There is a similar approach I once heard: Use 4 well-defined icons to encode any 32-bit value. You would only need 256 distinct icons to have a complete coverage. Remembering 4 icons like that is fairly easy. Plus if your application only uses 256 or 65K errors, you can reduce the number of icons to 1-2.
* Advertising
"Windows better than everything..." advertising campaign might be one. Massive consumer bombardment with "Windows 7" ads similar to what they did with Windows 95.
* Hardware
Microsoft might follow Apple and Oracle, and start making their own hardware. Massively parallel chips geared towards both the vector and regular computations would be one idea.
Alternatively, their own servers (totally not their market segment, plus they will aggravate their relationship with Dell and others, so this is less likely.
* M & A...
Buying Yahoo? Or better yet - Novel? That would be an interesting development. Novel has very little influence compared to their former glory, yet some of their technology (Moonlight?) might be valuable to MS. So instead of discrediting Mono project, MS might simply jump on it and start offering various open source solutions... I know I'm daydreaming...
The process of the bill writing seems to me to be very similar with how the Wikipedia articles get started / mature. Wikipedia API was designed specifically to work with the bulk data (see http://en.wikipedia.org/w/api.php ) - we can just adapt a similar approach.
(Shameless plug: I was the dev who implemented the original wiki api)
Classic Quest for Glory is out
on
Quests
·
· Score: 5, Informative
Recently a non-profit group, after 8 years in the making, finally released a remake of the old Sierra game Quest For Glory II - with a point-and-click interface. I was a big fan in the days, so in case anyone interested - http://agdinteractive.com/
Am I correct that with Verizon, being one of the few remaining non-GSM providers, CDMA is scheduled to disappear from North America? Does it still mean the multiple frequencies will remain, thus EU-GSM will stay on separate frequencies with US-GSM ?
I like RIAA no more than an average slashdoter, and IANAL, but I don't think this defense will hold - lots of companies in an industry pull their resources together for a common goal. For example, the popular "Got milk?" commercial was licensed by many industry players - effectively doing similar thing -- http://en.wikipedia.org/wiki/Got_Milk%3F
I am sure milk producers are competing with one another, but their common goal is to make the market larger, and that is what the commercials pursued.
I suspect Microsoft needs a common programming platform, and its Mono they are after.
The adaption of.NET in the enterprise was very slow, mostly because most backends have been exclusively Unix/Linux based. Having two infrastructures at the same time is fairly expensive to maintain - an enterprise basically needs two groups of admins.
Mono solves the problem of running.net on Unix, but its legal status makes many people worried, thus Java is much heavier present in the enterprise, thus eventually it will get to the point of having as nice UI as WinForms from both visual and developer's perspective. The moment it happens, being a cross-platform Java will run on both Unix & Windows - not good for MS.
This partnership sends a clear message to all enterprise architects: Mono is OK, we won't sue you. The extent of this is unclear... Will wait and see:)
There is now Firefox 2 support enabled on all Wiki*edia sites. To use, navigate to http://en.wikipedia.org/ (or any other language/project), click the search engine selector button in the upper right corner, and click "add wikipedia". The added bonus is that auto-suggest is also working - as you type you search, it will provide a list of page titles that begin with the typed letters.
One note - the timeout is set to 500ms, which is not too long (especially when the entire slashdot visits wiki). To make it longer, open firefox_install_dir\components\nsSearchSuggestions .js, and edit the "_suggestionTimeout: 500" line. Something like 2000 works fine for me.
I recently enabled support for the new Firefox 2.0 auto-suggest search engine feature on all Wikimedia servers. Wikipedia will provide suggestions to your search as you type in the search box. To enable, visit any Wiki-site (i.e. http://en.wikipedia.org/ ), and click the Engine Selector button (to the left of the search box). Click "Add Wikipedia". Afterwards, when you start typing in the search box while having Wikipedia engine selected, titles will automatically appear. Sometimes a FF restart is needed for the feature to begin to work. If you have any questions or suggestions, leave me a comment at http://en.wikipedia.org/wiki/User_talk:Yurik . At some point more relevant search will be implemented as well.
a. Not under econ embargo
b. Pressured by WTO and specific countries (US) to play nice with regards to IP
c. China the government not the same as China the people - the government doesn't care as much about the money, they care much more about control, and OSS gives them full control. How many governments would want a closed source software from another country used in their military?
It sounds like a great opportunity for open source software.
Speaking from personal experience from the good ol' Russia, I would disagree that open source software will proliferate. Strike it as flamebate, but given the choice of ANY software available for FREE (beer), the software that has the highest number of the most "common" applications will become ubiquitous. This means - everyone will have windows, photoshop, office plus whatever else that has high value, without any regards to price/advertising. Average Joe might not want to invest his time into less polished Linux for desktop, thus M$ is what everyone will have. Apparently, to the average consumer, the value of OSS is not stability or openness, but the word FREE.
I might be wrong, but i think what this proposal talks about is not just kazaa. Kazaa came into the picture to support it, but it goes far beyond that -- ISPs will try to detect filesharing participation, for ANY network! (edonkey, etc) You can use anything you like, but unless the request is "HTTP" or "SSH" or some other "legitimate" protocol, they can start charging extra.
Of course we can create networks that relies only on the http requests, but protocol would have to have some sort of distinct feature that ISPs can filter on once the network becomes big.
This might be similar to the law that canadians have to pay a CDR fee that automatically goes to the "artists", and, as a result, would legitimize the downloads (i already paid for the music i downloaded because my ISP charged me for it). What it also means is that the big companies will receive free money for doing nothing at all.
(singing) "It's the end of the life as we know it..."
Wasn't Russian space station placed into orbit with exactly the same expectations, and had people living onboard for very extended (over 6 months) periods? The new station is in reality not very different from MIR -- longer orbital stay, prolonged human presence, etc, but nothing fundamentally changed.
At some point, the station will come down or destroyed, and the "forever" will be over. If not, Russians might claim they started the permanent human presence in space, only it had been interrupted due to technical difficulties a few times.
IP is already used internally. There is no reason to integrate it into MSDEV; it has its own editor, compiler, and debuger. Its editor is fundamentally different from MSDEV text based environment. Yes, 2-3 years if a fair estimate, despite the begining of the productions. Yes, it really blows -- no competitors in sight, MS will have the monopoly on new development paradigm and everyone will bitch again.
Funny enough, I know of a project that might revolutionize code development, and it is being developed at everyone's favorite software company, with the research help from Oxford University programming tools group. Yep, you guessed it - Microsoft!
Intentional Programming Official M$ research site has severely limited info on the project, some research papers, etc, but does not present the true picture of the potential. Another good site has some more info. The project recently moved to development stage, and the transfer order sent to every VP has a very good description of some of the features. I had a chance to look at the beta version, but NDA doesn't permit me to talk of much... My guess is, M$ will try to develop and use this thing for their internal development to gain the edge on the rest of dev community. There is really no other reason they should hide all the demos and descriptions of it.
The brain behind the project is Charles Simonyi, M$ chef technology guy, who was one of the researchers working on WYSIWYG at Xerox, and formalized Hungarian Notation. He has written some papers on the topic of Intentional Programming.
This thing is by far superior to code parsers and other on-the-fly code generators. Very good presentation is here - you will need a free Power Point viewer. Code can be written in any language, with any syntax. It can assist with every structured text, even HTML and XML. Internally everything is stored similar to compiler-generated parse tree. Every item is a reference. Identifiers can be renamed at any moment, without complicated Search and Replace. Code can easily be moved between projects because it simply re-computes the references. Every chunk of code can be made into a function with a stroke of a key. At a higher level, methods for performing tasks can be formalized in such a way as to have no concrete data types, only abstractions (somewhat akin to STL, but taken to the language level and extended). I do not like some of the things M$ does, but cannot ignore a good idea when I see one.
Sounds like this is turning into a job fair. Instead of discussing the issue at hand (shortage vs. plentitude of geeks in this country) we are submitting our resumes!!! Maybe./ should go under ".com"?
Just to "be original": i am looking for a good c++ position. Have been developing proffessional software since i was 12, and can prove it too!
There is a similar discussion i started here, proposing an alternative to Napster's centralized database. IRC's idea is very good, but it still relies on centralized servers, that potentally could altered to refuse any standartized trafic from MP3 search clients.
From the discussion so far, it seams that FreeNet has several great ideas in progress, that theoretically could be used for our cause. The actual media files do not need to be uploaded, only an XML label for a file with host IP and a limited TTL. It can be propagated through the network, and will get refreshed on a regular basis if the provider stays on-line. If the nodes move quickly enough, and the search engine has enough flexibility, this addition to freenet might benefit it get jumpstarted.
This method is not very secure -- the host IP addresses will be in the plain view, but I doubt much can be done if pear-to-pear connection is desired.
I've just looked at freenet, and it seam to have several major problems:
1. It seams to be very susceptible to "flood attacks". A corporation can inject a large piece of random junk into the net, and later have a very large number of clients all over the world continuously request it. This will force immediate propagation, which in turn will push useful info off the storage of limited size. Since such malicious network can easily be organized by even a small size company, and the amounts of random junk is infinite, no useful info will ever survive in such scenario.
2. Media files are very large, and users will always like to use mp3 files locally. Freenet approach does not allow the user to listen to her files, while at the same time allowing others to download them. Doing so would save substantial amount of storage resources.
Napster has two major weaknesses: it's owned by a single company and has a central search database. Big $$ can easily fight a small company by simply overspending it on legal expenses, and shutting down the "brain" -- main SQL computers.
Proposal: an open source, distributed database search engine, where the search is performed in parallel by a large number of logged-in clients, and results are combined on the requesting client. When a new client runs, it uploads file info onto several other clients, whose location is found from the cached results of the last run / public web sites / hidden ftp servers / information from friends / whatever else you might come up with. Search can be performed akin the infamous worm viruses with graph search depth limitations.
As for the name, we can call it World Wide Search (WWS), or some other silly choice to attract attention. Obviously, this thing has to be portable enough to run on all platforms, including non-Linux.
E-mail me if you are interested. Remove "SPAM" twice to reply.
Napster has two major weaknesses: it's owned by a single company and has a central search database. Big $$ can easily fight a small company by simply overspending it on legal expenses, and shutting down the "brain" -- main SQL computers.
Proposal: an open source, distributed database search engine, where the search is performed in parallel by a large number of logged-in clients, and results are combined on the requesting client. When a new client runs, it uploads file info onto several other clients, whose location is found from the cached results of the last run / public web sites / hidden ftp servers / information from friends / whatever else you might come up with. Search can be performed akin the infamous worm viruses with graph search depth limitations.
As for the name, we can call it World Wide Search (WWS), or some other silly choice to attract attention. Obviously, this thing has to be portable enough to run on all platforms, including non-Linux.
Slashdot Mirror
At one time Prince was hip and suddenly he became outdated...
There is a similar approach I once heard: Use 4 well-defined icons to encode any 32-bit value. You would only need 256 distinct icons to have a complete coverage. Remembering 4 icons like that is fairly easy. Plus if your application only uses 256 or 65K errors, you can reduce the number of icons to 1-2.
I can see a few alternatives:
* Advertising
"Windows better than everything..." advertising campaign might be one. Massive consumer bombardment with "Windows 7" ads similar to what they did with Windows 95.
* Hardware
Microsoft might follow Apple and Oracle, and start making their own hardware. Massively parallel chips geared towards both the vector and regular computations would be one idea.
Alternatively, their own servers (totally not their market segment, plus they will aggravate their relationship with Dell and others, so this is less likely.
* M & A... ... I know I'm daydreaming...
Buying Yahoo? Or better yet - Novel? That would be an interesting development. Novel has very little influence compared to their former glory, yet some of their technology (Moonlight?) might be valuable to MS. So instead of discrediting Mono project, MS might simply jump on it and start offering various open source solutions
The process of the bill writing seems to me to be very similar with how the Wikipedia articles get started / mature. Wikipedia API was designed specifically to work with the bulk data (see http://en.wikipedia.org/w/api.php ) - we can just adapt a similar approach.
(Shameless plug: I was the dev who implemented the original wiki api)
Recently a non-profit group, after 8 years in the making, finally released a remake of the old Sierra game Quest For Glory II - with a point-and-click interface. I was a big fan in the days, so in case anyone interested - http://agdinteractive.com/
Am I correct that with Verizon, being one of the few remaining non-GSM providers, CDMA is scheduled to disappear from North America? Does it still mean the multiple frequencies will remain, thus EU-GSM will stay on separate frequencies with US-GSM ?
I like RIAA no more than an average slashdoter, and IANAL, but I don't think this defense will hold - lots of companies in an industry pull their resources together for a common goal. For example, the popular "Got milk?" commercial was licensed by many industry players - effectively doing similar thing -- http://en.wikipedia.org/wiki/Got_Milk%3F
I am sure milk producers are competing with one another, but their common goal is to make the market larger, and that is what the commercials pursued.
Wikimedia Foundation already has a project called WikiSpecies -- http://species.wikimedia.org/wiki/Main_Page . Not sure how different that project will be.
I suspect Microsoft needs a common programming platform, and its Mono they are after.
.NET in the enterprise was very slow, mostly because most backends have been exclusively Unix/Linux based. Having two infrastructures at the same time is fairly expensive to maintain - an enterprise basically needs two groups of admins.
.net on Unix, but its legal status makes many people worried, thus Java is much heavier present in the enterprise, thus eventually it will get to the point of having as nice UI as WinForms from both visual and developer's perspective. The moment it happens, being a cross-platform Java will run on both Unix & Windows - not good for MS.
:)
The adaption of
Mono solves the problem of running
This partnership sends a clear message to all enterprise architects: Mono is OK, we won't sue you. The extent of this is unclear... Will wait and see
There is now Firefox 2 support enabled on all Wiki*edia sites. To use, navigate to http://en.wikipedia.org/ (or any other language/project), click the search engine selector button in the upper right corner, and click "add wikipedia". The added bonus is that auto-suggest is also working - as you type you search, it will provide a list of page titles that begin with the typed letters.
s .js, and edit the "_suggestionTimeout: 500" line. Something like 2000 works fine for me.
One note - the timeout is set to 500ms, which is not too long (especially when the entire slashdot visits wiki). To make it longer, open firefox_install_dir\components\nsSearchSuggestion
--Yurik / http://en.wikipedia.org/wiki/User:Yurik
I recently enabled support for the new Firefox 2.0 auto-suggest search engine feature on all Wikimedia servers.
Wikipedia will provide suggestions to your search as you type in the search box. To enable, visit any Wiki-site (i.e. http://en.wikipedia.org/ ), and click the Engine Selector button (to the left of the search box). Click "Add Wikipedia". Afterwards, when you start typing in the search box while having Wikipedia engine selected, titles will automatically appear. Sometimes a FF restart is needed for the feature to begin to work. If you have any questions or suggestions, leave me a comment at http://en.wikipedia.org/wiki/User_talk:Yurik . At some point more relevant search will be implemented as well.
China is a slightly different case:
a. Not under econ embargo
b. Pressured by WTO and specific countries (US) to play nice with regards to IP
c. China the government not the same as China the people - the government doesn't care as much about the money, they care much more about control, and OSS gives them full control. How many governments would want a closed source software from another country used in their military?
It sounds like a great opportunity for open source software.
Speaking from personal experience from the good ol' Russia, I would disagree that open source software will proliferate. Strike it as flamebate, but given the choice of ANY software available for FREE (beer), the software that has the highest number of the most "common" applications will become ubiquitous. This means - everyone will have windows, photoshop, office plus whatever else that has high value, without any regards to price/advertising. Average Joe might not want to invest his time into less polished Linux for desktop, thus M$ is what everyone will have. Apparently, to the average consumer, the value of OSS is not stability or openness, but the word FREE.
<begin flame here>
I might be wrong, but i think what this proposal talks about is not just kazaa. Kazaa came into the picture to support it, but it goes far beyond that -- ISPs will try to detect filesharing participation, for ANY network! (edonkey, etc) You can use anything you like, but unless the request is "HTTP" or "SSH" or some other "legitimate" protocol, they can start charging extra.
Of course we can create networks that relies only on the http requests, but protocol would have to have some sort of distinct feature that ISPs can filter on once the network becomes big.
This might be similar to the law that canadians have to pay a CDR fee that automatically goes to the "artists", and, as a result, would legitimize the downloads (i already paid for the music i downloaded because my ISP charged me for it). What it also means is that the big companies will receive free money for doing nothing at all.
(singing) "It's the end of the life as we know it..."
At some point, the station will come down or destroyed, and the "forever" will be over. If not, Russians might claim they started the permanent human presence in space, only it had been interrupted due to technical difficulties a few times.
--
I am American: don't flame me, use the flag!
IP is already used internally. There is no reason to integrate it into MSDEV; it has its own editor, compiler, and debuger. Its editor is fundamentally different from MSDEV text based environment. Yes, 2-3 years if a fair estimate, despite the begining of the productions.
Yes, it really blows -- no competitors in sight, MS will have the monopoly on new development paradigm and everyone will bitch again.
Intentional Programming Official M$ research site has severely limited info on the project, some research papers, etc, but does not present the true picture of the potential. Another good site has some more info. The project recently moved to development stage, and the transfer order sent to every VP has a very good description of some of the features. I had a chance to look at the beta version, but NDA doesn't permit me to talk of much... My guess is, M$ will try to develop and use this thing for their internal development to gain the edge on the rest of dev community. There is really no other reason they should hide all the demos and descriptions of it.
The brain behind the project is Charles Simonyi, M$ chef technology guy, who was one of the researchers working on WYSIWYG at Xerox, and formalized Hungarian Notation. He has written some papers on the topic of Intentional Programming.
This thing is by far superior to code parsers and other on-the-fly code generators. Very good presentation is here - you will need a free Power Point viewer. Code can be written in any language, with any syntax. It can assist with every structured text, even HTML and XML. Internally everything is stored similar to compiler-generated parse tree. Every item is a reference. Identifiers can be renamed at any moment, without complicated Search and Replace. Code can easily be moved between projects because it simply re-computes the references. Every chunk of code can be made into a function with a stroke of a key. At a higher level, methods for performing tasks can be formalized in such a way as to have no concrete data types, only abstractions (somewhat akin to STL, but taken to the language level and extended). I do not like some of the things M$ does, but cannot ignore a good idea when I see one.
Hmmm....
./ should go under ".com"?
Sounds like this is turning into a job fair. Instead of discussing the issue at hand (shortage vs. plentitude of geeks in this country) we are submitting our resumes!!! Maybe
Just to "be original": i am looking for a good c++ position. Have been developing proffessional software since i was 12, and can prove it too!
There is a similar discussion i started here, proposing an alternative to Napster's centralized database. IRC's idea is very good, but it still relies on centralized servers, that potentally could altered to refuse any standartized trafic from MP3 search clients.
From the discussion so far, it seams that FreeNet has several great ideas in progress, that theoretically could be used for our cause. The actual media files do not need to be uploaded, only an XML label for a file with host IP and a limited TTL. It can be propagated through the network, and will get refreshed on a regular basis if the provider stays on-line. If the nodes move quickly enough, and the search engine has enough flexibility, this addition to freenet might benefit it get jumpstarted.
This method is not very secure -- the host IP addresses will be in the plain view, but I doubt much can be done if pear-to-pear connection is desired.
I've just looked at freenet, and it seam to have several major problems:
1. It seams to be very susceptible to "flood attacks". A corporation can inject a large piece of random junk into the net, and later have a very large number of clients all over the world continuously request it. This will force immediate propagation, which in turn will push useful info off the storage of limited size. Since such malicious network can easily be organized by even a small size company, and the amounts of random junk is infinite, no useful info will ever survive in such scenario.
2. Media files are very large, and users will always like to use mp3 files locally. Freenet approach does not allow the user to listen to her files, while at the same time allowing others to download them. Doing so would save substantial amount of storage resources.
Napster has two major weaknesses: it's owned by a single company and has a central search database. Big $$ can easily fight a small company by simply overspending it on legal expenses, and shutting down the "brain" -- main SQL computers.
Proposal: an open source, distributed database search engine, where the search is performed in parallel by a large number of logged-in clients, and results are combined on the requesting client. When a new client runs, it uploads file info onto several other clients, whose location is found from the cached results of the last run / public web sites / hidden ftp servers / information from friends / whatever else you might come up with. Search can be performed akin the infamous worm viruses with graph search depth limitations.
As for the name, we can call it World Wide Search (WWS), or some other silly choice to attract attention. Obviously, this thing has to be portable enough to run on all platforms, including non-Linux.
E-mail me if you are interested.
Remove "SPAM" twice to reply.
Proposal: an open source, distributed database search engine, where the search is performed in parallel by a large number of logged-in clients, and results are combined on the requesting client. When a new client runs, it uploads file info onto several other clients, whose location is found from the cached results of the last run / public web sites / hidden ftp servers / information from friends / whatever else you might come up with. Search can be performed akin the infamous worm viruses with graph search depth limitations.
As for the name, we can call it World Wide Search (WWS), or some other silly choice to attract attention. Obviously, this thing has to be portable enough to run on all platforms, including non-Linux.
E-mail me if you are interested.