It's not about bloatware in terms of resources (well, maybe for the author, but not for myself and many others) -- it's about visual bloat -- flexibility in the WM that doesn't serve any functional purpose. BlackBox and other traditional window managers are just as functionally bloated, even if they don't use more computational resources.
I am a little confused as well. I believe Mallinson's name is on the document at the bottom of the page, which is a solicitation for more submissions. In that context, it is completely okay that it's a MS employee that wrote the document.
It seems more likely that Don Funk wrote the article, though I'm not entirely clear on who Don Funk is -- donfu@microsoft.com? OTOH, searching Google his name shows up lots of places -- is it a name turned into a standard filler (like "foobar")? Or does he write lots of documentation, and uses his own name in example sometimes?
Now I suspect Don Funk wrote the article, and Mallinson is taking credit for it because the entire article is obviously a complete lie if Funk wrote it. The other possibility is that Funk wrote parts of it as documentation, and Mallinson used that content to produce the "letter".
The other thing that links Mallinson to the editing of the content, not the content itself, is the italicized comment on the bottom talking about converting the author to Pocket PC. Someone else referenced this site where this same Mallinson is referred to as an expert with Pocket PCs. Though, when I think of the possibilities, it again becomes likely that she already had all the content for a second article ready, and was setting it up for a followup.
The whole Pocket PC thing makes it seem very likely that Mallinson is simply lying about converting from Mac (at least to XP). But it would offend people more if a man wrote an article, and then was turned into a woman in a cynical attempt to make XP seem more soft and accessible. Probably several people wrote the article, and then one person put it together -- Dun Funk seems like his name is attached to technical things, so I would imagine he would not have put the article together, but would have written some of the more technical material in the article.
Bitkeeper (the gratis version, anyway) only restricts you from using it to develop a competing project, not from using one.
That both those freedoms are essential is a defining aspect of Free Software. That both those freedoms are important is a defining aspect of a true member of the community.
It's funny how much people will bitch when they're not the ones that have to deal with the inadequacies of $OSS_SM_TOOL when it comes to kernel development.
RMS's integrity with respect to this is unquestionable -- he restricts himself solely to using free software, regardless of the flaws in that software. He has also puts a great deal of effort into encouraging and supporting the filling in of holes in available free software. In fact, I think this has been one of the most important things that GNU did -- in the early years, people in the GNU project spent considerable time implementing really boring code that needed to be implemented to create a complete system, laying the groundwork for the completely free systems we can use today.
I would expect that in addition to this post to the kernel list, RMS is also doing what he can to support BK alternatives. But that probably would not be as public or controversial.
I think the reason people aren't getting excited about Parrot is that it hasn't had any significant progress. Skepticism among Python developers has been high (by Python developers I mean the people who work on writing the Python interpreter), and there's been criticism that work has been directed towards optimization before any real language can be targetted for Parrot.
Parrot would be cool, but a free CLR might come around a lot faster.
I'm also unclear also how languages with significantly different semantics will interact.
Ruby and Python, for instance, could interact very easily, since they have similar semantics. But Perl, for instance, considers "12" and 12 to be pretty much equivalent. But when you have some function written in Python, and you pass a string instead of a number, you're likely to get an error (or worse). So now the Perl programmer has to be aware of the languages he's interacting with, which is part of what we were trying to avoid.
It's a hard problem, and I don't know that it's easy to solve.
Re:An indictment of the Python programming languag
on
The Python Cookbook
·
· Score: 2
Oh! A language debate! I love those, though a more thoughtful and informed critique would have made for a better starting point... and this is just a troll (moreso since it's an AC) but I'm feeling bored.
Python provides no apparatus--other than indentation--to delimit the scope of a control flow construct (loop, conditional branch, etc.).
No one who actually spends time writing Python code has any real problem with the indentation. Some like it more than others, but anyone can learn to deal with it. If you don't indent your code properly anyway, then you are the poor programmer.
The issue of tabs and spaces being mixed is well known and debated in the Python community, and mixing is considered very poor form. There are code checkers to avoid this, and Python can be run to reject mixing as a syntax error (personally, I hope this becomes the default in a later release).
I was shocked... there was no means to determine to what class an object belongs based upon its handle.
I'm not entirely clear what you mean by this, "handle" is not clear term. Perhaps you mean, based on the variable name, you cannot tell the class? (Neither by name, nor by looking back in the code for a declaration on that name)
This is true. Perhaps you do not understand the style of programming that Python encourages -- it's roots lie in Lisp and Smalltalk, not C or Fortran. I.e., fully dynamic typing, where an object's type is incidental, but it's properties are essential. If you don't understand that kind of programming, lack of typing will seem like a deficiency.
Almost incredibly, Python does not support declarations of variables: variables merely spring into being when they are first referenced.
Not true. Variables spring into being when you initialize them. Python is not like PHP or Perl in this regard -- there is no default value for an uninitialized variable.
Lack of declarations is not a large source of bugs in actual use, and those bugs that do exist are shallow and easy to fix. Deep bugs are dangerous, but this does not often lead to deep bugs (though language that allow uninitialized variables can get themselves into trouble).
The author of Python has the nerve to trumpet the amazing flexibility of its data structures... I hate to burst his bubble, but he doesn't offer anything that LISP didn't offer forty years ago.
It's built-in data structures are useful, and C (and even Java) are simply lame not to include them directly in the language. Lisp is another matter (though, again, the literal data types available are slim). Python tends to have fairly Lisp-like semantics, and that's okay. Designers of Python have never claimed to be revolutionary -- rather, they have tried to take the best features of languages that have come before, and create a language that puts them together in a pleasing way (unlike, say, Perl that takes every feature, throws them into a big heap, and calls it freedom).
( ) delimits tuples as well as expressions leads to the need to write ( x , )... With a bit more thought, something more professional could have been conceived.
Such as? [ ] and { } are taken (and by data structures that are used more often than tuples). It works, and you get over the (x,) thing really quickly.
Obviously from this critique, you have never seriously used the language. You only are able to critique its outermost veneer of syntax, and you don't have the knowledge to make any comment on the power of its semantics. The wise language designer knows semantics are more important than syntax (though syntax can get in the way, Lisp being the primary example).
Their tools depended on Python 1.5.2. Which would be fine if they had just used #!/usr/bin/python1.5 at the top of their scripts.
The real problem was they used #!/usr/bin/python , and if you wanted to use a more modern Python as the system python (i.e., named simply "python") then you'd break system scripts. It's never been a problem otherwise to have two different versions of Python installed.
The HURD no longer has any political purpose -- it is continued purely for technical reasons, because some people are interested in it (the nature of free software projects). FSF, GNU, and RMS all have absolutely no problem with the Linux kernel on political/ideological grounds. There may be disagreement on binary modules and development tools (i.e. BK), but those things are not the same as the kernel itself.
In fact, if the BSD kernel had clear licensing and legal status when the HURD was first being developed, it probably would have been used instead.
The other side of the coin, though, is that Click-N-Run, unlike the similarly priced Windows, actually provides functional applications. This isn't just about installing applications easily, it's about having a bunch of applications available to use. $130 isn't a bad price.
But all the more reason Lindows won't market to Linux users -- we already know all those applications exist, and we aren't going to pay extra for them. But Windows users don't know that. They won't necessarily consider the situation as free-but-hard, or expensive-but-easy -- and even if they do, isn't that why many haven't switched to Linux earlier?
I read the review, and it wasn't clear to me how this was significantly better than plain Debian. It seems to have a better installer -- and that's nice, but it's just not enough to make another distro worth it. They mentioned an admin tool, but again... that's just one tool (but if it was all-encompassing, that would be cool, but I didn't get that impression).
Sure, there's lots of apps -- but that's because it's Debian. Why is it really better?
By any chance, can somehow give a link to a good reference on how to set up your own CA? mod-ss-makecert makes self-signing really easy, but I have no idea what's involved with making a CA.
Re:"Benefits" of killing the Alpha and PA-RISC...
on
Itanium Problems
·
· Score: 2
My vague impression was that they didn't offer a complete package for the high end, and they didn't offer a good price/compatibility for the low end. To make it a really good server you need all the hardware to work together, like Sun offers (or even HP or IBM). But I never heard any great praises for the rest of the hardware in an Alpha-based computer -- it seemed pretty commodity. You don't always need good hardware when just a fast processor is called for -- but the market for pure processing power doesn't seem that large. Scientists, mostly, and maybe for CGI.
So even when it was most competitive, Alpha wasn't substantially better than a PC. It just had a better chip, which wasn't nearly enough to distinguish it.
Meaningful demonstrations in a non-democratic power structure can only occur if there is a disruption of the normal operation of society.
In a democratic society, demonstrations are important simply because they express the will of the demonstrators. However, it should be clear that there is not even the barest veneer of democracy when it comes to certain issues in this country -- international fiscal policy being one of them (IMF, World Bank, etc).
In such a situation a demonstration can serve three purposes that I can think of:
It brings people of like mind together, so they can make connections and be more productive in their efforts. This is not an end unto itself, however, as no one will be productive if they just spend time meeting with each other.
It can inform the public about the issue at hand. This hasn't been terribly effective. The corporate media deserves a lot of blame for this, but by no means is that the only problem -- demonstrations aren't a good way to educate. And even an educated public still has to do something -- education is not an end unto itself.
The demonstration can disrupt society, i.e., blocking traffic, causing meetings to be held in remote locations, etc. Sometimes you can disrupt only the targetted people/organizations/events -- but usually not, since the security response will make this difficult. So you are forced to disrupt all of society. The powers that be by definition have a great deal at stake in the functioning of society. While this may be an indirect effort, it is the only effort that actually does anything. It is the only function of a demonstration that is an end unto itself.
As others have pointed out, DoS has always been the primary method of peaceful civil disobedience.
And education is not the answer to all political problems. The powers that be are not ignorant of their own actions (as much as some would like to believe that if they knew the results of their actions they would somehow reform themselves). Many people are ignorant, but even as more are educated, it's essential that they do something more with their education than stew. Other suggestions are welcome, though, if you have some great idea. (Personally, I'd like to see more moral attacks on the individuals who are doing the most wrong to the world -- name names, publish pictures, picket their homes, etc)
I'm not sure this is the most effective civil disobedience proposed, but at least it's an attempt.
I think a lot of browsers are emphasizing UI elements more than Technology. I use Galeon more for its UI than its speed -- the tabbing works Just Right, and overall it pleases me.
You can't take all the UI innovation and just put it all in one Super Browser. That will suck -- UI is about what you leave out, not just what you include. No one really knows what The Goal is for all these -- there is no clear vision of what The Most Usable browser will look like. So we have lots of people experimenting with different ideas.
Does this computer even have a PCI slot? I could only get to the press release, but it seems to imply you can only really expand it via PCMCIA or USB. (Are there TV tuners for PCMCIA?)
I know lawyers can't call it legal advice -- but lawyers can inform the public. They mostly choose not to, as an informed public would need less lawyers. At least, that's my (admittedly cynical) take on it.
Posting a question like this in a public forum isn't going to give you a solid answer. I think most people who do so realize this. But it can give you a direction to consider, and it can help inform other people who have similar concerns now or in the future.
I hate to get into meta-discussion, but someone else in this thread has already suggested that they go to the Labour Relations Board, or Labour Standards Branch, which will not charge for advice -- unlike a lawyer. Sounds like better advice then yours.
If you don't have any real advice to give -- and you didn't -- then why not just refrain from posting? Ask Slashdot does not seek to get the opinion of every reader -- only those that have something of use to say. Legal questions are of important to us all, and useful for people to understand before they are confronted with a problem -- with some legal knowlege you can act with more confidence and have some legal predictability in your life.
But instead there's always tons of people like yourself who give blanket non-advice, and attack the forum -- if you don't want to see legal advice on/., then just close your damn eyes and scroll past.
Not that your comment is better or worse than any of the common, equivalent comments... but it got tiring a long time ago.
Except that it seems like most parts of the PATRIOT Act also involve secrecy, and a general denial of any citizen participation. As in this case with libraries -- we have no idea how often this is being used. We cannot have any idea, because it is illegal for the librarians to tell anyone that this is happening.
This isn't getting anything out into the open. Sure, the government was doing bad things before -- all well documented, almost never with reprocusions -- but now that documentation itself is becoming illegal. We can inform on each other, but now it's becoming illegal to inform on the government.
This is a huge step in the wrong direction. The old system wasn't good -- police perjured themselves constantly -- but now that isn't the case, because the police are no longer being asked questions, so they can't lie. How is this possibly better?
I remember Logo as being less useful than Basic - what really sticks in my mind is the brain-dead syntax for string literals, where the first space character ends the string. Hello? Strings with spaces in them?
Traditionally you use a list in place of a string, like PRINT [How are you doing today?] -- all items of the list are implicitly quoted, so that's a list of five words. Many modern Logos have alternate syntaxes for expressing strings with spaces.
There might be ways around that; there might be decent control structures (not just REPEAT 50), ways of creating data structures like arrays, and some means of calling external procedures or operating system routines. But they didn't get much emphasis in Logo programming textbooks, perhaps because they'd differ from one pl atform to another.
Like most Lisps, you can define your own control structures -- all the traditional ones are usually included. For instance, you can implement WHILE like:
TO WHILE:cond:body
IF RUN:cond [RUN:body
WHILE:cond:body]
END
Of course, you will run out of stack if your logo is not tail-recursive (UCBLogo is, many others are not). But with WHILE and some other primitives, you can implement most other control structures without needing tail recursion.
The only data structure is usually the list. This serves as an array, and can do other things as well. Definitely old school Lisp. It's entirely sufficient, though associative arrays are convenient as well (can be implemented with lists, but annoying).
External routines are, as always, entirely up to the implementation. Many have plugins, usually there is no way to call arbitrary external code.
Ah, looking at the old Logo book on my bookshelf I see that the language had list handling similar to Lisp, its parent. But much clunkier and less elegant. Logo may be a functional language but I wouldn't call it a decent one.
It's not meant to be a great language -- it's an expedient language for young children. One example of this is the extensive use of abreviations (e.g., FD for FORWARD). Real programmers don't like these in their languages -- but real programmers type much, much faster than young children:)
I think Logo is very similar to Tcl, except where Tcl uses strings, Logo uses lists. Both have extremely simple syntaxes and semantics, and Logo is one of the purer dialects of Lisp -- but it's sometimes a bit awkward, and it's performance is fairly limited.
I've seen sophisticated computer usering children still get excited about programming, even when their creations pale next to the games they play. Creating things can be tremendously fun.
Some of the commercial Logos (like Microworlds and Terrapin Logo) have considerably more multimedia capabilities. In some ways the language as implemented in those products is dumbed-down, though compared to any "game construction kit" or similar product, they offer tremendous flexibility. They are still real programming environments.
StarLogo is more of a computer science kind of Logo -- all about massive parrallelism. Which can be very fun and motivating as well, but probably requires a different personality to enjoy. The Free logos are still pretty much old-school, triangles on the screen kind of deals.
Of course, get her Lego Logo kits, and no one can resist that.
Logo is more appropriate for teaching mathematics and algorithms than programming -- that has always been it's motivation. As such, I think it is considerably better than Basic or Pascal -- it's higher level, and has a much easier learning curve.
I think Logo should really be seen as a pre-algebra and geometry tool. It's algebra that a child is likely to understand -- somewhat imperative, and easy to experiment with. It's an easy introduction to variables and function composition, possible to do at an age when most children would be totally lost by algebra (moreso when you consider how many teenagers never really get algebra -- maybe most teenagers, as I'm afraid many of my friends' eyes glaze over if they are asked to remember any alegbra as an adult).
It's fine to give things away in order to receive goodwill. And of course pure altruism is also fine. The real problem is when you give someone a gift that is meant to trap -- and it doesn't matter if you are a monopoly or not.
The problem with Windows give-aways is that they close off other avenues of computer knowledge that a person might otherwise find. For instance, Windows uses its own terminology in many places for no good reason (for instance, calling hyperlinks "shortcuts"). I feel this is a specific attempt to miseducate people so they have difficulties with other environments.
Other companies do this too. AOL does it all the time, crippling their users understanding of how the internet works (while pretending it's because they want to make the internet easier). Many companies use binary formats meant to trap data. The MS Visual development tools trap developers, but a lot of other development environments have tried to do the same thing. There was a time when Troll Tech was using Qt in the same way (though they've thankfully reformed). Maybe this is a form of "leveraging", but I don't keep up on the subtleties of business lingo.
When someone gives away a product like this, it is not charity. It is just a loss-leader (and with software, there often isn't any real loss involved). It doesn't matter if you are a monopoly or not -- a loss-leader isn't charity, and the receiver should be suspicious. Microsoft receives extra scrutiny, because they have sinister intentions (as has been evidenced many times over). Sun can be sinister too -- their management of Java in particular, but that could certainly extend into other areas.
But maybe not. I don't think StarOffice is a loss-leader, nor much of Solaris (it is too similar to other Unices). They may be looking solely for goodwill, or they may be competing with Microsoft -- they can compete through giving without committing the same sins.
VNC isn't particularly fast, and it can't compensate for the "1 user per computer" paradigm that many OSes are built with. For example, using X11 clients, several people can all use GUI applications from a single remote computer; VNC would realistically only allow one person to use the desktop at a time because it's basically a KVM-over-ethernet.
VNC can be made faster if you tune it for a particular environment. The server and client find a lowest-common-denominator of communication, so you can increase the efficiency of either side without losing backward compatibility (and if both server and client support high level graphical descriptions and toolkit integration, you should see speedups).
Of course, VNC cannot make Windows multiuser. But VNC does not impose any single-user paradigm on a computer. It is entirely possible to have multiple VNC clients connecting to multiple VNC servers on the same computer, each as a different user.
VNC, as it is, does not integrate terribly well with the native user paradigm -- it won't start up new sessions when a user first logs in, for instance. This is not a big problem, just one that no one has bothered to solve.
In any case, a replacement for X11 is likely to have its client-server capabilities anyway. But without apps, nobody's going to bother to switch no matter how much better such a replacement might be.
Any realistic plan to recreate a graphics system involves porting the major toolkits -- at least GTK and Qt. GTK I know has a significant abstraction layer which hides Xlib from the program entirely. By porting GTK you will have ported most GTK applications -- and people are doing just this (the Win32 porting effort, and the handheld efforts). I believe Qt is similar (especially since it's already cross-platform).
Yes, several older toolkits would be left out. That is not a significant problem.
Slashdot Mirror
It's not about bloatware in terms of resources (well, maybe for the author, but not for myself and many others) -- it's about visual bloat -- flexibility in the WM that doesn't serve any functional purpose. BlackBox and other traditional window managers are just as functionally bloated, even if they don't use more computational resources.
It seems more likely that Don Funk wrote the article, though I'm not entirely clear on who Don Funk is -- donfu@microsoft.com? OTOH, searching Google his name shows up lots of places -- is it a name turned into a standard filler (like "foobar")? Or does he write lots of documentation, and uses his own name in example sometimes?
Now I suspect Don Funk wrote the article, and Mallinson is taking credit for it because the entire article is obviously a complete lie if Funk wrote it. The other possibility is that Funk wrote parts of it as documentation, and Mallinson used that content to produce the "letter".
The other thing that links Mallinson to the editing of the content, not the content itself, is the italicized comment on the bottom talking about converting the author to Pocket PC. Someone else referenced this site where this same Mallinson is referred to as an expert with Pocket PCs. Though, when I think of the possibilities, it again becomes likely that she already had all the content for a second article ready, and was setting it up for a followup.
The whole Pocket PC thing makes it seem very likely that Mallinson is simply lying about converting from Mac (at least to XP). But it would offend people more if a man wrote an article, and then was turned into a woman in a cynical attempt to make XP seem more soft and accessible. Probably several people wrote the article, and then one person put it together -- Dun Funk seems like his name is attached to technical things, so I would imagine he would not have put the article together, but would have written some of the more technical material in the article.
I would expect that in addition to this post to the kernel list, RMS is also doing what he can to support BK alternatives. But that probably would not be as public or controversial.
Parrot would be cool, but a free CLR might come around a lot faster.
I'm also unclear also how languages with significantly different semantics will interact. Ruby and Python, for instance, could interact very easily, since they have similar semantics. But Perl, for instance, considers "12" and 12 to be pretty much equivalent. But when you have some function written in Python, and you pass a string instead of a number, you're likely to get an error (or worse). So now the Perl programmer has to be aware of the languages he's interacting with, which is part of what we were trying to avoid.
It's a hard problem, and I don't know that it's easy to solve.
The issue of tabs and spaces being mixed is well known and debated in the Python community, and mixing is considered very poor form. There are code checkers to avoid this, and Python can be run to reject mixing as a syntax error (personally, I hope this becomes the default in a later release).
I'm not entirely clear what you mean by this, "handle" is not clear term. Perhaps you mean, based on the variable name, you cannot tell the class? (Neither by name, nor by looking back in the code for a declaration on that name)This is true. Perhaps you do not understand the style of programming that Python encourages -- it's roots lie in Lisp and Smalltalk, not C or Fortran. I.e., fully dynamic typing, where an object's type is incidental, but it's properties are essential. If you don't understand that kind of programming, lack of typing will seem like a deficiency.
Not true. Variables spring into being when you initialize them. Python is not like PHP or Perl in this regard -- there is no default value for an uninitialized variable.Lack of declarations is not a large source of bugs in actual use, and those bugs that do exist are shallow and easy to fix. Deep bugs are dangerous, but this does not often lead to deep bugs (though language that allow uninitialized variables can get themselves into trouble).
It's built-in data structures are useful, and C (and even Java) are simply lame not to include them directly in the language. Lisp is another matter (though, again, the literal data types available are slim). Python tends to have fairly Lisp-like semantics, and that's okay. Designers of Python have never claimed to be revolutionary -- rather, they have tried to take the best features of languages that have come before, and create a language that puts them together in a pleasing way (unlike, say, Perl that takes every feature, throws them into a big heap, and calls it freedom). Such as? [ ] and { } are taken (and by data structures that are used more often than tuples). It works, and you get over the (x,) thing really quickly.Obviously from this critique, you have never seriously used the language. You only are able to critique its outermost veneer of syntax, and you don't have the knowledge to make any comment on the power of its semantics. The wise language designer knows semantics are more important than syntax (though syntax can get in the way, Lisp being the primary example).
The real problem was they used #!/usr/bin/python , and if you wanted to use a more modern Python as the system python (i.e., named simply "python") then you'd break system scripts. It's never been a problem otherwise to have two different versions of Python installed.
In fact, if the BSD kernel had clear licensing and legal status when the HURD was first being developed, it probably would have been used instead.
But all the more reason Lindows won't market to Linux users -- we already know all those applications exist, and we aren't going to pay extra for them. But Windows users don't know that. They won't necessarily consider the situation as free-but-hard, or expensive-but-easy -- and even if they do, isn't that why many haven't switched to Linux earlier?
I can seriously imagine using this to install Debian, and then basically turning it back into a Debian machine. Is that how you are using it?
Sure, there's lots of apps -- but that's because it's Debian. Why is it really better?
By any chance, can somehow give a link to a good reference on how to set up your own CA? mod-ss-makecert makes self-signing really easy, but I have no idea what's involved with making a CA.
So even when it was most competitive, Alpha wasn't substantially better than a PC. It just had a better chip, which wasn't nearly enough to distinguish it.
In a democratic society, demonstrations are important simply because they express the will of the demonstrators. However, it should be clear that there is not even the barest veneer of democracy when it comes to certain issues in this country -- international fiscal policy being one of them (IMF, World Bank, etc).
In such a situation a demonstration can serve three purposes that I can think of:
- It brings people of like mind together, so they can make connections and be more productive in their efforts. This is not an end unto itself, however, as no one will be productive if they just spend time meeting with each other.
- It can inform the public about the issue at hand. This hasn't been terribly effective. The corporate media deserves a lot of blame for this, but by no means is that the only problem -- demonstrations aren't a good way to educate. And even an educated public still has to do something -- education is not an end unto itself.
- The demonstration can disrupt society, i.e., blocking traffic, causing meetings to be held in remote locations, etc. Sometimes you can disrupt only the targetted people/organizations/events -- but usually not, since the security response will make this difficult. So you are forced to disrupt all of society. The powers that be by definition have a great deal at stake in the functioning of society. While this may be an indirect effort, it is the only effort that actually does anything. It is the only function of a demonstration that is an end unto itself.
As others have pointed out, DoS has always been the primary method of peaceful civil disobedience.And education is not the answer to all political problems. The powers that be are not ignorant of their own actions (as much as some would like to believe that if they knew the results of their actions they would somehow reform themselves). Many people are ignorant, but even as more are educated, it's essential that they do something more with their education than stew. Other suggestions are welcome, though, if you have some great idea. (Personally, I'd like to see more moral attacks on the individuals who are doing the most wrong to the world -- name names, publish pictures, picket their homes, etc)
I'm not sure this is the most effective civil disobedience proposed, but at least it's an attempt.
You can't take all the UI innovation and just put it all in one Super Browser. That will suck -- UI is about what you leave out, not just what you include. No one really knows what The Goal is for all these -- there is no clear vision of what The Most Usable browser will look like. So we have lots of people experimenting with different ideas.
Does this computer even have a PCI slot? I could only get to the press release, but it seems to imply you can only really expand it via PCMCIA or USB. (Are there TV tuners for PCMCIA?)
Posting a question like this in a public forum isn't going to give you a solid answer. I think most people who do so realize this. But it can give you a direction to consider, and it can help inform other people who have similar concerns now or in the future.
If you don't have any real advice to give -- and you didn't -- then why not just refrain from posting? Ask Slashdot does not seek to get the opinion of every reader -- only those that have something of use to say. Legal questions are of important to us all, and useful for people to understand before they are confronted with a problem -- with some legal knowlege you can act with more confidence and have some legal predictability in your life.
But instead there's always tons of people like yourself who give blanket non-advice, and attack the forum -- if you don't want to see legal advice on /., then just close your damn eyes and scroll past.
Not that your comment is better or worse than any of the common, equivalent comments... but it got tiring a long time ago.
This isn't getting anything out into the open. Sure, the government was doing bad things before -- all well documented, almost never with reprocusions -- but now that documentation itself is becoming illegal. We can inform on each other, but now it's becoming illegal to inform on the government.
This is a huge step in the wrong direction. The old system wasn't good -- police perjured themselves constantly -- but now that isn't the case, because the police are no longer being asked questions, so they can't lie. How is this possibly better?
TO WHILE :cond :body :cond [RUN :body
WHILE :cond :body]
IF RUN
END
Of course, you will run out of stack if your logo is not tail-recursive (UCBLogo is, many others are not). But with WHILE and some other primitives, you can implement most other control structures without needing tail recursion.
The only data structure is usually the list. This serves as an array, and can do other things as well. Definitely old school Lisp. It's entirely sufficient, though associative arrays are convenient as well (can be implemented with lists, but annoying).
External routines are, as always, entirely up to the implementation. Many have plugins, usually there is no way to call arbitrary external code.
It's not meant to be a great language -- it's an expedient language for young children. One example of this is the extensive use of abreviations (e.g., FD for FORWARD). Real programmers don't like these in their languages -- but real programmers type much, much faster than young childrenI think Logo is very similar to Tcl, except where Tcl uses strings, Logo uses lists. Both have extremely simple syntaxes and semantics, and Logo is one of the purer dialects of Lisp -- but it's sometimes a bit awkward, and it's performance is fairly limited.
Some of the commercial Logos (like Microworlds and Terrapin Logo) have considerably more multimedia capabilities. In some ways the language as implemented in those products is dumbed-down, though compared to any "game construction kit" or similar product, they offer tremendous flexibility. They are still real programming environments.
StarLogo is more of a computer science kind of Logo -- all about massive parrallelism. Which can be very fun and motivating as well, but probably requires a different personality to enjoy. The Free logos are still pretty much old-school, triangles on the screen kind of deals.
Of course, get her Lego Logo kits, and no one can resist that.
I think Logo should really be seen as a pre-algebra and geometry tool. It's algebra that a child is likely to understand -- somewhat imperative, and easy to experiment with. It's an easy introduction to variables and function composition, possible to do at an age when most children would be totally lost by algebra (moreso when you consider how many teenagers never really get algebra -- maybe most teenagers, as I'm afraid many of my friends' eyes glaze over if they are asked to remember any alegbra as an adult).
The problem with Windows give-aways is that they close off other avenues of computer knowledge that a person might otherwise find. For instance, Windows uses its own terminology in many places for no good reason (for instance, calling hyperlinks "shortcuts"). I feel this is a specific attempt to miseducate people so they have difficulties with other environments.
Other companies do this too. AOL does it all the time, crippling their users understanding of how the internet works (while pretending it's because they want to make the internet easier). Many companies use binary formats meant to trap data. The MS Visual development tools trap developers, but a lot of other development environments have tried to do the same thing. There was a time when Troll Tech was using Qt in the same way (though they've thankfully reformed). Maybe this is a form of "leveraging", but I don't keep up on the subtleties of business lingo.
When someone gives away a product like this, it is not charity. It is just a loss-leader (and with software, there often isn't any real loss involved). It doesn't matter if you are a monopoly or not -- a loss-leader isn't charity, and the receiver should be suspicious. Microsoft receives extra scrutiny, because they have sinister intentions (as has been evidenced many times over). Sun can be sinister too -- their management of Java in particular, but that could certainly extend into other areas.
But maybe not. I don't think StarOffice is a loss-leader, nor much of Solaris (it is too similar to other Unices). They may be looking solely for goodwill, or they may be competing with Microsoft -- they can compete through giving without committing the same sins.
Of course, VNC cannot make Windows multiuser. But VNC does not impose any single-user paradigm on a computer. It is entirely possible to have multiple VNC clients connecting to multiple VNC servers on the same computer, each as a different user.
VNC, as it is, does not integrate terribly well with the native user paradigm -- it won't start up new sessions when a user first logs in, for instance. This is not a big problem, just one that no one has bothered to solve.
Any realistic plan to recreate a graphics system involves porting the major toolkits -- at least GTK and Qt. GTK I know has a significant abstraction layer which hides Xlib from the program entirely. By porting GTK you will have ported most GTK applications -- and people are doing just this (the Win32 porting effort, and the handheld efforts). I believe Qt is similar (especially since it's already cross-platform).Yes, several older toolkits would be left out. That is not a significant problem.