Well, this is akin to the "give a man a fish or teach a man to fish" argument. I think that there is a scarcity of Microsoft expertise in Mauritius too, causing it to be too expensive anyway, let alone payed-for Linux expertise.
But here's the difference: giving third-world countries Microsoft is "giving them a fish" because it is closed. Giving third-world countries Free alternatives is "teaching them to fish", because it's Open. Just as it initially costs more to teach fishing than to give away fish, so it costs more at the start to set up Linux et. al. than to install Microsoft stuff. But the end result is no dependence by the Mauritians on Microsoft.
I think you've hit upon it there, litterally. I believe the word "cyber" was coined by AI researchers for the ability of computers to interact with humans, either via a human interface, or by acting human. Later, it turned more towards embedding computers into humans as a form of prosthetic (a la "$6 millon man") or to build composite computer-humans (cyborg).
Lately I think the US "cyber security" push has one of these aims:
To control cyborg's access to the net?
To curb "cyberterrorism" -- the attack on America by those cyborgs?
To promote safe use of teledildonics within federal agencies, a sort of "monica lewinsky" protocol aimed at avoiding future political embarrasments.
Obviously, the most likely of these is the third aim...;-)
Actually I would say that Homeland Security is all about enforcing the US Government's control over it's own people, and a prime example of the Freedom that most US Citizens NO LONGER HAVE. Witness:
* The DMCA * The PATRIOT act * The increasing biocontrols at air and sea ports * Mandatory fingerprints for all US citizens entering or leaving the country * The scary ability that US police shows portray of any US citizen being seconds away from a database search, and the apparent acceptance by Hollywood that this is normal and good * Unjustified arrests of Americans at protests * Unexplained (and probably unjust) deportations of Americans from other countries, for apparent civil disobedience.
Homeland Security has done nothing about the safety of US Citizens because it is not really about that (that's just the excuse). It is in response to terrorism launched by naturalised americans against America.
I am not an American. I am living in a country that also enjoys the same Freedom by Constitutional right that Americans worship, only for Australia it was done without a war and without ammendments. I feel sorry for Americans as I watch their freedom being erroded by a runaway dictator president who was not even elected by the People of America. I feel shocked that so many Americans feel that they are still "the land of the free". And I watch in horror as my own country follows that same path.
I've made some suggestions elsewhere on this story, but actually, the more I think about this, the more I feel you are right.
It sounds a bit like a "checklist feature" that the faculty head is requesting for his degree offerings. I don't know how it could add to the relevance or effectiveness of the overall degree, especially within the confines of University guidelines.
More value could be added to the degree by having some practical unit where students contribute to some OSS projects, rather than by having a course devoted to OSS theory. The trickiest part about that though would be marrying the concepts of "open source" to the university precept of "plagerise not". While I would have loved to have worked on an OSS project for my course-work at Uni, I can see it would have been very hard to mark any work I submitted to the project as having been done by me!
Maybe it would be best to get some sort of credit for working on things like the Google coding challenges, either during a degree or as part of entry selection criteria.
Lecture 3 is an interesting one. Many would argue that there are "too many" OSS licenses already. While I don't agree with that, I do feel there is a benefit in using a license that is "OSS Certified" over rolling your own.
Perhaps this one could focus more on sellecting some of the "less-free" licenses, since many are written for "common" reasons. However, we risk be getting into the realms of a law degree, which is too much focus for a CS student. Very tricky balance there. Actually, perhaps a law professor could help to craft these two lessons better?
Better to merge it with 3 and focus on using refactoring tools and pretty-printers to reformat your code to the project's stated standards, especially since every project is different, but most do seem to be consistent within the project (it seems to be a criteria for path acceptance).
A general appreciation for coding standards, why they are different even for the same language, and how to write good code that can be made to conform to a given OSS project, would be better than focussing on, say the Sun Java programming standards, or the K&R verses C++ style relligious war.
Suits by contrast would use words like "empower", "dashboard", "collaborate", "monetize" "consumer market" and other marketroid speek. Which says a lot about the site's blogger, I think. He's no geek.
If he were a geek looking for a "swiss army knife" web browser, he wouldn't go past Mozilla propper, not Firefox:-)
That's a great idea. Also useful for lessenning the amount of "friendly fire" sustained by ground forces.
Btw, how successful has aircraft FOF detection been historically on the battlefield, and how easy would it be to defeat this as an implementation of FOF for troups? I suppose it would depend on the strength of the encryption. Also I suppose this system would fail if radio jamming or silence were being employed to confuse enemies. But once the airwaves were in control, I suppose it would be good for "mopping up".
not sure, but after reading the article, it definately falls under the heading of "bad reporting", so it's maybe not so off-topic after all!
Seriously, I'm no MS fan but I couldn't find how this article was living up to it's headline. More like, that MS had to start over because they finally recognised the problem with a monolithic approach.
Reading between the (rather repetative) lines of this article, It sort of sounds to me like MS have finally discovered some of the benefits of modular programming, in a 1970s Unix style of "everything should do one thing well". Gosh, it only took them 30 years. And still they have fears that their so-called "inovation" may be crippled by the new approach. It seems to highlight a serious mis-understanding that MS engineers have about how to produce code, that we could see from the end results, but could only speculate over before.
It may surprise you to know just how many businesses are running on in-house software that is controlled and maintained by just one person. And how many of these Subject Matter Experts are very close to retirement age. (Hint: much of it has to do with COBOL and a deep understanding of the applications' business domains that comes from decades of experience).
In order to be a good mentor and raise an apprentice or three, you need the time and motivation to do it. Many businesses I have worked for are in favour of an apprentiship programme in principle, until someone has to either pay for it directly, or sacrifice time-to-market. When you start asking them to think of it as insurance, the smarter businesses come around more, but still the time pressures are very high.
The motivation for the expert varies: sometimes, if one's job is not satisfactory, one could be motivated to train an apprentice so that there is an exit plan within the company. Most simply look for work outside their current company. On the other hand, in the current IT job market, nobody wants to look replaceable.
In the case of kernel maintenance, it sounds to me like Andrew has definate goals that he wants to achieve. So he is motivated to stay. What he needs is a way to take some of the drudge-work away so he can concentrate on those goals. He has identified one way (submitters do the testing that he has to do). Perhaps we could suggest others, rather than worry that "the process" is broken, or argue over the benefits of testing?
So, I read the first 3 paras and gave up. It's all about the reviewer... this is why I hate reviews, their all so full of themselves. Think I'll just move along now.
I see where you're comming from: definately as a global society there should be more focus on preventing things that are clearly wrong, rather than concentrating on RIAA/MPAA's supposed minor losses.
However, from a pragmatic view, at this stage there isn't a way to police it that isn't overly ownerous to Limewire (or others), or heavily manual (which implies invasion of privacy to non-offenders).
A keyword or points system is too easy (even for people with cro-magnon brains like child porn pushers) to defeat. Remember also that there are (at least, I'm not an expert!) two types of child porn offenses:
1. People misleading children to view offensive material 2. People posting offensive material containing children
If I were trying to get kids to view porn, I would probably append keywords interesting to children, having nothing to do with the actual image content. If I were trying to share child pornography, and stupid enough to try it online, I would probably not say "get your child porn here, payments to account X", either, or anything even remotely like that, for fear of being caught. So by implementing this loose protection, I've not detered anybody, I've only caught the extremely stupid offenders (admitedly, this might be a considerable portion of that goup...) and I've inconvenienced Limewire and potentially their customers for no gain.
One day, we may have the technology to enforce such forms of online censorship, but until then, it is up to parents to protect their kids from child porn type 1 by treating the Net as PG rated, and it's up to police and everybody to protect kids from becomming the subject of child porn photography type 2 and other abuse, directly. This is not an application for which computers are an even mildly effective or appropriate tool, and I would be very concerned if one day, parents and carers did rely on computers for this sort of child minding.
Well, I guess they sort-of have, since now if you want to share child pron on Limewire, you have to have a license...:-)
Seriously though, how is one to police the content of a JPEG or MPEG? It's not going to have some meta-data attached saying "look out, this is a kiddie porn picture!" More likely it'll have mis-leading meta-data, if any. And, while artificial vision is getting a lot better, I don't see how you could build a reliable "child porn" filter to scan all JPEG uploads, let alone any movies. What will the computer be looking for? Or do you propose outsourcing overseas to some manual eye-ball scans of all picture/movie/sound uploads?
You then get into all sorts of "civil liberty" problems, not to mention data quality (too many false-positives and false-negatives), and leaving Limewire users wide-open to attack from civil libitarians making pay-offs to the outsourced scanners, especially targetting any convicted pedofiles.
I still say, in the area of "censoring" the net against child porn, it would be far more effective to leave the porn online, mark it, and then track the people who download it. Even then, most of the black-market kid porn is not done online anyway: it's done my snail-mailing CD-Rs full of the crap. Also, this approach of capturing the kid-porn users online does nothing to help the children who are the subject of the material.
Um, doesn't that same 97% of the demographic just download a pre-compiled binary? They won't have to eat the tutti-frutti.
However, I agree that getting the Moz developers to compile as suggested is probably going to be only a little less difficult.
I also think that compiler checks aren't a silver bullet, though it would probably help guard against heap overruns (by crashing the prog if an exploit is attempted, I presume). That's Java-thinking, we won't go there...;-)
I don't know about WebDAV, but googling for "emacs webdav mode" brings up a few hits. It seems you aren't alone there and people's mileage varies with the current offerings. Since I know nothing about WebDAV beyond what it is, I can't recommend, sorry.
But I hate the fonts too and "invested" some time to figure it out.
Newer X versons of Emacs (v 21 and up, maybe earlier) have the notion of "faces" to describe the fonts used by the editor. You can click-and-drool the config for your "default face" like so (these instructions are for GNU Emacs/X11, though I think it's the same for Xemacs):
0) Open the Options menu, choose Customize Emacs > Browse Customization Groups. A new buffer opens, containing the *Customize Browser* buffer, in Custom mode. This mode lets you click on Emacs buttons to select things.
1) Expand the Faces group (by clicking on the [+] button next to it, and then expand Basic Faces
2) Click on the Face button next to "Default". Other faces inherit basic options from this face, so for a global change like font size, this is the place to go.
3) you can change the values for the font Height, Width, Colo[u]r etc. I find the "misc-fixed" font family with a Height of 173 1/10pt (that is, 17.3 pt) to be readable and still small enough to fit two 80-char wide Emacs windows side-by-side on my 1600x1200 screen (handy for ediff).
4) When you are happy with your selections, press the State button and choose "Save for Future Sessions" from the pop-up menu. Emacs will write the necessary stuff to your ~/.emacs file which is loaded each time Emacs starts.
5) Once it's in your.emacs file, you can transfer to other machines (either the whole file, or just the font bit written by the Custom buffer). Look for the call to custom-set-faces. Mine follow (which, together with the default colours results in an LCARS-like appearance):
(custom-set-faces ;; custom-set-faces was added by Custom -- don't edit or cut/paste it! ;; Your init file should contain only one such instance. '(default ((t (:stipple nil:background "Black":foreground "Orange":inverse-video nil:box nil:strike-through nil:overline nil:underline nil:slant normal:weight normal:height 180:width normal:family "misc-fixed")))) '(cursor ((t (:background "yellow")))) '(highlight ((((class color) (background dark)) (:background "blue")))) '(mouse ((t (:background "green")))))
Spell checking isn't useful either, actually, because English is inconsistent across national borders. I quit using Word's spell check after version 6, when the stupid thing doesn't remember your language settings and insists on "US-English" with "zee"s everywhere and all the colours spelt wrong. And spell checkers actually are a hindrance in some places: take using the wrong word (principle v. principal). A grammar check won't pick up that "It's the principal that counts" is incorrect, either.
On the other hand, the presence of a very bad spell check has prompted me to learn some spelling now. I find that reading a lot helps, since I get used to how the words should look. I still make mistakes though.
It's sort of like quantised music synthesisers. Very cool in the 1980's, but it lead to very "square" unnatural, boring sounding tunes (take "Axel-F" for instance, which actually sounds better now that the "crazy frog" is ever so slightly out of time).
I think we should just drop English as the Internet language and adopt Loglan (www.loglan.org). That's what it was invented for.
Well I share your point. I think if you read the GNOME HCI guidlines, you'll see where they are aiming, though. And yes, presently there is a ways to go.
I'm more into KDE myself, and it's very much like Windows, though that's not why I like it. I like konquerer better than nautilus, mainly. But I like Aqua better than everything else I've seen, and GNOME seems to be headed that way, from my viewpoint, so I keep my eye on it. And the initiative of freedesktop.org means that the two work well together, which I think is fantastic.
Still, each to their own. Where would you like GUIs to go? It's a very interesting research question. So far there is nothing perfect, perhaps their won't ever be. Maybe we'll eventually end up with something like Plan9 or LCARS (which both feature hightly configurable, pannel-like interfaces, though LCARS is fiction and heavily supplanted by an oral interface), or maybe we'll develop something else entirely. But so long as we are stuck with a 2D pallette/mindset, we won't get a lot further anyway.
However for a "mainstream" GUI, where should GNOME/KDE/Commercials be aiming? For a future that is not supported by current tech, or for something that's the best we can do within current limitations?
yup, which is how I read it, and asumed a different poster than SpiderMonkey.
I think it's a reflection on the sad events in recent history: we've all been burned by Microsoft, and with the SCO lunacy and patents on just about anything, it's probably how we are being conditioned.
I'm going to work out how to file a patent on "Method to file obvious techniques with USPTO in such a way that the USPTO will not recognise it as obvious or frivolous". Then the next patent Microsoft submits I can sue them over....
My take is that it's a pledge, not a contract. It can't be a contract unless there are some joint parties to sign an agreement. Consideration is not necessary but a written contract between parties is required for it to be legally binding.
My take on this pledge is that CA won't sue you, unless you sue first, and then CA will use these pledged patents as amunition, if you use any of them. It's sort of like the patent clause on GPL3.
If you want it to be legally binding, then ideally the Patent Commons should be an entity, like the OSDL or FSF (though probably not actually one of these) which can sign a contract on behalf of a group of OSS developers. I would like to see a development in this direction. If companies just gave their patents away to Public Domain, then they lose the ability to "fight back" if somebody else sues first.
Slashdot Mirror
The software licenses you in America too, matey... or have you not seen the Microserf EULA or used a Winblows XP machine for more than 30 days?
Well, this is akin to the "give a man a fish or teach a man to fish" argument. I think that there is a scarcity of Microsoft expertise in Mauritius too, causing it to be too expensive anyway, let alone payed-for Linux expertise.
But here's the difference: giving third-world countries Microsoft is "giving them a fish" because it is closed. Giving third-world countries Free alternatives is "teaching them to fish", because it's Open. Just as it initially costs more to teach fishing than to give away fish, so it costs more at the start to set up Linux et. al. than to install Microsoft stuff. But the end result is no dependence by the Mauritians on Microsoft.
I would say the repeated application by Zonk of dupes constitutes a standard of sorts... just a poor one.
Lately I think the US "cyber security" push has one of these aims:
Obviously, the most likely of these is the third aim... ;-)
Actually I would say that Homeland Security is all about enforcing the US Government's control over it's own people, and a prime example of the Freedom that most US Citizens NO LONGER HAVE. Witness:
* The DMCA
* The PATRIOT act
* The increasing biocontrols at air and sea ports
* Mandatory fingerprints for all US citizens entering or leaving the country
* The scary ability that US police shows portray of any US citizen being seconds away from a database search, and the apparent acceptance by Hollywood that this is normal and good
* Unjustified arrests of Americans at protests
* Unexplained (and probably unjust) deportations of Americans from other countries, for apparent civil disobedience.
Homeland Security has done nothing about the safety of US Citizens because it is not really about that (that's just the excuse). It is in response to terrorism launched by naturalised americans against America.
I am not an American. I am living in a country that also enjoys the same Freedom by Constitutional right that Americans worship, only for Australia it was done without a war and without ammendments. I feel sorry for Americans as I watch their freedom being erroded by a runaway dictator president who was not even elected by the People of America. I feel shocked that so many Americans feel that they are still "the land of the free". And I watch in horror as my own country follows that same path.
I've made some suggestions elsewhere on this story, but actually, the more I think about this, the more I feel you are right.
It sounds a bit like a "checklist feature" that the faculty head is requesting for his degree offerings. I don't know how it could add to the relevance or effectiveness of the overall degree, especially within the confines of University guidelines.
More value could be added to the degree by having some practical unit where students contribute to some OSS projects, rather than by having a course devoted to OSS theory. The trickiest part about that though would be marrying the concepts of "open source" to the university precept of "plagerise not". While I would have loved to have worked on an OSS project for my course-work at Uni, I can see it would have been very hard to mark any work I submitted to the project as having been done by me!
Maybe it would be best to get some sort of credit for working on things like the Google coding challenges, either during a degree or as part of entry selection criteria.
Lecture 3 is an interesting one. Many would argue that there are "too many" OSS licenses already. While I don't agree with that, I do feel there is a benefit in using a license that is "OSS Certified" over rolling your own.
Perhaps this one could focus more on sellecting some of the "less-free" licenses, since many are written for "common" reasons. However, we risk be getting into the realms of a law degree, which is too much focus for a CS student. Very tricky balance there. Actually, perhaps a law professor could help to craft these two lessons better?
I agree.
But point 2 is religious territory!
Better to merge it with 3 and focus on using refactoring tools and pretty-printers to reformat your code to the project's stated standards, especially since every project is different, but most do seem to be consistent within the project (it seems to be a criteria for path acceptance).
A general appreciation for coding standards, why they are different even for the same language, and how to write good code that can be made to conform to a given OSS project, would be better than focussing on, say the Sun Java programming standards, or the K&R verses C++ style relligious war.
Suits by contrast would use words like "empower", "dashboard", "collaborate", "monetize" "consumer market" and other marketroid speek. Which says a lot about the site's blogger, I think. He's no geek.
If he were a geek looking for a "swiss army knife" web browser, he wouldn't go past Mozilla propper, not Firefox :-)
That's a great idea. Also useful for lessenning the amount of "friendly fire" sustained by ground forces.
Btw, how successful has aircraft FOF detection been historically on the battlefield, and how easy would it be to defeat this as an implementation of FOF for troups? I suppose it would depend on the strength of the encryption. Also I suppose this system would fail if radio jamming or silence were being employed to confuse enemies. But once the airwaves were in control, I suppose it would be good for "mopping up".
not sure, but after reading the article, it definately falls under the heading of "bad reporting", so it's maybe not so off-topic after all!
Seriously, I'm no MS fan but I couldn't find how this article was living up to it's headline. More like, that MS had to start over because they finally recognised the problem with a monolithic approach.
Reading between the (rather repetative) lines of this article, It sort of sounds to me like MS have finally discovered some of the benefits of modular programming, in a 1970s Unix style of "everything should do one thing well". Gosh, it only took them 30 years. And still they have fears that their so-called "inovation" may be crippled by the new approach. It seems to highlight a serious mis-understanding that MS engineers have about how to produce code, that we could see from the end results, but could only speculate over before.
Which might be interpreted as an hostile threat! Thank goodness it was launched in the seventies!
It may surprise you to know just how many businesses are running on in-house software that is controlled and maintained by just one person. And how many of these Subject Matter Experts are very close to retirement age. (Hint: much of it has to do with COBOL and a deep understanding of the applications' business domains that comes from decades of experience).
In order to be a good mentor and raise an apprentice or three, you need the time and motivation to do it. Many businesses I have worked for are in favour of an apprentiship programme in principle, until someone has to either pay for it directly, or sacrifice time-to-market. When you start asking them to think of it as insurance, the smarter businesses come around more, but still the time pressures are very high.
The motivation for the expert varies: sometimes, if one's job is not satisfactory, one could be motivated to train an apprentice so that there is an exit plan within the company. Most simply look for work outside their current company. On the other hand, in the current IT job market, nobody wants to look replaceable.
In the case of kernel maintenance, it sounds to me like Andrew has definate goals that he wants to achieve. So he is motivated to stay. What he needs is a way to take some of the drudge-work away so he can concentrate on those goals. He has identified one way (submitters do the testing that he has to do). Perhaps we could suggest others, rather than worry that "the process" is broken, or argue over the benefits of testing?
So, I read the first 3 paras and gave up. It's all about the reviewer... this is why I hate reviews, their all so full of themselves. Think I'll just move along now.
I see where you're comming from: definately as a global society there should be more focus on preventing things that are clearly wrong, rather than concentrating on RIAA/MPAA's supposed minor losses.
However, from a pragmatic view, at this stage there isn't a way to police it that isn't overly ownerous to Limewire (or others), or heavily manual (which implies invasion of privacy to non-offenders).
A keyword or points system is too easy (even for people with cro-magnon brains like child porn pushers) to defeat. Remember also that there are (at least, I'm not an expert!) two types of child porn offenses:
1. People misleading children to view offensive material
2. People posting offensive material containing children
If I were trying to get kids to view porn, I would probably append keywords interesting to children, having nothing to do with the actual image content. If I were trying to share child pornography, and stupid enough to try it online, I would probably not say "get your child porn here, payments to account X", either, or anything even remotely like that, for fear of being caught. So by implementing this loose protection, I've not detered anybody, I've only caught the extremely stupid offenders (admitedly, this might be a considerable portion of that goup...) and I've inconvenienced Limewire and potentially their customers for no gain.
One day, we may have the technology to enforce such forms of online censorship, but until then, it is up to parents to protect their kids from child porn type 1 by treating the Net as PG rated, and it's up to police and everybody to protect kids from becomming the subject of child porn photography type 2 and other abuse, directly. This is not an application for which computers are an even mildly effective or appropriate tool, and I would be very concerned if one day, parents and carers did rely on computers for this sort of child minding.
Well, I guess they sort-of have, since now if you want to share child pron on Limewire, you have to have a license... :-)
Seriously though, how is one to police the content of a JPEG or MPEG? It's not going to have some meta-data attached saying "look out, this is a kiddie porn picture!" More likely it'll have mis-leading meta-data, if any. And, while artificial vision is getting a lot better, I don't see how you could build a reliable "child porn" filter to scan all JPEG uploads, let alone any movies. What will the computer be looking for? Or do you propose outsourcing overseas to some manual eye-ball scans of all picture/movie/sound uploads?
You then get into all sorts of "civil liberty" problems, not to mention data quality (too many false-positives and false-negatives), and leaving Limewire users wide-open to attack from civil libitarians making pay-offs to the outsourced scanners, especially targetting any convicted pedofiles.
I still say, in the area of "censoring" the net against child porn, it would be far more effective to leave the porn online, mark it, and then track the people who download it. Even then, most of the black-market kid porn is not done online anyway: it's done my snail-mailing CD-Rs full of the crap. Also, this approach of capturing the kid-porn users online does nothing to help the children who are the subject of the material.
Probably a lot sooner than NetBeans.... :-)
Um, doesn't that same 97% of the demographic just download a pre-compiled binary? They won't have to eat the tutti-frutti.
;-)
However, I agree that getting the Moz developers to compile as suggested is probably going to be only a little less difficult.
I also think that compiler checks aren't a silver bullet, though it would probably help guard against heap overruns (by crashing the prog if an exploit is attempted, I presume). That's Java-thinking, we won't go there...
me too.
I think 1.1.5 is a back-port of some stuff? The homepage mentions OpenDocument support.
This new slashdot layout is freakin me out... looks cool, but I have to look around to find things again. Strange parallels with OOo...
But I hate the fonts too and "invested" some time to figure it out.
Newer X versons of Emacs (v 21 and up, maybe earlier) have the notion of "faces" to describe the fonts used by the editor. You can click-and-drool the config for your "default face" like so (these instructions are for GNU Emacs/X11, though I think it's the same for Xemacs):
0) Open the Options menu, choose Customize Emacs > Browse Customization Groups. A new buffer opens, containing the *Customize Browser* buffer, in Custom mode. This mode lets you click on Emacs buttons to select things.
1) Expand the Faces group (by clicking on the [+] button next to it, and then expand Basic Faces
2) Click on the Face button next to "Default". Other faces inherit basic options from this face, so for a global change like font size, this is the place to go.
3) you can change the values for the font Height, Width, Colo[u]r etc. I find the "misc-fixed" font family with a Height of 173 1/10pt (that is, 17.3 pt) to be readable and still small enough to fit two 80-char wide Emacs windows side-by-side on my 1600x1200 screen (handy for ediff).
4) When you are happy with your selections, press the State button and choose "Save for Future Sessions" from the pop-up menu. Emacs will write the necessary stuff to your ~/.emacs file which is loaded each time Emacs starts.
5) Once it's in your .emacs file, you can transfer to other machines (either the whole file, or just the font bit written by the Custom buffer). Look for the call to custom-set-faces. Mine follow (which, together with the default colours results in an LCARS-like appearance):
Spell checking isn't useful either, actually, because English is inconsistent across national borders. I quit using Word's spell check after version 6, when the stupid thing doesn't remember your language settings and insists on "US-English" with "zee"s everywhere and all the colours spelt wrong. And spell checkers actually are a hindrance in some places: take using the wrong word (principle v. principal). A grammar check won't pick up that "It's the principal that counts" is incorrect, either.
On the other hand, the presence of a very bad spell check has prompted me to learn some spelling now. I find that reading a lot helps, since I get used to how the words should look. I still make mistakes though.
It's sort of like quantised music synthesisers. Very cool in the 1980's, but it lead to very "square" unnatural, boring sounding tunes (take "Axel-F" for instance, which actually sounds better now that the "crazy frog" is ever so slightly out of time).
I think we should just drop English as the Internet language and adopt Loglan (www.loglan.org). That's what it was invented for.
I'm more into KDE myself, and it's very much like Windows, though that's not why I like it. I like konquerer better than nautilus, mainly. But I like Aqua better than everything else I've seen, and GNOME seems to be headed that way, from my viewpoint, so I keep my eye on it. And the initiative of freedesktop.org means that the two work well together, which I think is fantastic.
Still, each to their own. Where would you like GUIs to go? It's a very interesting research question. So far there is nothing perfect, perhaps their won't ever be. Maybe we'll eventually end up with something like Plan9 or LCARS (which both feature hightly configurable, pannel-like interfaces, though LCARS is fiction and heavily supplanted by an oral interface), or maybe we'll develop something else entirely. But so long as we are stuck with a 2D pallette/mindset, we won't get a lot further anyway.
However for a "mainstream" GUI, where should GNOME/KDE/Commercials be aiming? For a future that is not supported by current tech, or for something that's the best we can do within current limitations?
um, my spelling is intolerable...? whoops! :-)
yup, which is how I read it, and asumed a different poster than SpiderMonkey.
I think it's a reflection on the sad events in recent history: we've all been burned by Microsoft, and with the SCO lunacy and patents on just about anything, it's probably how we are being conditioned.
I'm going to work out how to file a patent on "Method to file obvious techniques with USPTO in such a way that the USPTO will not recognise it as obvious or frivolous". Then the next patent Microsoft submits I can sue them over....
Good point.
My take is that it's a pledge, not a contract. It can't be a contract unless there are some joint parties to sign an agreement. Consideration is not necessary but a written contract between parties is required for it to be legally binding.
My take on this pledge is that CA won't sue you, unless you sue first, and then CA will use these pledged patents as amunition, if you use any of them. It's sort of like the patent clause on GPL3.
If you want it to be legally binding, then ideally the Patent Commons should be an entity, like the OSDL or FSF (though probably not actually one of these) which can sign a contract on behalf of a group of OSS developers. I would like to see a development in this direction. If companies just gave their patents away to Public Domain, then they lose the ability to "fight back" if somebody else sues first.