Actually, the first Palm was pretty good from a user's point of view. But in order to create a good, cheap, low-power handheld, Palm cut many corners when it came to programmability and extensibility.
*NIX is fine; the libraries aren't
on
Agenda VR3 Review
·
· Score: 1
Various forms of UNIX used to run on small PDP-11s with less compute power than your PalmPilot. UNIX is a fine system for those kinds of machines: it's simple, secure, mature, and standardized. PalmOS, in comparison, is non-standard and rather restrictive in terms of its memory model. EPOC is simply non-standard and proprietary, for no good reason.
But if you want to run UNIX on a handheld, you need new libraries, applications, and GUI code, simply because the kind of applications you are going to write are different.
Whether Linux fits the bill, I don't know. While Linux may be lean by modern OS standards, as far as traditional UNIX goes, it's still pretty bloated and complex. But I think as a kernel, Linux probably still stands up pretty well to WinCE.
Re:Going to wait for a post-release review
on
Agenda VR3 Review
·
· Score: 2
The reason these things are slow is because they run X.
That's a lame excuse. I've run a commercial X11 server on a 386/20 with 4Mbyte in the past and it worked fine. XFree86+FLTK may not be adapted to the needs of small devices, but current "small" devices are positively luxurious compared to the kind of machines people used to run X11 on. (I suspect part of the overhead is that FLTK, in the name of cross-platform compatibility, just ignores many of the facilities X11 already has built in.)
Of course, it's "too simplistic". But Xerox was the first company to create a complete, usable, high quality desktop environment. That is what Apple copied.
What's important about this is that Apple's myth of being the savior for "the rest of us" by inventing a simple GUI are false. Apple simply undercut the competition on price by delivering a cheap imitation. In the process, they cut many corners and delivered something that was a neat engineering achievement (the 128k Mac) but whose software architecture has been haunting the industry ever since. It lives on not only in MacOS 9 (thankfully, a line that's at an end now) but also everything up to Windows ME (which took many of its bad design ideas from MacOS).
Justice costs time and money, you have to pick your business partners carefully, and it's much harder to profit from your inventions if you don't already have lots of money. It isn't fair, but nobody has figured out how to make things work any better.
His advice is something you really, really should to take to heart. You may choose not to follow it, and there may be good reasons to, but you run significant risks if you do.
As for Hotline, I think the claims of "innovation" are somewhat overblown. Hotline probably succeeded because it was a well-written piece of software and was at the right place at the right time, not because it did something nobody else had done before. Several other academic groups and companies, however, have been working on similar functionality for years.
But again, I think the reaction that somehow Apple stole theme.org's hardware steps way over the line of common sense.
But it doesn't seem implausible for Apple to get a court injunction and have the police pull the plug. I was assuming that that was what people implied happened.
Some go far as to just blindly hate everything that is big and corporate.
And some people misinterpret the relationship between consumers and corporations as one between natural persons. It isn't. It is perfectly legitimate for Apple to consider having themes.org shut down if they violate their copyright, and it is perfectly legitimate for people to discuss the possibility. "Apple" is not a person whose feelings can be hurt by such discussions.
Apple can't legally "grab" their hardware, but if they could make a case that themes.org keeps violating their intellectual property, they could probably get some local police department to pull the plug. In addition, Apple has a pretty rocky history when it comes to intellectual property claims against others. And Jobs isn't exactly known for being the most restrained executive in the world either.
So, yes, many people are suspicious of Apple, and based on Apple's history, there is some reason for that. I think one can ask with equal justification why so many people are so in love with Apple. The truth is probably somewhere in between: Apple is primarily a money-making venture; innovation, law suits, image, style, quality, open source, and other issues are merely means to that end. Sometimes, they may be genuine, sometimes they are merely faked.
Knowing that you can sell something is not the same as knowing that someone wants it. In fact, it is in Microsoft's business interest to force this technology onto users whether they want it or not, so that Microsoft can then in the future make money on lots of day-to-day transactions.
I think a lot of the bad taste ascribed to Americans is simply due to the lack of alternatives. Many people would like something better, but companies are not filling the need because profit maximization does not completely coincide with satisfying customer demand.
This is the typical Microsoft end-run around standards. There is nothing new about trying to give people the ability to coordinate their calendars, integrate messaging, etc. The W3C is coming out with standards that allow this to happen in a distributed, vendor neutral way.
What Microsoft is doing is convenient: centralize it all on Microsoft servers and Microsoft standards. Forget about federation, server-to-server protocols and all that. What Microsoft is doing is also cheaper in the short run an quicker to market (which is why it will likely beat open standards). Nobody but Microsoft can deliver this, not because they have any better technology, but because they have the market position.
The loser is the consumer, who will be denied any kind of market choice again: your choice may be to buy Microsoft or not schedule any appointments with your doctor, dentist, or insurance broker.
On the bright side: there is a good chance that this will not fly. With always-on Internet connections, people can control their data themselves. Even without any privacy incentives, answering machines still sell well, despite personal voice mail offerings. Many people will probably prefer to keep their personal data in cheap, secure Internet servers in their home, no larger and no more complex than an answering machine.
How do you "accidentally" configure a router or DNS server?
Easy: some older Linux distributions would automatically install set up "routed" and advertise a route to the Internet if you happened to have two network interfaces (beats me why).
Or you might intend to configure something in "linuxconf" and accidentally choose the wrong
default.
Some distributions install a DNS server if you install everything, and they may well turn it on, serving your/etc/hosts file (by no means an unreasonable thing to do at home).
That isn't even taking into account the possibility that anything that you install, or the ActiveX component you run in IE or your Outloook mail attachment, might carry a computer virus that turns your machine into something that tries to break in on other machines.
I'm not defending Schwartz's conduct--I think what he did was stupid and he probably should have been fired and made to pay to clean up his mess. However, many people do need to install new software as part of their job (Intel is very much into Java, Linux, and some open source); if an employer is so prone to misjudgement, I wouldn't feel safe doing such work as part of my job there.
Think about all the data that the government can get:
point-of-sales data
ticket data
toll both data
cell phone location
traffic cameras and license plate information
public and private surveillance cameras
All that information will likely be stored in perpetuity in various databases, searchable and indexed.
Once that information is on-line and on some kind of network, you can track most people's movements in real time, and get video footage about their whereabouts at any time in their lives. With automatic face recognition and automatic tracking technologies, you can even automate the process of finding people in individual shots, so that compiling a complete videography of someone's life during a particular time period will take little more than a single database query.
The potential for abuse by the government is enormous: blackmail of political opponents, selective enforcement by police, plea bargains by prosecutors involving the threat of disclosure of embarrassing video footage, etc. And since "normal" people have no access to that data and can't go on fishing expeditions, they will be severely disadvantaged when it comes to legal action.
The governments et. al have a hard enough job actually getting the film out when needed in a crime that people are jumping up and down about.
Online image and video databases, face recognition, and other technologies make that really easy. Maybe your government doesn't have that now, but they will, soon.
Schwartz's actions were stupid, and the Oregon legal system malfunctioned. But the immediate cause of this problem was Intel. Intel's actions were callous and unnecessary. They could have treated this as a contractual matter, fired Schwartz, and presented him with a bill for a security consultant to come in and clean up after him, a response that would have been severe and heavy-handed enough.
Why should you worry about this if you don't run Crack? Because there are lots of other mistakes and activities that could be misconstrued as illegal computer activity:
You set up your new Linux system and configure it accidentally as a DNS server.
You set up your new Linux system and configure it accidentally as a router.
You accidentally create a user with the same user ID as someone else and NFS mount a remote file system.
You install "everything" from a Linux distribution and get a number of well-known tools for network and protocol debugging that also double as tools for breaking into other computers (nmap, tcpdump, netgrep, etc.).
You accidentally and repeatedly try to connect to someone elses X11, VNC, or remote PC server.
You have a script that automatically downloads and/or distributes data through FTP sites, but you get the password wrong and it keeps looping, logging hundreds or thousands of failed login attempts.
You may, in fact, want to run programs like Saint to check the security of your own system.
You have to be able to rely on your employer to behave reasonably even when you make a mistake. When it comes down to it, a company like Intel will be able to present enough evidence and experts in court to make just about anything look like illegal activity to a non-technical jury.
Intel didn't have enough of a clue to distinguish harmful activity from stupid mistake in this case. That means that if you are going to do anything non-trivial with software (like run Linux, run X11, run VNC, write scripts), given their past performance, there is a good chance that they will again behave in a haphazard and unpredictable way.
Working for Intel seems to expose you to the risk of getting a criminal record for a mistake. I don't think that's the kind of "benefit" I want from an employer. I'd look elsewhere for a job.
In fact, I find this aspect of it much more intriguing than the issue of whether Curl will succeed. I have little doubt that Curl will go down in flames. The question to me is: why is Berners-Lee listed as a founder and what are all those people thinking?
A new gaming platform is a hard sell no matter how you try it. But VC funding may not have been the best way of building it, since for something like that to succeed may easily take a decade of patient, steady development and marketing. The best way for that might have been inside a traditional, large company. Or it might have required a different strategy: develop a Linux gaming platform first, release a bunch of games over the years, make just enough money to break even, and eventually release your own box. That's, incidentally, what Microsoft has been doing.
But VCs want it all: a company worth hundreds of millions of dollars within a few years. It should have been clear from the start that that was very unlikely to work for Indrema. Not impossible, but very unlikely.
A startup is not the only way to build a successful company. Most companies, and most of the biggest companies, have grown gradually over many years. It gives them time to build good products and a good reputation, where startups rely on buzz and marketing blitzes.
Think about that before you take the VC money. If you really care about your product and want it to come to market, a startup may not be the way to go. Of course, if you want a lottery ticket with better chances than the state lottery, the startup is still a good choice.
As for Indrema, let's hope that any efforts they have made on the software side will benefit Linux gaming in the long run.
The problem with earth sized planets out by themselves in the habitable zone is that they may not end up being particularly habitable--people have postulated that an earth-size planet needs a moon in order to be habitable and protected from various ill effects.
However, a very large moon orbiting a gas giant in the habitable zone is an interesting possibility. The gas giant might do the right thing to the moon to stabilize its axis and protect it from space debris (but, then, it might do other bad things).
The Palm is a great PDA. If calendaring, a TODO list, and a few other little apps is what you want, get a Palm. I have one and love it.
However, for custom applications/vertical applications, the Palm is unimpressive: low resolution screen, underpowered processor, limited OS, limited memory management. Something Linux based, with standard APIs, X11, and a good range of scripting languages is a better choice for that.
Devices like the Agenda are good for people who write custom/vertical applications. Development for something with standard POSIX APIs, a 32bit flat address space, and a higher resolution screen is a lot easier than for something like the Palm.
For now, the Palm will still give you better calendaring and better battery life.
The problem with Palm applications is not whether the source is available, but simply that the OS and APIs themselves are so old-fashioned and inflexible. It's more like programming DOS than a modern OS.
You can't compare the Palm and something like the Agenda.
Palm does a limited set of things extremely well. But Palm is not a convenient platform to write custom applications for: its screen, processor, and operating system are way too limited and non-standard. Something like the Agenda uses standard APIs, can deal with bigger programs, and allows people to use standard development tools.
In the long run, devices like the Palm won't survive because their functionality will be subsumed by devices like the Agenda. In fact, I would be surprised if even PalmOS 5 wasn't a completely different system (EPOC? Linux?) that ran the old applications in emulation mode. Until then, you have to make a choice.
(As for WinCE, that's just a lost cause as far as I'm concerned.)
The term "shareware" implies, as I stated, that it is not to be sold or included in a distribution without the author's permission.
That makes no sense. What kind of distribution are you thinking of? Can SIMTEL put it on their FTP site? What about CNET? What if SIMTEL repackages their FTP archive on CD-ROM? What if Walnut Creek sells those CD-ROMs?
The meaning of shareware is that something can be distributed, in any form in which people like. With shareware, the author receives money directly from the end user. If the distributor needs to negotiate with the author, it's not shareware.
The Internet used to be built around P2P services: most hosts on the Internet ran FTP, telnet, SMTP, talk, and other such services. These services allowed me to connect directly to your machine to send files, receive files, and talk to you. Look, ma, no central server in the loop. Amazing--all that in 1980, if not earlier.
The profusion of hosts with impoverished services is a more recent phenomenon, brought about by intermittent dial-up connections, dumbed-down nearly-single-tasking operating systems, and greedy ISPs.
Maybe now that even mass market operating systems are catching up, that people are getting permanent connections to the Internet, and that bandwidth is increasing, P2P will come back again.
Other than let you fiddle around with bits and copy rectangular regions back and forth, does SDL even give you 2D graphics? You know, like lines, rectangles, circles, gradients, colors, alpha transparency? I didn't see anything in the APIs last I looked.
Slashdot Mirror
Actually, the first Palm was pretty good from a user's point of view. But in order to create a good, cheap, low-power handheld, Palm cut many corners when it came to programmability and extensibility.
But if you want to run UNIX on a handheld, you need new libraries, applications, and GUI code, simply because the kind of applications you are going to write are different.
Whether Linux fits the bill, I don't know. While Linux may be lean by modern OS standards, as far as traditional UNIX goes, it's still pretty bloated and complex. But I think as a kernel, Linux probably still stands up pretty well to WinCE.
That's a lame excuse. I've run a commercial X11 server on a 386/20 with 4Mbyte in the past and it worked fine. XFree86+FLTK may not be adapted to the needs of small devices, but current "small" devices are positively luxurious compared to the kind of machines people used to run X11 on. (I suspect part of the overhead is that FLTK, in the name of cross-platform compatibility, just ignores many of the facilities X11 already has built in.)
What's important about this is that Apple's myth of being the savior for "the rest of us" by inventing a simple GUI are false. Apple simply undercut the competition on price by delivering a cheap imitation. In the process, they cut many corners and delivered something that was a neat engineering achievement (the 128k Mac) but whose software architecture has been haunting the industry ever since. It lives on not only in MacOS 9 (thankfully, a line that's at an end now) but also everything up to Windows ME (which took many of its bad design ideas from MacOS).
His advice is something you really, really should to take to heart. You may choose not to follow it, and there may be good reasons to, but you run significant risks if you do.
As for Hotline, I think the claims of "innovation" are somewhat overblown. Hotline probably succeeded because it was a well-written piece of software and was at the right place at the right time, not because it did something nobody else had done before. Several other academic groups and companies, however, have been working on similar functionality for years.
But it doesn't seem implausible for Apple to get a court injunction and have the police pull the plug. I was assuming that that was what people implied happened.
And some people misinterpret the relationship between consumers and corporations as one between natural persons. It isn't. It is perfectly legitimate for Apple to consider having themes.org shut down if they violate their copyright, and it is perfectly legitimate for people to discuss the possibility. "Apple" is not a person whose feelings can be hurt by such discussions.
So, yes, many people are suspicious of Apple, and based on Apple's history, there is some reason for that. I think one can ask with equal justification why so many people are so in love with Apple. The truth is probably somewhere in between: Apple is primarily a money-making venture; innovation, law suits, image, style, quality, open source, and other issues are merely means to that end. Sometimes, they may be genuine, sometimes they are merely faked.
I think a lot of the bad taste ascribed to Americans is simply due to the lack of alternatives. Many people would like something better, but companies are not filling the need because profit maximization does not completely coincide with satisfying customer demand.
What Microsoft is doing is convenient: centralize it all on Microsoft servers and Microsoft standards. Forget about federation, server-to-server protocols and all that. What Microsoft is doing is also cheaper in the short run an quicker to market (which is why it will likely beat open standards). Nobody but Microsoft can deliver this, not because they have any better technology, but because they have the market position.
The loser is the consumer, who will be denied any kind of market choice again: your choice may be to buy Microsoft or not schedule any appointments with your doctor, dentist, or insurance broker.
On the bright side: there is a good chance that this will not fly. With always-on Internet connections, people can control their data themselves. Even without any privacy incentives, answering machines still sell well, despite personal voice mail offerings. Many people will probably prefer to keep their personal data in cheap, secure Internet servers in their home, no larger and no more complex than an answering machine.
I find gun violence, shootings, and murder a lot more "indecent" than even graphic depictions of sex. Maybe we need to change our community standards.
Easy: some older Linux distributions would automatically install set up "routed" and advertise a route to the Internet if you happened to have two network interfaces (beats me why). Or you might intend to configure something in "linuxconf" and accidentally choose the wrong default. Some distributions install a DNS server if you install everything, and they may well turn it on, serving your /etc/hosts file (by no means an unreasonable thing to do at home).
That isn't even taking into account the possibility that anything that you install, or the ActiveX component you run in IE or your Outloook mail attachment, might carry a computer virus that turns your machine into something that tries to break in on other machines.
I'm not defending Schwartz's conduct--I think what he did was stupid and he probably should have been fired and made to pay to clean up his mess. However, many people do need to install new software as part of their job (Intel is very much into Java, Linux, and some open source); if an employer is so prone to misjudgement, I wouldn't feel safe doing such work as part of my job there.
- point-of-sales data
- ticket data
- toll both data
- cell phone location
- traffic cameras and license plate information
- public and private surveillance cameras
All that information will likely be stored in perpetuity in various databases, searchable and indexed.Once that information is on-line and on some kind of network, you can track most people's movements in real time, and get video footage about their whereabouts at any time in their lives. With automatic face recognition and automatic tracking technologies, you can even automate the process of finding people in individual shots, so that compiling a complete videography of someone's life during a particular time period will take little more than a single database query.
The potential for abuse by the government is enormous: blackmail of political opponents, selective enforcement by police, plea bargains by prosecutors involving the threat of disclosure of embarrassing video footage, etc. And since "normal" people have no access to that data and can't go on fishing expeditions, they will be severely disadvantaged when it comes to legal action.
Online image and video databases, face recognition, and other technologies make that really easy. Maybe your government doesn't have that now, but they will, soon.
Why should you worry about this if you don't run Crack? Because there are lots of other mistakes and activities that could be misconstrued as illegal computer activity:
You have to be able to rely on your employer to behave reasonably even when you make a mistake. When it comes down to it, a company like Intel will be able to present enough evidence and experts in court to make just about anything look like illegal activity to a non-technical jury.
Intel didn't have enough of a clue to distinguish harmful activity from stupid mistake in this case. That means that if you are going to do anything non-trivial with software (like run Linux, run X11, run VNC, write scripts), given their past performance, there is a good chance that they will again behave in a haphazard and unpredictable way.
Working for Intel seems to expose you to the risk of getting a criminal record for a mistake. I don't think that's the kind of "benefit" I want from an employer. I'd look elsewhere for a job.
In fact, I find this aspect of it much more intriguing than the issue of whether Curl will succeed. I have little doubt that Curl will go down in flames. The question to me is: why is Berners-Lee listed as a founder and what are all those people thinking?
But VCs want it all: a company worth hundreds of millions of dollars within a few years. It should have been clear from the start that that was very unlikely to work for Indrema. Not impossible, but very unlikely.
A startup is not the only way to build a successful company. Most companies, and most of the biggest companies, have grown gradually over many years. It gives them time to build good products and a good reputation, where startups rely on buzz and marketing blitzes.
Think about that before you take the VC money. If you really care about your product and want it to come to market, a startup may not be the way to go. Of course, if you want a lottery ticket with better chances than the state lottery, the startup is still a good choice.
As for Indrema, let's hope that any efforts they have made on the software side will benefit Linux gaming in the long run.
However, a very large moon orbiting a gas giant in the habitable zone is an interesting possibility. The gas giant might do the right thing to the moon to stabilize its axis and protect it from space debris (but, then, it might do other bad things).
However, for custom applications/vertical applications, the Palm is unimpressive: low resolution screen, underpowered processor, limited OS, limited memory management. Something Linux based, with standard APIs, X11, and a good range of scripting languages is a better choice for that.
For now, the Palm will still give you better calendaring and better battery life.
The problem with Palm applications is not whether the source is available, but simply that the OS and APIs themselves are so old-fashioned and inflexible. It's more like programming DOS than a modern OS.
Palm does a limited set of things extremely well. But Palm is not a convenient platform to write custom applications for: its screen, processor, and operating system are way too limited and non-standard. Something like the Agenda uses standard APIs, can deal with bigger programs, and allows people to use standard development tools.
In the long run, devices like the Palm won't survive because their functionality will be subsumed by devices like the Agenda. In fact, I would be surprised if even PalmOS 5 wasn't a completely different system (EPOC? Linux?) that ran the old applications in emulation mode. Until then, you have to make a choice.
(As for WinCE, that's just a lost cause as far as I'm concerned.)
That makes no sense. What kind of distribution are you thinking of? Can SIMTEL put it on their FTP site? What about CNET? What if SIMTEL repackages their FTP archive on CD-ROM? What if Walnut Creek sells those CD-ROMs?
The meaning of shareware is that something can be distributed, in any form in which people like. With shareware, the author receives money directly from the end user. If the distributor needs to negotiate with the author, it's not shareware.
The profusion of hosts with impoverished services is a more recent phenomenon, brought about by intermittent dial-up connections, dumbed-down nearly-single-tasking operating systems, and greedy ISPs.
Maybe now that even mass market operating systems are catching up, that people are getting permanent connections to the Internet, and that bandwidth is increasing, P2P will come back again.
Other than let you fiddle around with bits and copy rectangular regions back and forth, does SDL even give you 2D graphics? You know, like lines, rectangles, circles, gradients, colors, alpha transparency? I didn't see anything in the APIs last I looked.