The Columbia Accident Investigation Board has recommended that NASA establish an independent Technical Engineering Authority. This would put responsibility for technical matters where it rightly belongs -- with the engineers who, because they know how the space shuttle was designed, also know best how it can fail. Without that knowledge, another fatal accident is inevitable.
Appointment to the TEA will be a political process controlled by management. The so called engineers will be nothing but management lackeys and the same problems will occur. The real engineers still won't have a voice.
Agreed. In writing, however you should not send it via snail mail, use fax. To two places,
the CEOs office and tech support.
Get a fax number from the tech support person for submitting a report. Ask for the fax number of the CEOs office. If they won't provide it, hang up, call the front office
and ask them for the number without explaining
why. Most will provide this number.
The report should be as simple as possible
with enough details that they can verify that
they are under attack.
Provide a method for contacting you by phone.
Identify yourself by your customer ID.
The cover page to the CEO should state clearly
in a short note:
that their ISP service is under attack now and
passwords and possibly credit card information
are being stolen from customers.
If you must include speculation on how this
has been accomplished put it at the end of the
report, along with speculation of damages you
don't really know about.
You've done your bit. If the ISP doesn't do theirs after this, find another.
You should follow this up later with an email
or a normal letter asking to be informed of the
action that was taken and the results.
In 1992 I was briefly mucking around in the set top
box world. It became apparent that people behave as if they have
an entertainment budget. At the time it was something similar to:
$35/mth - Cable TV
$30/mth - Movies (Rentals, Theaters..)
$250/yr - Equipment, TV, VCR, Set Top Box, GameBox
I don't believe it has changed much. If the person intended on buying a new TV/VCR every three years
then they would only spend $750 on both. Every ten years - $2500.
The other factor was an equipment threashold of around $250. If it
was more expensive than this people thought too much about the purchase,
below it they would buy on impulse. This doomed lots of game boxes that
tried to sell above this price.
The trouble with introducing a new service (payTV) or a new product
(like a set top box) is that it had to fit into this budget by replacing
something that was already there. HDTV and DTV are facing the same problem.
Tivo clearly got it. Pricing less than $250, replaced the VCR.
DTV tuner (4:3) is at least $100 more expensive (now) than the normal
cable TV tuner - for what additional value? That is $100 they have to find from somewhere else in
the budget. People are good at maximizing their
entertainment value within the budget.
HDTV 16:9 tubes are at least $500 more expensive - too much thinking.
The FCC is trying to speed up the adoption but are proving to be
rather toothless, the old deadline was the end of this year. 2006? 2010
more like it.
The broadcast companies were given the option of using the 6Mhz bandwidth for
one HDTV signal or 5 DTV signals... guess which one they prefer? Where can you sell more advertizing space.
There is also the confusion of will DVD/PVR work with the new HDTV set. The cable companies aren't helping either. Why would they want to replace their infrastructure, they already own
their customers.
Until HDTV reaches these threasholds and budget constraints
it's not time to buy for most
people. The FCC won't, or won't be allowed to, force the switch.
It would have been faster and less traumatic
if the FCC had said "on YYMMDD everything HDTV
or DTV" a couple of years ago instead of the
slow phase in. Networks would have been scrambling to
subsidize the converters and new TVs so their
audience and advertizing revenue didn't get hurt.
Jim Gettys
posted
the font family "Vera" was derived
from: "Prima". I obviously got the list
of fonts wrong. If you want to see them try:
http://www.myfonts.com/fonts/bitstream/prima-sans/ $24.75
As you can see - there are "bold oblique" versions
of "Sans" and "Mono" and only a bold version of the serif font. No symbols. Unfortunately this will probably
mean that the Serif font will look bad when using
italic/oblique and will turn people off using it
for printing documents. Still, they are nice fonts
on the screen and Jim Lyles is cleaning them up.
The 10 fonts are all from the same family "Vera". Hopefully they look good enough on the screen and on paper that people won't mind using them.
There are at three major styles "Serif", "Sans" and "Mono", with three minor styles "regular", "italic" and "bold". Thats 9 fonts. I would guess the 10th is a set of symbols.
I haven't been able to find samples of the family on either bitstreams site or myFonts.com
so I would also guess that the font is renamed
for copyright purposes from something else.
Infect the net in 15 minutes. 15 seconds if we believe some other papers.
Have the worm/virus announce itself by killing off the most common other viruses. That's all. The logs would go quiet as Nimda, Code Red, Klez and Bugbear stop. It would take a while for people to figure out that a new overlord was in town.
If all it did was 'Benevolent' would there be any action to stop it? Whoever was controlling it would be outside the law.
Appple assumes that the mouse has only one button. In that case a single menu at the top
of the screen might be more productive.
Two buttons allows instant access to a menu without moving the mouse. If this was standardized like the menu bar layout then I would expect the right mouse button menu to be much faster than the one menu at the top.
There is a much bigger issue here. Digital images should not be admissible in court. Anyone with a simple image editor can add/remove things. Photographs at least have a negative that will show any tampering.
Until digital cameras "sign" every image with
a unique id, a time stamp and a hash of the
image don't expect the courts to accept the images as evidence. The camera should also
have some way of verifying that it took the
photo image and that it hasn't been altered.
Digital video would have to have every frame
signed. Once you can verify that the original
video sources are free of tampering then there
shouldn't be any problem with showing a 3D
fly through providing the court can locate the
original view that contributed to some item of interest.
Its one one the main reasons projects fail in the long term
Can you point to one project that has succeeded because of HCI? In my experience all the recommendations by HCI types were irrelevant to the success of the project.
Way back in the late 1980's a japanese graphical multiplayer game existed. It didn't have a real currency but players could collect different heads. A real market for heads developed, people held parties to show their friends their head collections. A marfia existed that would target new players and steal their heads. It was clear even then that people were getting hooked.
The comparison of multi-player games to chemical additions is wrong only in the mechanism. It's much closer to gambling on races, sports events, or at casinos. You have the same ideas of wealth, and the risk of something with a promise of reward. You can build a 'life' that is different to your real world. Going to Vegas is appealing not just for the gambling but also for the total disconnect from your real life. For a couple of days you can be in a glamorous life where people (casino staff) treat you well.
The companies that host and develop these games do indead build into them things that appeal to the players. It is unfortunate that these things are also the same things that addictive personalities get hooked by.
The real question is whether these companies have a responsibility to society. We as a society have placed requirements on casinos to watch out for excessive gamblers. Race tracks have lists of people for whom the courts have banned placing bets. Bars are required to cut off drunks and can be held liable for drunk drivers.
All of the above we as a society have decided are the responsibility of the companies that provide these pleasures. Multi-player game companies will eventually have to accept some form of social responsibility to help stop the addicts.
I do not believe Sony is responisble for this suicide. I do believe that the courts should be able to get the records related to a player for the investigation of death or crimes. I also believe that we as a society will require these companies to do something to protect the addicts from themselves, not because the companies are at fault but because we consider the cost to society too high.
Watching a normal movie the wireless access doesn't provide much useful unless you are completely bored and want the sports results. Ordering food and drinks that get bought to the tables is mildly useful as a way of reducing noise and interruptions.
The best use of this type of improvement are group presentations where a short movie can be played to an audience who can then submit questions and observations about the presentation while it occurs to them.
Open discussions where the audience can submit questions to a moderator, the moderator choses the interesting questions to the speaker who can control what is shown on the large screen, call up movie clips and provide references to the audience via the network.
The real question is does a movie theater provide a cost advantage for setting up this type of conference over say a conference room in a typical venue?
For 50 million we could easily produce this device using readily available components. It's really just an assembly job. You don't have to invent a single thing.
Start manufacturing the devices in 2008. This gives you 2 years to work out the bugs.
The software could easily be written today and would run at acceptable speeds on today's PDAs. Keep a small team of 4-5 programers to write the software over the next 8 years.
The hardware either doubles in speed, halves in size, or halves in price every 18 months. If you wait until 2008 before manufacturing the devices they could be 1/16th the cost. In practice it'd be more like 1/8th given some of the other components like the case and battery.
If I can buy a Palm for $200 retail, the cost is well below $100 today. In 2008 the cost of the harware will be below $50 easily.
Forget wireless - the coverage will never be that great in low population density areas. Instead every night plug the thing into a phone line while it recharges. Upload the collected surveys and download the next day's targets and maps etc.
The whole device could be assembled from off the shelf components and manufactured in 2008. You'd only need 1 or 2 people to co-ordinate the manufacturing.
The only problem would be managing the politics. Keeping the specifications of what the device should be capable of to what is needed for the task. Every new person on the government side is going to want their own pet idea implemented.
Time to start a small company of no more than 10 people. Any takers?
Unix was a step forward from OSs of the time that had complex filesystems
that supported metadata and structured files. We are just repeating the
mistakes of 40 years ago because most programers have no idea of the history
of operating systems.
As soon as you need to move a file between two machines you have to invent
a flattened form of the file and all it's metadata into a bytestream. This
bytestream can then be sent over a network, or modem, or put onto a tape. Unix
removed the complexity of doing this from the OS and made every file take this
flattened form. If a user program wanted to impose a more complex structure
inside the file that was ok too.
Once everything is a bytestream to the OS other things like pipes between
processes become easy and simple. It doesn't matter if the processes are on
the same machine or on the other side of a network.
It turned out that the majority of data people stored did not need anything
more complex than a flat file with some header information at the front. Unix
was a good balance.
When we need a finer level of access to data we turn to databases. They are
a whole extra level of complexity.
Sarcasm: Forget about embedding the web browser into the OS let us put
the whole database in there as well. If we just apply a simple encryption
scheme to the content we can use the law to prosecute illegal copying of media
too. We will own the world...
Who decided that the value of an ad-free Slashdot was equivalent to that of a typical magazine?
There is a general failure in all of the new entertainment ventures. They forget that people have a real or imagined budget for their entertainment:
Spending $20 here means that something else will lose $20. A decision to buy a new magazine means I'll drop one of the other ones I used to pay for.
It's not just money, it is also time. I have a limited amount of time for entertainment. Time I spend watching a DVD is time I don't spend watching cable TV.
People don't like metered services. It's too complicated and unpredictable. It doesn't allow the easy comparison of two services. Look at the failure of metered bandwidth ISPs.
While you may be putting a price on the value of advertising, because you are counting page views you are encouraging a mental model of pricing the content. The content comes from us and we'll start wanting something in return for moderating and posting. If people that are active and valuable contributors start to feel that they are being used they will go elsewhere where they feel that they are not being used. There will always be places on the web where people can post their opinions for free.
By ignoring economics, human psychology and history, Slashdot will become another lesson on how-not-to-make-it-pay.
The author has over simplified multiple things in his article. Every simplification has enabled him to claim things that on the surface seem reasonable but in practice are false. To further increase the noise he has a lot of his 'facts' and 'assumptions' incorrect.
1) "J2EE and.Net--will essentially control the programming languages market"
How does he define the programming languages market? There are more people employed writing "vertical applications". Their choice of language and libraries is typically dictated by legacy: fortran, cobol, lisp, C, pascal, etc. What about embedded systems: assembly language, C, C++, Forth. High performance computing using vectors and parallel algorithms have a whole set of specialized languages.
2) "Still, it is an amazing achievement to be able to support different languages on the same runtime, right? I certainly think it is, but others would disagree."
Where has this guy been? By the early 1980s there were compilers for Pascal, C and Fortran that compiled to P-code (a common runtime) and were either compiled to native machine code or ran in an interpreter. Not a new concept. It is just that people are ready to accept the cost today.
3) discussion of what is the natural descendant of C++.
Who really cares? Once you've chosen a language to write in, it doesn't matter how it came to be. It supports a set of programming constructs and has a number of libraries available for use (that may support the type of program you are writing). The evolution of the language is completely irrelevant.
Language designers are very aware of other languages (more so than Liotta) and will borrow/steal ideas and syntax that they like. The language is designed with a particular purpose in mind. There are literally hundreds of domain specific languages that work better than C#, Java, VB for their intended task. They may be similar to another language, or use similar constructs but they are not the same.
4) "With.Net, there is only a single runtime (functionality), but different language syntaxes (look and feel) can make use of it."
There has always been a single runtime, the machine code. It hasn't prevented languages from having different semantics from one another..Net and JVM won't prevent it either. The syntax of a language is almost accidental from the point of view of the language designer.
I could keep going but I'm also over simplifying so I'll stop.
Having the source code available enables you to fix the bugs that matter to your business. Even if you can't do it yourself, you can pay me to fix it for you. If you are not willing to pay then the problem is not affecting your bottom line.
Without the source code you have one choice: go to the application developer; hope they will pay attention to your problem, or pay them a huge fee to produce a one-off version for you. Their priorities do not match yours.
Any business would favor OSS if they thought through the issues. What manufacturer/business really wants to be dependant on a single source supplier who receives revenue based on flaws the supplier leaves in the product?
The real issue is that while I can implement a tail call implementation using the existing CLI/JVM instructions it isn't very efficient. Without a tail call instruction in the JVM/CLI the runtime compiler/optimizer is prevented from recognizing the tail call and emitting better machine code.
Programs written in functional language use call instructions much more than traditional ALGOL like programs. The overhead of the call matters much more than you would expect.
Microsoft's implementation of the CLI could (and probably already does) have extra instructions that are not documented publicly. MS could then produce a.NET module that uses them but doesn't work on other implementations of the CLI.
If you want to gain the same advantage then you'll have to licence the info from MS because they'll have the instructions patented.
If the module is critical to the.NET world then MS has effectively locked people onto their platform because that module doesn't run on other CLI platforms.
This is exactly what MS was trying with the Java VM before Sun stopped them.
Younger audiences do have 'traditions'. Think back to your early school days. There were games and behaviors that were passed on from one year to the next. These 'traditions' have been propagating through this age group for decades.
Gameboy is firmly established in these younger age groups (esp under 12).
- AndrewN
The article misses the real problem with computer programing: the immaturity of the industry.
I believe that there are three types of programers:-
Programmers: May have a degree in some computer related field, but in practice do not use the 'science' or 'engineering' in their daily work. They are always under some time pressure to get things done. Engineers: Use formal specifications and mathematics to design and build programs. Most of them are not working in the commercial world. (No UML is not a formal specification.) Scientists: Look for better algorithms, write papers about their research, typically employed by universities.
So here is the problem:
'programers' write code that is only as good as the best code they have been exposed to. They learn by osmosis. They almost never read research papers, instead they read opinion pieces about the tools they do use.
'engineers' write code that is correct and bug free and take forever to get anything done. As a result they are not wanted in the typical software development team.
'scientists' write code that illustrates their particular area. It's often bug ridden when outside of the test domain and unmaintainable.
Both 'engineers' and 'scientists' write papers for their peers and are not targeted at 'programers' in the field. They may as well be using a different language when it comes to trying to comprehend some of the papers.
There is very little cross over between the different schools of programing. Almost none of the thousands of engineering/science papers published have had any effect on the world of practical programing.
While there are various associations of computing professionals, not one of them provides an industry recognized certification that tests the knowledge of a programmer/engineer. (please no MSCE/MCSE wars)
Given the above it's not suprising that for any given team of programers they will produce code that contains bugs, is insecure, and doesn't use 'best practices'. The tools we use have been written by us, not by engineers, and fulfill a different purpose, not correctness or security but speed of development.
One other point: Why spend time on security of an application when you know the OS and the other applications are gaping security holes anyway. Why put an expensive shiny lock on the back door when the front door has a "please come in" sign?
- AndrewN
Not going to happen for a long time
on
Think And Click
·
· Score: 1
The biggest problem that still needs to be solved is the implants that pick up signals cause the surrounding tissue to degrade and become useless after about 6 months.
Spend the time to learn how to use it, then it stops working.
However once they sort this out - sign me up.
The PDA market isn't looking for a computer that needs to be replaced every three years. You should think of it as a consumer device like a VCR or TV, replaced every 5 to 7 years. Hopefully Palm realizes this. They should start to see the repeat business in the next 2 to 3 years.
I've had my Palm Pro since 97, with one hardware upgrade to add extra memory and the IR port. The only reason I bought it was that I knew if I broke or lost it I could restore everything to the latest model. I have a fixed set of applications that cover all my daily note taking, contacts and reminders.
Color would be nice but there is no extra functionality with it. A higher resolution would help reading text documents. However it works the way it is. I'd rather have the battery life.
If I need more complex processing or better graphics, and so on, I'll switch to my laptop or workstation. The PocketPC holds no appeal for me.
Palm made a consumer device that matched what people wanted. Repeat business will happen if Palm keeps their head, brand loyalty still exists.
If you are a hardware manufacturer then the OS software is an expence you have to pay for. Both Sun and Apple treat their software divisions as a non-revenue producing expences.
Separating the OS and the Hardware groups allows them to follow different and competing sales agendas. The OS group can now support the hardware platforms (Sony, IBM, Handspring, Nokia & Palm) that the market is willing to buy. The hardware group can focus on producing the most cost effective platform for a PDA.
Slashdot Mirror
Sandwiching and Suspensions is a later article by the same guy using elastic to suspend the drives to reduce noise transmission.
Appointment to the TEA will be a political process controlled by management. The so called engineers will be nothing but management lackeys and the same problems will occur. The real engineers still won't have a voice.
Agreed. In writing, however you should not send it via snail mail, use fax. To two places, the CEOs office and tech support.
Get a fax number from the tech support person for submitting a report. Ask for the fax number of the CEOs office. If they won't provide it, hang up, call the front office and ask them for the number without explaining why. Most will provide this number.
The report should be as simple as possible with enough details that they can verify that they are under attack.
Provide a method for contacting you by phone. Identify yourself by your customer ID.
The cover page to the CEO should state clearly in a short note: that their ISP service is under attack now and passwords and possibly credit card information are being stolen from customers.
If you must include speculation on how this has been accomplished put it at the end of the report, along with speculation of damages you don't really know about.
You've done your bit. If the ISP doesn't do theirs after this, find another.
You should follow this up later with an email or a normal letter asking to be informed of the action that was taken and the results.
In 1992 I was briefly mucking around in the set top box world. It became apparent that people behave as if they have an entertainment budget. At the time it was something similar to:
I don't believe it has changed much. If the person intended on buying a new TV/VCR every three years then they would only spend $750 on both. Every ten years - $2500.
The other factor was an equipment threashold of around $250. If it was more expensive than this people thought too much about the purchase, below it they would buy on impulse. This doomed lots of game boxes that tried to sell above this price.
The trouble with introducing a new service (payTV) or a new product (like a set top box) is that it had to fit into this budget by replacing something that was already there. HDTV and DTV are facing the same problem.
Tivo clearly got it. Pricing less than $250, replaced the VCR.
DTV tuner (4:3) is at least $100 more expensive (now) than the normal cable TV tuner - for what additional value? That is $100 they have to find from somewhere else in the budget. People are good at maximizing their entertainment value within the budget.
HDTV 16:9 tubes are at least $500 more expensive - too much thinking.
The FCC is trying to speed up the adoption but are proving to be rather toothless, the old deadline was the end of this year. 2006? 2010 more like it.
The broadcast companies were given the option of using the 6Mhz bandwidth for one HDTV signal or 5 DTV signals ... guess which one they prefer? Where can you sell more advertizing space.
There is also the confusion of will DVD/PVR work with the new HDTV set. The cable companies aren't helping either. Why would they want to replace their infrastructure, they already own their customers.
Until HDTV reaches these threasholds and budget constraints it's not time to buy for most people. The FCC won't, or won't be allowed to, force the switch.
It would have been faster and less traumatic if the FCC had said "on YYMMDD everything HDTV or DTV" a couple of years ago instead of the slow phase in. Networks would have been scrambling to subsidize the converters and new TVs so their audience and advertizing revenue didn't get hurt.
Bit of a ramble ...
Jim Gettys posted the font family "Vera" was derived from: "Prima". I obviously got the list of fonts wrong./ $24.75
If you want to see them try:
http://www.myfonts.com/fonts/bitstream/prima-sans
- Vera Sans
- Vera Sans Oblique
- Vera Sans Bold
- Vera Sans Bold Oblique
Designer: Jim Lyleshttp://www.myfonts.com/fonts/bitstream/prima-sans- monospace/ $99
- Vera Sans Monospace
- VeraSansMono BT Oblique
- VeraSansMono BT Bold
- VeraSansMono BT Bold Oblique
Designer: Sue Zafaranahttp://www.myfonts.com/fonts/bitstream/prima-serif / $49.50
- Vera Serif
- Vera Serif Bold
Designer: Jim LylesAs you can see - there are "bold oblique" versions of "Sans" and "Mono" and only a bold version of the serif font. No symbols. Unfortunately this will probably mean that the Serif font will look bad when using italic/oblique and will turn people off using it for printing documents. Still, they are nice fonts on the screen and Jim Lyles is cleaning them up.
I thought the pricing for Prima was interesting.
The 10 fonts are all from the same family "Vera". Hopefully they look good enough on the screen and on paper that people won't mind using them.
There are at three major styles "Serif", "Sans" and "Mono", with three minor styles "regular", "italic" and "bold". Thats 9 fonts. I would guess the 10th is a set of symbols.
I haven't been able to find samples of the family on either bitstreams site or myFonts.com so I would also guess that the font is renamed for copyright purposes from something else.
Infect the net in 15 minutes. 15 seconds if we believe some other papers.
Have the worm/virus announce itself by killing off the most common other viruses. That's all. The logs would go quiet as Nimda, Code Red, Klez and Bugbear stop. It would take a while for people to figure out that a new overlord was in town.
If all it did was 'Benevolent' would there be any action to stop it? Whoever was controlling it would be outside the law.
Appple assumes that the mouse has only one button. In that case a single menu at the top of the screen might be more productive.
Two buttons allows instant access to a menu without moving the mouse. If this was standardized like the menu bar layout then I would expect the right mouse button menu to be much faster than the one menu at the top.
There is a much bigger issue here. Digital images should not be admissible in court. Anyone with a simple image editor can add/remove things. Photographs at least have a negative that will show any tampering.
Until digital cameras "sign" every image with a unique id, a time stamp and a hash of the image don't expect the courts to accept the images as evidence. The camera should also have some way of verifying that it took the photo image and that it hasn't been altered.
Digital video would have to have every frame signed. Once you can verify that the original video sources are free of tampering then there shouldn't be any problem with showing a 3D fly through providing the court can locate the original view that contributed to some item of interest.
Its one one the main reasons projects fail in the long term
Can you point to one project that has succeeded because of HCI? In my experience all the recommendations by HCI types were irrelevant to the success of the project.
Way back in the late 1980's a japanese graphical multiplayer game existed. It didn't have a real currency but players could collect different heads. A real market for heads developed, people held parties to show their friends their head collections. A marfia existed that would target new players and steal their heads. It was clear even then that people were getting hooked.
The comparison of multi-player games to chemical additions is wrong only in the mechanism. It's much closer to gambling on races, sports events, or at casinos. You have the same ideas of wealth, and the risk of something with a promise of reward. You can build a 'life' that is different to your real world. Going to Vegas is appealing not just for the gambling but also for the total disconnect from your real life. For a couple of days you can be in a glamorous life where people (casino staff) treat you well.
The companies that host and develop these games do indead build into them things that appeal to the players. It is unfortunate that these things are also the same things that addictive personalities get hooked by.
The real question is whether these companies have a responsibility to society. We as a society have placed requirements on casinos to watch out for excessive gamblers. Race tracks have lists of people for whom the courts have banned placing bets. Bars are required to cut off drunks and can be held liable for drunk drivers.
All of the above we as a society have decided are the responsibility of the companies that provide these pleasures. Multi-player game companies will eventually have to accept some form of social responsibility to help stop the addicts.
I do not believe Sony is responisble for this suicide. I do believe that the courts should be able to get the records related to a player for the investigation of death or crimes. I also believe that we as a society will require these companies to do something to protect the addicts from themselves, not because the companies are at fault but because we consider the cost to society too high.
Watching a normal movie the wireless access doesn't provide much useful unless you are completely bored and want the sports results. Ordering food and drinks that get bought to the tables is mildly useful as a way of reducing noise and interruptions.
The best use of this type of improvement are group presentations where a short movie can be played to an audience who can then submit questions and observations about the presentation while it occurs to them.
Open discussions where the audience can submit questions to a moderator, the moderator choses the interesting questions to the speaker who can control what is shown on the large screen, call up movie clips and provide references to the audience via the network.
The real question is does a movie theater provide a cost advantage for setting up this type of conference over say a conference room in a typical venue?
For 50 million we could easily produce this device using readily available components. It's really just an assembly job. You don't have to invent a single thing.
Start manufacturing the devices in 2008. This gives you 2 years to work out the bugs.
The software could easily be written today and would run at acceptable speeds on today's PDAs. Keep a small team of 4-5 programers to write the software over the next 8 years.
The hardware either doubles in speed, halves in size, or halves in price every 18 months. If you wait until 2008 before manufacturing the devices they could be 1/16th the cost. In practice it'd be more like 1/8th given some of the other components like the case and battery. If I can buy a Palm for $200 retail, the cost is well below $100 today. In 2008 the cost of the harware will be below $50 easily.
Forget wireless - the coverage will never be that great in low population density areas. Instead every night plug the thing into a phone line while it recharges. Upload the collected surveys and download the next day's targets and maps etc.
The whole device could be assembled from off the shelf components and manufactured in 2008. You'd only need 1 or 2 people to co-ordinate the manufacturing.
The only problem would be managing the politics. Keeping the specifications of what the device should be capable of to what is needed for the task. Every new person on the government side is going to want their own pet idea implemented.
Time to start a small company of no more than 10 people. Any takers?
Unix was a step forward from OSs of the time that had complex filesystems that supported metadata and structured files. We are just repeating the mistakes of 40 years ago because most programers have no idea of the history of operating systems.
As soon as you need to move a file between two machines you have to invent a flattened form of the file and all it's metadata into a bytestream. This bytestream can then be sent over a network, or modem, or put onto a tape. Unix removed the complexity of doing this from the OS and made every file take this flattened form. If a user program wanted to impose a more complex structure inside the file that was ok too.
Once everything is a bytestream to the OS other things like pipes between processes become easy and simple. It doesn't matter if the processes are on the same machine or on the other side of a network.
It turned out that the majority of data people stored did not need anything more complex than a flat file with some header information at the front. Unix was a good balance.
When we need a finer level of access to data we turn to databases. They are a whole extra level of complexity.
Who decided that the value of an ad-free Slashdot was equivalent to that of a typical magazine?
There is a general failure in all of the new entertainment ventures. They forget that people have a real or imagined budget for their entertainment:
People don't like metered services. It's too complicated and unpredictable. It doesn't allow the easy comparison of two services. Look at the failure of metered bandwidth ISPs.
While you may be putting a price on the value of advertising, because you are counting page views you are encouraging a mental model of pricing the content. The content comes from us and we'll start wanting something in return for moderating and posting. If people that are active and valuable contributors start to feel that they are being used they will go elsewhere where they feel that they are not being used. There will always be places on the web where people can post their opinions for free.
By ignoring economics, human psychology and history, Slashdot will become another lesson on how-not-to-make-it-pay.
The author has over simplified multiple things in his article. Every simplification has enabled him to claim things that on the surface seem reasonable but in practice are false. To further increase the noise he has a lot of his 'facts' and 'assumptions' incorrect.
1) "J2EE and .Net--will essentially control the programming languages market"
How does he define the programming languages market? There are more people employed writing "vertical applications". Their choice of language and libraries is typically dictated by legacy: fortran, cobol, lisp, C, pascal, etc. What about embedded systems: assembly language, C, C++, Forth. High performance computing using vectors and parallel algorithms have a whole set of specialized languages.
2) "Still, it is an amazing achievement to be able to support different languages on the same runtime, right? I certainly think it is, but others would disagree."
Where has this guy been? By the early 1980s there were compilers for Pascal, C and Fortran that compiled to P-code (a common runtime) and were either compiled to native machine code or ran in an interpreter. Not a new concept. It is just that people are ready to accept the cost today.
3) discussion of what is the natural descendant of C++.
Who really cares? Once you've chosen a language to write in, it doesn't matter how it came to be. It supports a set of programming constructs and has a number of libraries available for use (that may support the type of program you are writing). The evolution of the language is completely irrelevant.
Language designers are very aware of other languages (more so than Liotta) and will borrow/steal ideas and syntax that they like. The language is designed with a particular purpose in mind. There are literally hundreds of domain specific languages that work better than C#, Java, VB for their intended task. They may be similar to another language, or use similar constructs but they are not the same.
4) "With .Net, there is only a single runtime (functionality), but different language syntaxes (look and feel) can make use of it."
There has always been a single runtime, the machine code. It hasn't prevented languages from having different semantics from one another. .Net and JVM won't prevent it either. The syntax of a language is almost accidental from the point of view of the language designer.
I could keep going but I'm also over simplifying so I'll stop.
Amoeba is identical to the design of Plan9 with the following differences: transparent parallelism, built on a micro-kernel.
It's not that revolutionary, nor does it address the configuration and usability for the 'average' user.
Having the source code available enables you to fix the bugs that matter to your business. Even if you can't do it yourself, you can pay me to fix it for you. If you are not willing to pay then the problem is not affecting your bottom line.
Without the source code you have one choice: go to the application developer; hope they will pay attention to your problem, or pay them a huge fee to produce a one-off version for you. Their priorities do not match yours.
Any business would favor OSS if they thought through the issues. What manufacturer/business really wants to be dependant on a single source supplier who receives revenue based on flaws the supplier leaves in the product?
The real issue is that while I can implement a tail call implementation using the existing CLI/JVM instructions it isn't very efficient. Without a tail call instruction in the JVM/CLI the runtime compiler/optimizer is prevented from recognizing the tail call and emitting better machine code.
Programs written in functional language use call instructions much more than traditional ALGOL like programs. The overhead of the call matters much more than you would expect.
Microsoft's implementation of the CLI could (and probably already does) have extra instructions that are not documented publicly. MS could then produce a .NET module that uses them but doesn't work on other implementations of the CLI. .NET world then MS has effectively locked people onto their platform because that module doesn't run on other CLI platforms.
If you want to gain the same advantage then you'll have to licence the info from MS because they'll have the instructions patented. If the module is critical to the
This is exactly what MS was trying with the Java VM before Sun stopped them.
Younger audiences do have 'traditions'. Think back to your early school days. There were games and behaviors that were passed on from one year to the next. These 'traditions' have been propagating through this age group for decades.
Gameboy is firmly established in these younger age groups (esp under 12).
- AndrewN
The article misses the real problem with computer programing: the immaturity of the industry.
I believe that there are three types of programers :-
Programmers: May have a degree in some computer related field, but in practice do not use the 'science' or 'engineering' in their daily work. They are always under some time pressure to get things done.
Engineers: Use formal specifications and mathematics to design and build programs. Most of them are not working in the commercial world. (No UML is not a formal specification.)
Scientists: Look for better algorithms, write papers about their research, typically employed by universities.
So here is the problem:
'programers' write code that is only as good as the best code they have been exposed to. They learn by osmosis. They almost never read research papers, instead they read opinion pieces about the tools they do use.
'engineers' write code that is correct and bug free and take forever to get anything done. As a result they are not wanted in the typical software development team.
'scientists' write code that illustrates their particular area. It's often bug ridden when outside of the test domain and unmaintainable.
Both 'engineers' and 'scientists' write papers for their peers and are not targeted at 'programers' in the field. They may as well be using a different language when it comes to trying to comprehend some of the papers.
There is very little cross over between the different schools of programing. Almost none of the thousands of engineering/science papers published have had any effect on the world of practical programing.
While there are various associations of computing professionals, not one of them provides an industry recognized certification that tests the knowledge of a programmer/engineer. (please no MSCE/MCSE wars)
Given the above it's not suprising that for any given team of programers they will produce code that contains bugs, is insecure, and doesn't use 'best practices'. The tools we use have been written by us, not by engineers, and fulfill a different purpose, not correctness or security but speed of development.
One other point: Why spend time on security of an application when you know the OS and the other applications are gaping security holes anyway. Why put an expensive shiny lock on the back door when the front door has a "please come in" sign? - AndrewN
The biggest problem that still needs to be solved is the implants that pick up signals
cause the surrounding tissue to degrade and become useless after about 6 months.
Spend the time to learn how to use it, then it stops working.
However once they sort this out - sign me up.
The PDA market isn't looking for a computer that needs to be replaced every three years. You should think of it as a consumer device like a VCR or TV, replaced every 5 to 7 years. Hopefully Palm realizes this. They should start to see the repeat business in the next 2 to 3 years.
I've had my Palm Pro since 97, with one hardware upgrade to add extra memory and the IR port. The only reason I bought it was that I knew if I broke or lost it I could restore everything to the latest model. I have a fixed set of applications that cover all my daily note taking, contacts and reminders.
Color would be nice but there is no extra functionality with it. A higher resolution would help reading text documents. However it works the way it is. I'd rather have the battery life.
If I need more complex processing or better graphics, and so on, I'll switch to my laptop or workstation. The PocketPC holds no appeal for me.
Palm made a consumer device that matched what people wanted. Repeat business will happen if Palm keeps their head, brand loyalty still exists.
If you are a hardware manufacturer then the OS software is an expence you have to pay for. Both Sun and Apple treat their software divisions as a non-revenue producing expences.
Separating the OS and the Hardware groups allows them to follow different and competing sales agendas. The OS group can now support the hardware platforms (Sony, IBM, Handspring, Nokia & Palm) that the market is willing to buy. The hardware group can focus on producing the most cost effective platform for a PDA.