10.5.5 (Leopard) than on Ubuntu.
If Linux actually honors the SCHED_RR policy, and were to explicitly set it, it would probably do better on the SQL benchmarks; however, it would likely do worse than with SCHED_OTHER or SCHED_FIFO (not sure which is the implementation default in the current Linux scheduler), but certain workloads like to serialize their operations and get better cache line efficiencies from doing so. Oh, and if the default scoping is PTHREAD_SCOPE_SYSTEM vs. PTHREAD_SCOPE_PROCESS will also matter here (again, assuming it's supported).
A decent benchmark would explicitly set the policy and scope to the best value for the system it was running on. The FreeBSD numbers could have easily been better as well, by adjusting these values to explicit values, rather than taking the system defaults.
-- Terry