For instance office type docs Word, Star Writer, etc... can have the text extracted and used. mp3 files can have their id3 tags read and combined with more information via various online databases. Most image types contain a header section with various bits of information. Zip, bzip, gzip files can be opened and read. If you wanted to be really thourough, you could run something like "strings" against unknown binary types to match the data.
Granted I'm still a long way from a solution. I think that perhaps something like a vitual filesystem (personally I still like trees) could be created. This could be updated when new files are added or index the file system periodically. From here you could categorize files of different types. Given, this takes some effort up front but, should get smarter over time.
For each file you add, the system will try to auto-categorize. You verify if its correct or not and tell them system when its wrong.
When you need a file, just follow the category. This will probably work better for older files that you've forgotten about. The tree system is still there for when you know where it is.
A local search engine (which I'm working on BTW) could index the files so they are easier to find later as well.
Perhaps the "smart" behavior could be created by using a distributed bayesian filter against the contents of a file.
This worked for a while to categorize spam. Perhaps we could use it to categorize documents automagically. With the bonus that everytime its wrong, it learns more. The "distributed" part is where we share our filters and gain from each others effort of training the filters.
But, don't you realize? Dot com's are different.
They don't need a business plan, because dot coms are different. They don't need to make a profit, because dot coms are different. Providing products and services, is so passe, because dot coms are different.
Isn't that all we heard during the go-go bubble days.
You'll notice that having a business plan isn't part of the schedule.
This is really very cool. It seems a lot like how the Plan9 OS is able to treat items on the network as if they are local. Its interesting to see someone else recognize the value of the concept.
Of course, I worry about the security of ubiquitous networking.
Most ISP's run a pretty slim margin. The cost of investigation and possible litigation are more than they are willing to pay. Its cheapr to just concede.
Looking further. Is it the ISP's responsibility to fight for YOUR rights at their cost?
Now if our legal system made sense, people would not be able to sue an ISP if you placed said content their. The should sue YOU. Since they do sue the ISP, they are only going to try to protect themselves.
This means that the cost of fighting for your rights is yours, and yours alone. As it should be.
Personally, I tired of people whinning because others don't fight for their rights. Of course, I am one of those whacky individualists who thinks that it is no one's responsibility (except mine) to take care of me.
I guess I did....
I picked up one of those "lap desks" from the book store for this very purpose. It is basically, a small plasic deck with a pillow like surface on the bottom. It fits nicely in my backback or brief case by squishing the air out of the pillow part. Very useful for using the laptop on the bus.
Why didn't I think to patent "the application of the lap desk as a laptop insulator". I could be collecting some ill gotten litigation money.....
oh well, maybe next time
I have this idea of using clothes pins to keep bags of chips closed. Now if I were to make the clip wider... what? Doh!
I doubt it.
The companies I've worked for spent a huge effort to ensure that the *fringe* browsers are supported.
This is not comprehensive, but it does mean that the sites must work with Moz and Safari, as well as MSIE.
Companies that are interested in making money will not ignore users who are not using the mainstream browser. Often this has less to do with making money off of those individuals but, more to do with the geek influence factor.
My rough explanation of the "geek influence factor": RealNetworks (save the flames for a moment) has always offered a player for a large variety of platforms, including linux. They only make money off of the Windows player (the mac player consistently loses money). Why would they do this you ask?
Well, imagine you run a company and want to offer streaming media. WHo do you ask about which system is the way to go? Why your geeks. Your IT dept. Well, geeks tend not to use Windows (ok, posers do but,.....). The geeks will take a bit of a self centered view while still supporting the mainstream. THat would leave MS and Apple out in the cold.
Now lets add to this the recent offerings of Linows and the Java Desktop pre-installed on Walmart computers. Now that we see non-Windows desktops coming from Walmart for under $300, companies can not ignore supporting them. Their numbers will be too big.
The funny thing is, everbody sees dot net as trying to lock all into windows but, its the opposite. dot net uses a virtual machine (the CLR) this means that porting an app to a new OS is simply a matter of porting the CLR to the new platform. This is why you hear so much of dot net is a Java copy. I will argue that dot net is the MS application developers way of preparing to lose dependence on the Windows OS. Before you rant, think about the possibilies here.
It may appear a contradiction but, the possibility is very strong.
You are basically right. This is a classic case of different parts of the same company fighting with each other. Sony mp3 players and living room CD copiers vs Sony the RIAA member.
The Java camp really want to embrace Linux. They know that the success of Linux on the server is what made J2EE as big as it is today.
The SPARC/Solaris side of the business doesn't want to let go of a dying business. When you can get 10 x86 boxes with Linux for less than one SPARC with Solaris, its obvious where the market is going. And with the publicity of the success of the Google super cluster, people are taking it seriously.
I expect that the next *contrarion* software we'll see happens when Linux gets just a wee bit bigger. Lets face it, when Linux overtakes Mac on the desktop (this year?) we'll soon see MS offering their Office product for Linux. The guys who work on office want to grow their distribution. They could care less about the OS, or they wouldn't offer it for the Mac. (is dot net pre-emptive thinking here?)
I expect we have some interesting times ahead of us. We'll see big companies moving their internal fights into the public area.
Maybe its the result of a constant demand for newer versions and features.
Many companies (mine included) are under constant pressure to release new products with more features. Often this means that instead of fixing bugs making products better, we are working on the next release.
Perhaps the constant *beta* is a marketing ploy. Tell people its beta and they can't complain when bugs are found. So what if the product never comes out of beta.
Classic example. Has there ever been a non-beta version of ICQ?
This is an example of the "problem" with Linux (and the perpetually beta software packaged with it) going mainstream.
In the *old* days. When a Linux user had a problem with a program in a linux distro he/she fixed it, and sent a patch so that others could benefit from the fix. This resulted in these systems contiually getting better (the opposite of OS rot windows users are so familiar with).
Now that non developers are increasingly using these distros, there is a lot more complaining but, not an associated amount of fixing going on.
People seem to forget that the majority of the development on any of these distros is done for free (read Joe/Jane developer working in his/her free time). The professional developers working for RedHat, Mandrake, etc... are mostly building config tools.
The result of this, the developers who actually build the apps get more and more abuse, without a cooresponding amount of help. We've already seen many developers drop out of projects for this reason.
I would suggest the author and others who feel as he to think about this. If you want to make linux better without (actually doing it) writing code, encourage the developers. Let them know about things that don't work so well. If you want a new feature, try offering a bribe. Say, "I'll mail some beer to whoever implements......" or similar.
You are perhaps missing the bigger issue. Which comes down to two points.
1. It used to be this way
2. It should be easily configurable
However, I do agree with you that just spouting negative comments is unproductive. (read: if you piss off the developers they will be less accomodating) My suggestion to the author of the article is, if you don't like it, switch. Either to a previous version of Gnome or to another desktop. One of the great things about Linux is choice. Not every desktop/window manager has to do everything. You can pick the one that best meets your needs. Last I looked there are RPM's for KDE, IceWin, Blackbox, and at least 10 others. Instead of a rant he could say, "I don't like the default WM because...." and then chose one that fits him better.
Re:honest about Mac troubles
on
Fix a Troubled Mac
·
· Score: 2, Informative
amen
Ibet Mac OS8 is the only system that crashes more than windows.
Now that is an honor to hold
Slashdot Mirror
Now we just need google for the desktop...
Lets rephrase that to *some* binary files.
For instance office type docs Word, Star Writer, etc... can have the text extracted and used. mp3 files can have their id3 tags read and combined with more information via various online databases. Most image types contain a header section with various bits of information. Zip, bzip, gzip files can be opened and read. If you wanted to be really thourough, you could run something like "strings" against unknown binary types to match the data.
Granted I'm still a long way from a solution. I think that perhaps something like a vitual filesystem (personally I still like trees) could be created. This could be updated when new files are added or index the file system periodically. From here you could categorize files of different types. Given, this takes some effort up front but, should get smarter over time.
For each file you add, the system will try to auto-categorize. You verify if its correct or not and tell them system when its wrong.
When you need a file, just follow the category. This will probably work better for older files that you've forgotten about. The tree system is still there for when you know where it is.
A local search engine (which I'm working on BTW) could index the files so they are easier to find later as well.
Perhaps the "smart" behavior could be created by using a distributed bayesian filter against the contents of a file.
This worked for a while to categorize spam. Perhaps we could use it to categorize documents automagically. With the bonus that everytime its wrong, it learns more. The "distributed" part is where we share our filters and gain from each others effort of training the filters.
Thoughts?
But, don't you realize? Dot com's are different.
They don't need a business plan, because dot coms are different. They don't need to make a profit, because dot coms are different. Providing products and services, is so passe, because dot coms are different.
Isn't that all we heard during the go-go bubble days.
You'll notice that having a business plan isn't part of the schedule.
Basically, there are a few slight differences, but overall, they're the same language with different syntax.
You could argue this is more or less true for all object oriented languages.
This is really very cool. It seems a lot like how the Plan9 OS is able to treat items on the network as if they are local. Its interesting to see someone else recognize the value of the concept.
Of course, I worry about the security of ubiquitous networking.
You could also argue that a professional developer may not want to switch to a language where "Beginner" is part of its acronym.......
Basic (Beginners Allpurpose Symbolic Instruction Code)
Lets be real about this.
Most ISP's run a pretty slim margin. The cost of investigation and possible litigation are more than they are willing to pay. Its cheapr to just concede.
Looking further. Is it the ISP's responsibility to fight for YOUR rights at their cost?
Now if our legal system made sense, people would not be able to sue an ISP if you placed said content their. The should sue YOU. Since they do sue the ISP, they are only going to try to protect themselves.
This means that the cost of fighting for your rights is yours, and yours alone. As it should be.
Personally, I tired of people whinning because others don't fight for their rights. Of course, I am one of those whacky individualists who thinks that it is no one's responsibility (except mine) to take care of me.
I guess I did.... I picked up one of those "lap desks" from the book store for this very purpose. It is basically, a small plasic deck with a pillow like surface on the bottom. It fits nicely in my backback or brief case by squishing the air out of the pillow part. Very useful for using the laptop on the bus. Why didn't I think to patent "the application of the lap desk as a laptop insulator". I could be collecting some ill gotten litigation money..... oh well, maybe next time I have this idea of using clothes pins to keep bags of chips closed. Now if I were to make the clip wider... what? Doh!
Even the "Human" tested images don't work anymore.
Many porn sites have "clustered" humans by having people enter the text to get access.
We need a way to have a human test that also verifies that it is a human making the post.
The only thing I can think of is digitally signing everything. But, that would mean giving up the MYTH of being anonymous on the web.
---
I doubt it. The companies I've worked for spent a huge effort to ensure that the *fringe* browsers are supported. This is not comprehensive, but it does mean that the sites must work with Moz and Safari, as well as MSIE. Companies that are interested in making money will not ignore users who are not using the mainstream browser. Often this has less to do with making money off of those individuals but, more to do with the geek influence factor. My rough explanation of the "geek influence factor": RealNetworks (save the flames for a moment) has always offered a player for a large variety of platforms, including linux. They only make money off of the Windows player (the mac player consistently loses money). Why would they do this you ask? Well, imagine you run a company and want to offer streaming media. WHo do you ask about which system is the way to go? Why your geeks. Your IT dept. Well, geeks tend not to use Windows (ok, posers do but, .....). The geeks will take a bit of a self centered view while still supporting the mainstream. THat would leave MS and Apple out in the cold.
Now lets add to this the recent offerings of Linows and the Java Desktop pre-installed on Walmart computers. Now that we see non-Windows desktops coming from Walmart for under $300, companies can not ignore supporting them. Their numbers will be too big.
The funny thing is, everbody sees dot net as trying to lock all into windows but, its the opposite. dot net uses a virtual machine (the CLR) this means that porting an app to a new OS is simply a matter of porting the CLR to the new platform. This is why you hear so much of dot net is a Java copy. I will argue that dot net is the MS application developers way of preparing to lose dependence on the Windows OS. Before you rant, think about the possibilies here.
It may appear a contradiction but, the possibility is very strong.
You are basically right. This is a classic case of different parts of the same company fighting with each other. Sony mp3 players and living room CD copiers vs Sony the RIAA member.
The Java camp really want to embrace Linux. They know that the success of Linux on the server is what made J2EE as big as it is today.
The SPARC/Solaris side of the business doesn't want to let go of a dying business. When you can get 10 x86 boxes with Linux for less than one SPARC with Solaris, its obvious where the market is going. And with the publicity of the success of the Google super cluster, people are taking it seriously.
I expect that the next *contrarion* software we'll see happens when Linux gets just a wee bit bigger. Lets face it, when Linux overtakes Mac on the desktop (this year?) we'll soon see MS offering their Office product for Linux. The guys who work on office want to grow their distribution. They could care less about the OS, or they wouldn't offer it for the Mac. (is dot net pre-emptive thinking here?)
I expect we have some interesting times ahead of us. We'll see big companies moving their internal fights into the public area.
My opinion, its lots of free entertainment.....
Maybe its the result of a constant demand for newer versions and features.
Many companies (mine included) are under constant pressure to release new products with more features. Often this means that instead of fixing bugs making products better, we are working on the next release.
Perhaps the constant *beta* is a marketing ploy. Tell people its beta and they can't complain when bugs are found. So what if the product never comes out of beta.
Classic example. Has there ever been a non-beta version of ICQ?
This is an example of the "problem" with Linux (and the perpetually beta software packaged with it) going mainstream.
......" or similar.
In the *old* days. When a Linux user had a problem with a program in a linux distro he/she fixed it, and sent a patch so that others could benefit from the fix. This resulted in these systems contiually getting better (the opposite of OS rot windows users are so familiar with).
Now that non developers are increasingly using these distros, there is a lot more complaining but, not an associated amount of fixing going on.
People seem to forget that the majority of the development on any of these distros is done for free (read Joe/Jane developer working in his/her free time). The professional developers working for RedHat, Mandrake, etc... are mostly building config tools.
The result of this, the developers who actually build the apps get more and more abuse, without a cooresponding amount of help. We've already seen many developers drop out of projects for this reason.
I would suggest the author and others who feel as he to think about this. If you want to make linux better without (actually doing it) writing code, encourage the developers. Let them know about things that don't work so well. If you want a new feature, try offering a bribe. Say, "I'll mail some beer to whoever implements
Just my 2 cents.
You are perhaps missing the bigger issue. Which comes down to two points.
1. It used to be this way
2. It should be easily configurable
However, I do agree with you that just spouting negative comments is unproductive. (read: if you piss off the developers they will be less accomodating)
My suggestion to the author of the article is, if you don't like it, switch. Either to a previous version of Gnome or to another desktop.
One of the great things about Linux is choice. Not every desktop/window manager has to do everything. You can pick the one that best meets your needs.
Last I looked there are RPM's for KDE, IceWin, Blackbox, and at least 10 others.
Instead of a rant he could say, "I don't like the default WM because...." and then chose one that fits him better.
amen Ibet Mac OS8 is the only system that crashes more than windows. Now that is an honor to hold