The connection machine's processors were distributed among multiple cards, a single card contained 16 processors, the 1 bit processors were implemented using a ROM chip, I think.
UTF-16 is an encoding which explains how to map bytes to code-points (what you call characters), like UTF-8. UTF-16 encodes data in chunks of 16 bits, while UTF-8 encodes the data in chunks of 8 bits. UCS-2 was an encoding where only the 2^16 first code-points could be encoded, in the same way that ASCII is an encoding where only the first 2^7 code-points can be expressed, and ISO-latin only encodes the 2^8 first code-points. UCS-2 was an attempt to encode the "most common case" as you describe it. The problem is, in order to achieve this, Chinese and Japanese characters were crammed together (look up Han Unification) and were basically not usable. We are talking about around 1.5 billion people here. The fix was to add back the characters that had been removed, and go above the FFFF line.
As to why we need trading cards and smiley in Unicode, the reason is pretty simple: compatibility. The goal is to be able to convert all existing text data into Unicode, this is why DOS area block drawing are defined as codepoints. Emoji were added to add compatibility to the Japanese systems so that companies like Apple could enter that market with the iPhone, without this, iPhone users would not have been able to exchange messages with other users.
Remember that at one point in time, ASCII was the extended character set with unnecessary symbols like curly braces, this is why C++ compilers still have trigraph support
It is not really once code per country, ISBN started with a code per language zone, and switched to countries when they realised it could not scale, so codes 978-0 and 978-1 are for english (this includes the mysterious lands of united kingdom and australia), code 978-2 is for french, and so does 979-10, 978-3 is for german, the followin 978- prefixes are assigned to various countries. Note that the code is not assigned to the language of the book, but the dominant language of the country / publisher. So a swiss publisher can have a 978-2 book in english.
If prices of ISBN codes were really a problem, people could just publish in France, where ISBNs are free. Anyways nowadays ISBN are just a particular class of GTIN/EAN so I suspect one could just buy an EAN (UPC) code.
If we hit the reset button, can we also fix ASCII? it is by no mean the minimal set most english speakers think it is.
Why do we need a character to represent to 'v' one after the other? You could write 'w' with to 'v' and handle the ligature where it should be handled, at display time. There are so few words in English with the sequence vv that it makes no sense to have the special case coded in the encoding.
Also could we handle the dots on the characters 'i' and 'j' like the diacriticals they are? there should be first the the dotless 'i' and 'j' and the some character to add the dots, like all other diacriticals. Also move out the currency symbols ($ and £), they can be represented as text (USD and GBP), no point in have silly symbols in there. Also remove BELL (11), having a symbol for a bell (2407) might be bloated, but having one for the sound of a bell is absurd.
By the way, why do we need different code points for upper and lower case? They are just variants of each other anyways
Unicode is certainly messy, but plain ASCII is not much better: the most precious 127 code points of utf-8 are basically wasted to display 32 characters and a bit of punctuation, that is pretty bloated for me, we are just used to it
I was a fan of Elite on the C64, I tried Oolite, and it is indeed an excellent port of Elite.
I also found it to be no fun at all: requiring time I don't have anymore and lacking the richness I'm used to find in games nowadays.
The other game I loved on the C64 was paradroid, there is also an open source port (free Droid), same problem.
My expectations and sense of fun have changed in 20 years.
You would still have to prove that they are responsible for the hack. The fact that their legitimate (if silly) business benefits from some hacked code does not prove they are responsible for the hack.
Or turn the problem around: if one provider of telecom services is ever condemned without any other proof than the fact they benefit from a hack, the bad guy just change their business model to extortion.
I saw similar magnetic connectors in household appliances in Japan before Apple used them for laptops: my rice cooker and the water heater both had them.
This is one of the reasons instruments on the LHC are duplicated: Atlas and CMS.
Given the difference of energy, if the Higgs is found in one of the LHC experiments, it is doubtful that the experiment could be reproduced at Tevatron.
While the LHC is indeed not an EU project, most of girlintraining's remain valid. Politics and national pride play an important part in the internal workings of CERN, and could well explain the communication policy.
Calling people full of nonsense because they did not get some details right is not very polite, dear Anonymous Coward. Actually, there are quite a few building in Prévessin. And while CERN is definitely not an EU project, it is different from the WHO, or the UN in the sense that it has a geographic definition, it is called the European Organization for Nuclear Research. Some of the funding for CERN related projects, like the grids efforts, comes directly from the EU.
Indeed a brazillian windows computer might not be configured to input japanese text, but I don't think this is a reason for discarding the concept, this is merely a software limitation. I'm routinely typing japanese text using a swiss-french keyboard, you just need to have a proper OS. Interestingly, my keyboard does not have the square brackets, the @ or the ~ characters directly accessible (they are reached using the alt-key), but I'm not clamoring for e-mail addresses to stop using those caracters in e-mail addresses or urls.
At any rate there might be technical hurdles, but the core problem is social. Asian people want a system that supports their own text system for addresses because using roman, non-accented characters for the job is really a kludge. Converting kanjis into plain ASCII leads to ambiguous texts or multiple possible transcriptions. For instance the city of Tôkyô can be transcribed as Tokyo, Toukyou, Tohkyoh. Actually, the correct romanization involves the use of the macron accent which is not supported by slashdot's engine (and I don't feel like mucking with escape sequences).
So using their own character system for designating servers is my opinion a valid wish. While some urls might not be typable directly, this simply means a fallback mechanism is needed. At any rate, many url are so long as to be very difficult to type. So either we have a common international system to handle this legitimate need, or every country/company will hack around it, which will be much worse.
Re:And non-compatability with your existing music
on
The Zune Cometh
·
· Score: 1
When you rip your CD with iTunes, the resulting files are either plain old MP3, or unproctected AAC files.
You are right that decrypted pages are not cached in memory. I'm don't think the issue of caching disk blocks applies, as XNU uses a unified buffer cache. So if the one encrypted page has the same size as a unencrypted page and each page can be encrypted indenpendently, the mechanism is roughtly the same as for normal code pages: the are loaded and evicted as needed, simply the load phase includes a decryption phase.
The reason code segments are mapped read-only is because during the execution of a program the code is never changed. In fact, if you have sane system you don't even have the rights to overwrite the binary exectuable file. The virtual memory system never writes out the pages that make up the code segment of a program because they are already available on the disk, in the executable file they were loaded from.
I fail to see how having a fast disk array gives you any incentive to do additional copies of the code segement on disk...
I don't think that the fact that the pager does only reading means that binaries are in wired memory. Code segments are typically only mapped in read-only mode, so they are never written back to disk, nor should they be.
Just a technically, but Apache is not released under the GPL. Apache is released under the Apache License, which is, as far as I understand, closer to the BSD license than from the GPL license.
Maybe not everyone knows what OGG is, but just about everyone has heard MP3s that have been re-encoded too many times, and most people DO understand the difference between lossless and lossy formats.
Err, No.
Most people might know their music is represented by files, and know that compressed files are smaller. I doubt the majority of people realise that MP3 or JPEG are a lossy compression schemes, or that the degradation increases when the file are decoded and re-encoded multiple times. People don't want to know about those things, they want to listen to the music.
In some sense, I have the feeling we have gone full circle. What this amounts to is having the operating system and core libraries in some sort of pseudo-ROM device, that can be read fast and written a limited number of times. This is not so different from the classical macintosh computer that had a large part of the OS in ROM, the Mac classic had even a bootable disk image in ROM, plus ça change....
It's about time Microsoft seriously thought about re-architecting their operating system from the ground up.
I suspect that this is what they have been doing, and the reason they are late. The main issue of WINE and I suspect of any reasonably clean re-implementation of the Windows API will run into the same issues: emulating bugs and un-documented behavior. Rewriting a clean version is only half of the game, the other is to tell developers to stop using system X call and that behaviour Y will not be supported anymore. Either that, or your code needs to reproduce the undocumented behaviour which is difficult and will result in not so clean code.
Apple did clean up its API by moving from the classic Mac OS Toolbox to the Carbon API which is basically a cleaned up version of the former. The transition took time, and old programs had to run in a emulated version of the previous OS. It is also interesting that Apple chose this approach after the let's rewrite the OS from the group up, plan failed.
Following such path would be, I suspect, quite painful for Microsoft:
The move to OS X brought Mac users a lot of new features (basically many advantages of Unix), the advantages of Vista are not so clear, especially now that many features have been cut out.
The number of Windows applications is much larger, this means more program using strange calls, and more users and programmers complaining that their application does not run on the new OS.
I suspect that games would be particularly affected, support for games has always been an important factor for the acceptance of Microsoft operating systems.
Microsoft has many low-level API like DirectX that Apple did not have to worry about
If Microsoft succeeds and most applications start to rely on a cleaned up API, this API will be much more easy to reverse-engineer by projects like WINE.
I don't think that federalism is a relevant difference. Switzerland is also a federal state with national ID cards. Cards are national but issued by the cantons (equivalent of US states). Those ID cards are used roughly the same way driver licenses are used in the US, with the added advantage that the notion of identity is decoupled from the right to drive or your age. Also an ID card is sufficient to go to neighbouring countries.
I suspect the main difference between Switzerland and the US, beside size, obviously, is social. Switzerland is a settled country, where the government has a pretty good idea where its citizens are: people have ID cards, and are supposed to register in the place they live - most men also have to register with the army. All in all, people don't seem to worried about the government, but then again, Switzerland has a weak executive and direct democracy.
The funny thing for me is, the prospect of national ID cards raises such a ruckus, but nobody talks a lot about the requirements the US imposes on foreigner's passports. First the US required machine readable passports, and now it wants biometric information. Basically, the Swiss government will collect biometric information about me not because it wants to, but because of the US.
I like you idea about having both the funny and more straight-forward text in the same media, in alternate forms. It would indeed be very useful for non native speakers, or kids. In general more meta-data would be useful, like for instance pronunciation for foreign words (like using ruby annotation).
The problem is, it implies changing the relationship with text, there is not a single text that is written and delivered to the reader, but some set of texts. In some sense, this is the same change that is brought with real style-sheets, not one page layout, but some indications on how to lay-out the text. I doubt the average writer is ready for this changed of structure, so I doubt it would actually work: how many news site even use simple meta-tags like acronym?
I'm in a similar situation. I'm a swiss post-doc researcher in Japan. For day to day information (weather, maps, etc) I access of course, japanese web sites. I read blogs and news in three langages (english, french, german), with each langage spanning multiple countries. In general, the web is life line for expatriates. For my study of japanese, I use web sites in Japan but also abroad.
Still the most important thing is for work: I'm accessing web-site all over the world to get papers, either from University web-site or the web-sites of organizations like IEEE or ACM. Was the whole thing not put into place to help academic research? If the web would be really be split along political lines, research would be the first causality. Some of the largest online databases on genes or proteins are not in the US. Same goes for physics: the largest particle accelerator will not be in the US. Many academic projects are international, same goes for open-source projects.
This is because OSS tends to do a huge amount of software reuse. Windows and MacOS [X] software doesn't do that to the same extent.
I'm not convinced that actual software reuse is much more important on OSS software. In theory, this is true: open-source programs can share all libraries. In practice, the picture is more mixed, when you are targeting OS X or Windows, there is a set of libraries that you know will be present, so there is a strong incentive to use those. On the other hand, the OSS world is very fragmented, so the typical OSS installation might contain a lot of libraries with redundant functionality along with a few exotic libraries that are used by one or two programs.
The fact that some code is sitting in/lib is no guarantee that it is effectively reused, I suspect that factors like documentation, stability and age are very important factors.
Slashdot Mirror
The connection machine's processors were distributed among multiple cards, a single card contained 16 processors, the 1 bit processors were implemented using a ROM chip, I think.
As to why we need trading cards and smiley in Unicode, the reason is pretty simple: compatibility. The goal is to be able to convert all existing text data into Unicode, this is why DOS area block drawing are defined as codepoints. Emoji were added to add compatibility to the Japanese systems so that companies like Apple could enter that market with the iPhone, without this, iPhone users would not have been able to exchange messages with other users.
Remember that at one point in time, ASCII was the extended character set with unnecessary symbols like curly braces, this is why C++ compilers still have trigraph support
If prices of ISBN codes were really a problem, people could just publish in France, where ISBNs are free. Anyways nowadays ISBN are just a particular class of GTIN/EAN so I suspect one could just buy an EAN (UPC) code.
If we hit the reset button, can we also fix ASCII? it is by no mean the minimal set most english speakers think it is.
Why do we need a character to represent to 'v' one after the other? You could write 'w' with to 'v' and handle the ligature where it should be handled, at display time. There are so few words in English with the sequence vv that it makes no sense to have the special case coded in the encoding.
Also could we handle the dots on the characters 'i' and 'j' like the diacriticals they are? there should be first the the dotless 'i' and 'j' and the some character to add the dots, like all other diacriticals. Also move out the currency symbols ($ and £), they can be represented as text (USD and GBP), no point in have silly symbols in there. Also remove BELL (11), having a symbol for a bell (2407) might be bloated, but having one for the sound of a bell is absurd.
By the way, why do we need different code points for upper and lower case? They are just variants of each other anyways
Unicode is certainly messy, but plain ASCII is not much better: the most precious 127 code points of utf-8 are basically wasted to display 32 characters and a bit of punctuation, that is pretty bloated for me, we are just used to it
I was a fan of Elite on the C64, I tried Oolite, and it is indeed an excellent port of Elite. I also found it to be no fun at all: requiring time I don't have anymore and lacking the richness I'm used to find in games nowadays. The other game I loved on the C64 was paradroid, there is also an open source port (free Droid), same problem. My expectations and sense of fun have changed in 20 years.
Or turn the problem around: if one provider of telecom services is ever condemned without any other proof than the fact they benefit from a hack, the bad guy just change their business model to extortion.
I saw similar magnetic connectors in household appliances in Japan before Apple used them for laptops: my rice cooker and the water heater both had them.
This is one of the reasons instruments on the LHC are duplicated: Atlas and CMS. Given the difference of energy, if the Higgs is found in one of the LHC experiments, it is doubtful that the experiment could be reproduced at Tevatron.
While the LHC is indeed not an EU project, most of girlintraining's remain valid. Politics and national pride play an important part in the internal workings of CERN, and could well explain the communication policy.
Calling people full of nonsense because they did not get some details right is not very polite, dear Anonymous Coward. Actually, there are quite a few building in Prévessin. And while CERN is definitely not an EU project, it is different from the WHO, or the UN in the sense that it has a geographic definition, it is called the European Organization for Nuclear Research. Some of the funding for CERN related projects, like the grids efforts, comes directly from the EU.
So please try to be less impolite and arrogant...
Actually, for a french speaker, "utilisation" is very easy to understand, as it is a french-word...
That would explain why places like Switzerland or Singapore are so poor...
At any rate there might be technical hurdles, but the core problem is social. Asian people want a system that supports their own text system for addresses because using roman, non-accented characters for the job is really a kludge. Converting kanjis into plain ASCII leads to ambiguous texts or multiple possible transcriptions. For instance the city of Tôkyô can be transcribed as Tokyo, Toukyou, Tohkyoh. Actually, the correct romanization involves the use of the macron accent which is not supported by slashdot's engine (and I don't feel like mucking with escape sequences).
So using their own character system for designating servers is my opinion a valid wish. While some urls might not be typable directly, this simply means a fallback mechanism is needed. At any rate, many url are so long as to be very difficult to type. So either we have a common international system to handle this legitimate need, or every country/company will hack around it, which will be much worse.
When you rip your CD with iTunes, the resulting files are either plain old MP3, or unproctected AAC files.
You are right that decrypted pages are not cached in memory. I'm don't think the issue of caching disk blocks applies, as XNU uses a unified buffer cache. So if the one encrypted page has the same size as a unencrypted page and each page can be encrypted indenpendently, the mechanism is roughtly the same as for normal code pages: the are loaded and evicted as needed, simply the load phase includes a decryption phase.
I fail to see how having a fast disk array gives you any incentive to do additional copies of the code segement on disk...
I don't think that the fact that the pager does only reading means that binaries are in wired memory. Code segments are typically only mapped in read-only mode, so they are never written back to disk, nor should they be.
Just a technically, but Apache is not released under the GPL. Apache is released under the Apache License, which is, as far as I understand, closer to the BSD license than from the GPL license.
Most people might know their music is represented by files, and know that compressed files are smaller. I doubt the majority of people realise that MP3 or JPEG are a lossy compression schemes, or that the degradation increases when the file are decoded and re-encoded multiple times. People don't want to know about those things, they want to listen to the music.
In some sense, I have the feeling we have gone full circle. What this amounts to is having the operating system and core libraries in some sort of pseudo-ROM device, that can be read fast and written a limited number of times. This is not so different from the classical macintosh computer that had a large part of the OS in ROM, the Mac classic had even a bootable disk image in ROM, plus ça change....
Apple did clean up its API by moving from the classic Mac OS Toolbox to the Carbon API which is basically a cleaned up version of the former. The transition took time, and old programs had to run in a emulated version of the previous OS. It is also interesting that Apple chose this approach after the let's rewrite the OS from the group up, plan failed.
Following such path would be, I suspect, quite painful for Microsoft:
I suspect the main difference between Switzerland and the US, beside size, obviously, is social. Switzerland is a settled country, where the government has a pretty good idea where its citizens are: people have ID cards, and are supposed to register in the place they live - most men also have to register with the army. All in all, people don't seem to worried about the government, but then again, Switzerland has a weak executive and direct democracy.
The funny thing for me is, the prospect of national ID cards raises such a ruckus, but nobody talks a lot about the requirements the US imposes on foreigner's passports. First the US required machine readable passports, and now it wants biometric information. Basically, the Swiss government will collect biometric information about me not because it wants to, but because of the US.
The problem is, it implies changing the relationship with text, there is not a single text that is written and delivered to the reader, but some set of texts. In some sense, this is the same change that is brought with real style-sheets, not one page layout, but some indications on how to lay-out the text. I doubt the average writer is ready for this changed of structure, so I doubt it would actually work: how many news site even use simple meta-tags like acronym?
Still the most important thing is for work: I'm accessing web-site all over the world to get papers, either from University web-site or the web-sites of organizations like IEEE or ACM. Was the whole thing not put into place to help academic research? If the web would be really be split along political lines, research would be the first causality. Some of the largest online databases on genes or proteins are not in the US. Same goes for physics: the largest particle accelerator will not be in the US. Many academic projects are international, same goes for open-source projects.
I must say I love slashdot, where an article about UK politics displays with a US flag on top.
The fact that some code is sitting in /lib is no guarantee that it is effectively reused, I suspect that factors like documentation, stability and age are very important factors.