The KJV is in Early Modern English, not Old English. Old English is incomprehensible to modern speakers. u scealt witan over æm e u segest ær u spricst.
He didn't say "roughly 100% of users", he said "roughly 100% of users *I've talked to*" which means that his statistic could be completely correct (of course, only he would know because only he knows who he's talked to and what they've said).
This isn't about ethnicity, at all. It's about linguistic taxonomy. Now, obviously the speakers in various countries just call their language "English". I wouldn't say that I speak "American [English]" but simply "English", as would our friends on the other side of the pond. But for the purposes of linguistic classification, terms like British English (i.e., the English spoken in the British Isles*) and American English (i.e., the English spoken in the US) are quite useful, although they are a bit vague when it comes to serious dialect differences. And by the way, all dialects of English, including those spoken in the U.K. have changed a lot since the language first was spread beyond the Isles. So American English is no more derivative than what is spoken in London. In fact, American English has held on to many words, idioms and types of pronunciation that have been since modified in British English.
I think that people who get their panties in a wad over this should just mentally replace "British English" and "American English" with "English spoken in Britain/UK" and "English spoken in America", which is really all those phrases mean. It has nothing to do with which country is the progenitor of the language, or which country is more "right" about the language, etc. It's just for linguistic classification/clarification.
*Yes, I'm aware of the large dialect differences even within England, so British English is an even more vague term than American English.
You mean British English as opposed to American English, right? English is the term that refers to the West Germanic language the developed on the coast of the North Sea and spread into parts of the British Isles with the Germanic invasions of the first millenium AD, followed by further spread via the British Empire and later, the American "empire". A dialect of English spoken in a given country can be qualified with the name of the country, e.g., American English, Canadian English, British English, etc. if such specificity is required. Although, apparently, it seems, British people have a problem with this rather logical usage. I guess it's fair since we Americans have some strange usage patterns ourselves.
Except that this isn't at all equivalent to that. That would be the equivalent of moving the X server into the kernel, not just some directly hardware facing parts of the drivers.
I actually chose those examples because they also commonly have non-native plurals, but the point is taken. There are many others where the Latin plural is rare and/or pedantic-sounding ("formulae" is a good example, in fact) and a great deal more where it simply is not used (the plural of "creation" is not "creationes", but the native "creations" without exception).
The differences are both dialectical and based on register. Educated or written registers will prefer the Latin plurals, colloquial and other "non-standard" registers and dialects will generally not. The reality is, written standard English really is a different language from the varieties of spoken English and it's rate of change is held in check by generally respected standards, whereas there are no external standards for spoken English (despite the attempts of English teachers and parents of teenagers).
Yet another display of ignorance. It doesn't have anything to do with laziness or whatever stupid reason you have come up with for why we don't keep non-native plurals. The reason is, quite simply, that they aren't native plurals! We don't speak Latin, we speak English. We use English grammar, which has plurals in -s, not -a, -er or anything else (aside from some irregulars). The sole reason that we use the Latin plurals for *some* (but by no means all, or even a majority, of loanwords from Latin) is that Latin was held in such esteem during parts of the history of the language, that there was a tendency to force a non-native plural, which is treated as an irregular plural at best, and at worst, a separate word, leading to strange quadruplets like medium/mediums/media/medias. The natural order is not to use the non-native plural (and this is not just an English thing: even Latin itself generally used Latin inflections for Greek and other non-Latin loanwords*). Why is English singled out as being so bad here?
I am aware that in the Scandinavian languages, "kraken" is a singular with an attached article. And that's fine...in Scandinavian. But in English, that has no bearing because English does not operate that way. And that is the entire point of my post. It is not a flame against Scandinavian linguistics, or the history of the beast, but about the implication that English has somehow screwed it up, and furthermore, that English must respect Scandinavian noun inflections. That is the irrelevant pedantry which I was referring to.
* For some Greek loanwords, Latin actually employed the Greek declensions, but this was done rather haphazardly and inconsistently. Yes, the Romans even had this problem. My Latin grammar (Gildersleeve) says this about Greek nouns in Latin:
"Greek substantives, especially proper names, are commonly Latinised, and declined regularly according to their stem-characteristic. Many substantives, however, either retain their Greek form exclusively, or have the Greek and Latin forms side by side. These variations occur principally in the Singular, in the Plural, the declension is usually regular." (section 65)
We have borrowed words or phrases into this language that include non-native morphology. With the exception of some Latin and Greek plurals, we generally ignore the non-native morphology and use our own endings. So, for example, not only is it acceptable, but it is required to say "the La Nina" or "those La Ninas" and not "*La Nina" or "*the Nina" or "*those Las Ninas" or "*Estas Ninas". The lexical item is "La Nina", which cannot be decomposed into smaller morphemes like it can in Spanish.
Another example, also Spanish-related, is the presence of a number of words of Arabic origin in Spanish that begin with "al-", such as "algodon". Originally, this was the definite article in Arabic, but it is now a meaningless part of the word in Spanish and does not prevent the use of the native definite article.
Or, going back to English, it is generally correct to use a native s-plural for words of Latin origin, except in a small set of common loanwords: "formulas", "nexuses", "moratoriums", etc. Again, this is okay because English isn't Latin and isn't required to use Latin morphology. The fact that it does at all is a more a testament to the high standing Latin had and still has in our culture. Those non-native plurals are actually affectations, rather than the rule. You don't see people generally trying to use non-native plurals with words from other languages (the less important the language, the less likely we are to use anything other than the native s-plural).
So, my point is, it doesn't really matter what the original morphology was in the language we borrowed from. We borrowed the word as "kraken" and it is not decomposable into any smaller morphemes. The correct *English* plural is "krakens" and not "kraker", "kraks" or "krakulations".
I majored in linguistics in college, so I am aware of the way the sounds work. I once again apologize for not immediately breaking out IPA and discussing the finer points of vowel articulation.
I didn't mean the diphthong/ai/ notated in English, I meant the normal sound of that letter in other languages. I thought the fact that we already know what the name sounds like would make that clear, but I guess not.
You clearly do not know anything about Cyrillic. The 'H' you refer to is actually an 'n' and the backwards 'N' is actually a 'i' type sound. The 'C' is an 's'.
With good libraries or a language runtime, you shouldn't have to make any system calls. And the few times that you do, it can be wrapped in a thin abstraction layer within your program.
There are extended attributes and magic numbers. Some file managers seem to make limited use of these. I guess the problem with extended attributes is that they aren't guaranteed to be transferable along with the file when you, e.g., upload the file to a website, or transfer it to another computer. Only the main stream of a file is guaranteed to be transfered.
But make no mistake, the filesystem fully supports this kind of information. We just don't seem to make much use of it right now.
They aren't selling anything. They are just coming up with ideas to explain things many light years away (i.e., not particularly relevant to business practices).
There's no value in doing any of that for a graphics server. The only reason it's useful for web services is because web applications start at the server and interface with the client. Thus, a mechanism was needed to make it easier for more work to be pushed onto the client to avoid lots of roundtrips. Well, what do you know, that's already what's happened with X. Most rendering work is done client-side. Pushing it back onto the server would make a lot less sense and would result in a great deal more complexity and lowered performance. The server needs to be doing even less work than it already does not. The end-goal is to have the server basically multiplex windows and events and leave most of the rest to the clients, to the DRM/DRI or to a compositing manager. It shouldn't be executing arbitrary code, doing complex rendering, or managing video hardware.
The slowness is generally bad drivers and that's mostly the fault of the vendors, not X. For those of us with good drivers, performance is excellent, nearly on par with Windows.
As far as legacy API calls are concerned: leave them there. They don't hurt anything. No new code is optimized around them. Yes, it may bloat the server a little bit to provide the legacy emulation, but lets face it, any large infrastructure project like X, or an OS, or a language runtime, has to deal with legacy. That's a reality that Windows and OS X and every other system out there deal with. Get over it.
Network transparency is not a problem for several reasons. One, performance is not affected by network transparency. Communication is done through Unix sockets and shared memory if the app is on a local machine. This is pretty much the same as it is on OS X (IPC to WindowServer) and Windows, sort of. For 3d stuff, there's DRI which lets apps talk directly to the video hardware, which is marshaled by the DRM. Secondly, network transparency is really just a side effect of having a good design for communicating between apps and the server. So it doesn't really add any complexity to the system for most cases: you just connect to a server and then start calling Xlib API calls, or whatever. The security model IS complex, or was, but it's now being replaced with a simpler one and most people just use SSH anyways.
As far as the API is concerned, yes, it has some cruft in it. But guess what? It already has a working replacement: XCB. XCB implements the bare-bones protocol and can reduce latency significantly. All we need is for the toolkits to be ported to XCB, but nobody has done much work in this area (last I saw was a post to the Xorg mailing list a year or two ago where someone was trying to figure out what it would take to port GTK+ to XCB).
Every time you make a system call, you switch rings. Every time there's an interrupt, you switch rings. Having mode setting either in userspace or kernelspace is orthogonal to the issue of ring switching. It has everything to do with having the mode setting code in one central place (and the right place: the kernel should be dealing with hardware, not some userland program) so that X, the kernel and whoever else can access the video card in a uniform, secure and stable fashion. The reduction of flicker is just one of the many benefits of KMS.
That's almost exactly how it already is right now. This whole thread is a bit strange to me since everyone's complaints about X are based mostly on their misunderstanding of how it actually works.
Yes, I'm aware of the theming issues. I use the same theme on Ubuntu and Gentoo (Clearlooks "Classic"). On Ubuntu, it definitely feels slower. I didn't run GtkPerf on my own Ubuntu to get a true comparison yet, but I'd be surprised if the results were close to what I get on Gentoo.
Profiling would be worth it if I really cared about using Ubuntu. So far, for me at least, it offers almost no benefits, and a number of drawbacks, among them performance problems. I'll stick with Gentoo for the time being.
And no, I'm not a ricer. I just happen to find that Gentoo fights against me less than some other distros and that's why I keep using it. My CFLAGS are quite sane.
Slashdot Mirror
The KJV is in Early Modern English, not Old English. Old English is incomprehensible to modern speakers. u scealt witan over æm e u segest ær u spricst.
I think #1 humping and data freaking him out are two separate thoughts.
He didn't say "roughly 100% of users", he said "roughly 100% of users *I've talked to*" which means that his statistic could be completely correct (of course, only he would know because only he knows who he's talked to and what they've said).
This isn't about ethnicity, at all. It's about linguistic taxonomy. Now, obviously the speakers in various countries just call their language "English". I wouldn't say that I speak "American [English]" but simply "English", as would our friends on the other side of the pond. But for the purposes of linguistic classification, terms like British English (i.e., the English spoken in the British Isles*) and American English (i.e., the English spoken in the US) are quite useful, although they are a bit vague when it comes to serious dialect differences. And by the way, all dialects of English, including those spoken in the U.K. have changed a lot since the language first was spread beyond the Isles. So American English is no more derivative than what is spoken in London. In fact, American English has held on to many words, idioms and types of pronunciation that have been since modified in British English.
I think that people who get their panties in a wad over this should just mentally replace "British English" and "American English" with "English spoken in Britain/UK" and "English spoken in America", which is really all those phrases mean. It has nothing to do with which country is the progenitor of the language, or which country is more "right" about the language, etc. It's just for linguistic classification/clarification.
*Yes, I'm aware of the large dialect differences even within England, so British English is an even more vague term than American English.
You mean British English as opposed to American English, right? English is the term that refers to the West Germanic language the developed on the coast of the North Sea and spread into parts of the British Isles with the Germanic invasions of the first millenium AD, followed by further spread via the British Empire and later, the American "empire". A dialect of English spoken in a given country can be qualified with the name of the country, e.g., American English, Canadian English, British English, etc. if such specificity is required. Although, apparently, it seems, British people have a problem with this rather logical usage. I guess it's fair since we Americans have some strange usage patterns ourselves.
There are stable branches: older kernel releases. They keep getting bugfixes and security fixes for some time.
Except that this isn't at all equivalent to that. That would be the equivalent of moving the X server into the kernel, not just some directly hardware facing parts of the drivers.
I actually chose those examples because they also commonly have non-native plurals, but the point is taken. There are many others where the Latin plural is rare and/or pedantic-sounding ("formulae" is a good example, in fact) and a great deal more where it simply is not used (the plural of "creation" is not "creationes", but the native "creations" without exception). The differences are both dialectical and based on register. Educated or written registers will prefer the Latin plurals, colloquial and other "non-standard" registers and dialects will generally not. The reality is, written standard English really is a different language from the varieties of spoken English and it's rate of change is held in check by generally respected standards, whereas there are no external standards for spoken English (despite the attempts of English teachers and parents of teenagers).
Yet another display of ignorance. It doesn't have anything to do with laziness or whatever stupid reason you have come up with for why we don't keep non-native plurals. The reason is, quite simply, that they aren't native plurals! We don't speak Latin, we speak English. We use English grammar, which has plurals in -s, not -a, -er or anything else (aside from some irregulars). The sole reason that we use the Latin plurals for *some* (but by no means all, or even a majority, of loanwords from Latin) is that Latin was held in such esteem during parts of the history of the language, that there was a tendency to force a non-native plural, which is treated as an irregular plural at best, and at worst, a separate word, leading to strange quadruplets like medium/mediums/media/medias. The natural order is not to use the non-native plural (and this is not just an English thing: even Latin itself generally used Latin inflections for Greek and other non-Latin loanwords*). Why is English singled out as being so bad here?
I am aware that in the Scandinavian languages, "kraken" is a singular with an attached article. And that's fine...in Scandinavian. But in English, that has no bearing because English does not operate that way. And that is the entire point of my post. It is not a flame against Scandinavian linguistics, or the history of the beast, but about the implication that English has somehow screwed it up, and furthermore, that English must respect Scandinavian noun inflections. That is the irrelevant pedantry which I was referring to.
* For some Greek loanwords, Latin actually employed the Greek declensions, but this was done rather haphazardly and inconsistently. Yes, the Romans even had this problem. My Latin grammar (Gildersleeve) says this about Greek nouns in Latin: "Greek substantives, especially proper names, are commonly Latinised, and declined regularly according to their stem-characteristic. Many substantives, however, either retain their Greek form exclusively, or have the Greek and Latin forms side by side. These variations occur principally in the Singular, in the Plural, the declension is usually regular." (section 65)
-1, irrelevant pedantry
We have borrowed words or phrases into this language that include non-native morphology. With the exception of some Latin and Greek plurals, we generally ignore the non-native morphology and use our own endings. So, for example, not only is it acceptable, but it is required to say "the La Nina" or "those La Ninas" and not "*La Nina" or "*the Nina" or "*those Las Ninas" or "*Estas Ninas". The lexical item is "La Nina", which cannot be decomposed into smaller morphemes like it can in Spanish.
Another example, also Spanish-related, is the presence of a number of words of Arabic origin in Spanish that begin with "al-", such as "algodon". Originally, this was the definite article in Arabic, but it is now a meaningless part of the word in Spanish and does not prevent the use of the native definite article.
Or, going back to English, it is generally correct to use a native s-plural for words of Latin origin, except in a small set of common loanwords: "formulas", "nexuses", "moratoriums", etc. Again, this is okay because English isn't Latin and isn't required to use Latin morphology. The fact that it does at all is a more a testament to the high standing Latin had and still has in our culture. Those non-native plurals are actually affectations, rather than the rule. You don't see people generally trying to use non-native plurals with words from other languages (the less important the language, the less likely we are to use anything other than the native s-plural).
So, my point is, it doesn't really matter what the original morphology was in the language we borrowed from. We borrowed the word as "kraken" and it is not decomposable into any smaller morphemes. The correct *English* plural is "krakens" and not "kraker", "kraks" or "krakulations".
I majored in linguistics in college, so I am aware of the way the sounds work. I once again apologize for not immediately breaking out IPA and discussing the finer points of vowel articulation.
I didn't mean the diphthong /ai/ notated in English, I meant the normal sound of that letter in other languages. I thought the fact that we already know what the name sounds like would make that clear, but I guess not.
You clearly do not know anything about Cyrillic. The 'H' you refer to is actually an 'n' and the backwards 'N' is actually a 'i' type sound. The 'C' is an 's'.
Idiot. It's a possessive, where the apostrophe is required.
With good libraries or a language runtime, you shouldn't have to make any system calls. And the few times that you do, it can be wrapped in a thin abstraction layer within your program.
How about they not immigrate illegally and then steal copper? Seriously, WTF.
There are extended attributes and magic numbers. Some file managers seem to make limited use of these. I guess the problem with extended attributes is that they aren't guaranteed to be transferable along with the file when you, e.g., upload the file to a website, or transfer it to another computer. Only the main stream of a file is guaranteed to be transfered.
But make no mistake, the filesystem fully supports this kind of information. We just don't seem to make much use of it right now.
They aren't selling anything. They are just coming up with ideas to explain things many light years away (i.e., not particularly relevant to business practices).
There's no value in doing any of that for a graphics server. The only reason it's useful for web services is because web applications start at the server and interface with the client. Thus, a mechanism was needed to make it easier for more work to be pushed onto the client to avoid lots of roundtrips. Well, what do you know, that's already what's happened with X. Most rendering work is done client-side. Pushing it back onto the server would make a lot less sense and would result in a great deal more complexity and lowered performance. The server needs to be doing even less work than it already does not. The end-goal is to have the server basically multiplex windows and events and leave most of the rest to the clients, to the DRM/DRI or to a compositing manager. It shouldn't be executing arbitrary code, doing complex rendering, or managing video hardware.
That's way outdated. Modern X.org has fixed most of those problems.
The slowness is generally bad drivers and that's mostly the fault of the vendors, not X. For those of us with good drivers, performance is excellent, nearly on par with Windows.
As far as legacy API calls are concerned: leave them there. They don't hurt anything. No new code is optimized around them. Yes, it may bloat the server a little bit to provide the legacy emulation, but lets face it, any large infrastructure project like X, or an OS, or a language runtime, has to deal with legacy. That's a reality that Windows and OS X and every other system out there deal with. Get over it.
Network transparency is not a problem for several reasons. One, performance is not affected by network transparency. Communication is done through Unix sockets and shared memory if the app is on a local machine. This is pretty much the same as it is on OS X (IPC to WindowServer) and Windows, sort of. For 3d stuff, there's DRI which lets apps talk directly to the video hardware, which is marshaled by the DRM. Secondly, network transparency is really just a side effect of having a good design for communicating between apps and the server. So it doesn't really add any complexity to the system for most cases: you just connect to a server and then start calling Xlib API calls, or whatever. The security model IS complex, or was, but it's now being replaced with a simpler one and most people just use SSH anyways.
As far as the API is concerned, yes, it has some cruft in it. But guess what? It already has a working replacement: XCB. XCB implements the bare-bones protocol and can reduce latency significantly. All we need is for the toolkits to be ported to XCB, but nobody has done much work in this area (last I saw was a post to the Xorg mailing list a year or two ago where someone was trying to figure out what it would take to port GTK+ to XCB).
Every time you make a system call, you switch rings. Every time there's an interrupt, you switch rings. Having mode setting either in userspace or kernelspace is orthogonal to the issue of ring switching. It has everything to do with having the mode setting code in one central place (and the right place: the kernel should be dealing with hardware, not some userland program) so that X, the kernel and whoever else can access the video card in a uniform, secure and stable fashion. The reduction of flicker is just one of the many benefits of KMS.
That's almost exactly how it already is right now. This whole thread is a bit strange to me since everyone's complaints about X are based mostly on their misunderstanding of how it actually works.
Yes, I'm aware of the theming issues. I use the same theme on Ubuntu and Gentoo (Clearlooks "Classic"). On Ubuntu, it definitely feels slower. I didn't run GtkPerf on my own Ubuntu to get a true comparison yet, but I'd be surprised if the results were close to what I get on Gentoo.
Profiling would be worth it if I really cared about using Ubuntu. So far, for me at least, it offers almost no benefits, and a number of drawbacks, among them performance problems. I'll stick with Gentoo for the time being.
And no, I'm not a ricer. I just happen to find that Gentoo fights against me less than some other distros and that's why I keep using it. My CFLAGS are quite sane.