The French have had a history in blocking American Air-waves; I see no reason why the internet should be any different.. If the French Government doesn't like what America (or any other country) has to say, then they can simply close their ears as they've done for decades.. And if the people don't like this sort of "law and order".. Well, they can revolt (civilly of course:)
Seriously. As a yongster, I was appauled that the KKK could hold a march in DC. But as I've gotten older, I understand better.. It is not right to supress opinions (so long as those opinions do not migrate into harmful actions). Those that adore the Nazi regieme are entitled to their opinions. It is this very sort of "Treaty of Vers..(sp??)" punishment and attempts at covering up ugly parts of the world or history that has exemplified trouble.
This is obviously a trivial (albeit sensationalized) case of sensorship (who's to say that Nazi imagry is less offensive or more dangerous than child-porn.. Personally, I don't understand the puritanism of anti-pornography to begin with.. It's just sex.. But that's just my opinion).
Lastly, if France really wants national sensorship then have THEM put firewalls/Filters at their borders! Sensor at the client, not at the server.
If you have a 24/7 internet connection, then it's a fixed cost, and you can't factor in the "marginal cost" of the download of a distribution. This is especially true if you download it from work, where the fixed cost is paid by someone else.
Heck, I'm still at a University, where I don't even directly pay for the internet connection.. The fixed cost is bundled in with thousands of other costs. Being that I get a better part of a tax rebate on tuition costs, and my grants are never taxed, it's entirely possible to completely avoid taxation of free software alltogether.
This is obviously an extreme, and some people are plagued with pay-per-hour usage.. But even if you get a 50hr / month fixed rate, you can't economically consider that to be taxation of your individual online downloads.
I'm still waiting for the following scenereo: fiber to the home, but you pay for the BW at the switching station.. Charge people for BW just as you charge them for their electric bill. Most Americans don't consider their electric expense in their every-day lives, and I think the same would be true if we had pay-per-usage of the internet, so long as BW / cost is marginal. With this, governments could tax you just like they do your normal utilities, but this would still be such a marginally small amount, that it would be on par with electricity; once again.. The main advantage is that, just like electricity, you'd have a virtually unlimited potential, but you wouldn't want to abuse that potential often because of budgeting concerns.
MB's can be very expensive.. Especially when they're in low availability. When the Athalon came out, the MB's were almost as expensive as the chips.. Throw in the fact that they were buggy (since it was all brand new.. including the EV6 connectors). The P4 requires an all new powering structure, so it's possible that we're going to find MB's that blow up the CPU's.. Thus you might want to wait a few rounds before considering these guys (this is usually a good rule of thumb anyway).
Given that it's brand new, you're also going to be at the mercy of the MB manufacturers in terms of features.. If you like ultra-SCSI-RAID on your MB, it might be a while. There is a similar argument to be made for your case.
Expect to pay premium dollars for this combo.
The problem with the head-sink is logistics and ergonomics.. Where do you put everything? Unless you have a monster tower case, you're not going to be able to fit too many full lenght boards.. You have a massively over-heated PAIR of RDRAM RIMMS, a MASSIVE CPU, a larger than normal power-supply.. And then you get to have your first AGP card. A Mid-Tower is probably too small for these guys.. That might not be a problem for some.. But personally, I like stacking my computers, so it's a problem for people like me.
The heat-issue also means real-estate for lots of cooling.. And an increased risk of heart-failure.. So to speak.
My guess is that you're looking at $300 - $400 for the MB + case alone. That's more than I usually spend on an entire bare-bones system. Throw in roughly $600 for each of CPU and memory, and you've got a rather large handi-cap for an almost miniscule performance gain.
As for the pricing of their CPUs... To be fair, they _have_ to depretiate their costs.. It's not cheap to design and entire CPU from scratch.. Remember, that this is their FIRST CPU redisign since the pentium Pro some 5 years ago!! Everything else (with the exception of the vapor-ware Italium) has been an add-on to their old archetecture.
I don't know what their margins are per CPU - I wouldn't be surprised if it was 100% - But when you take in the cost of multiple billions of dollars, it's not all money in the bank. AT&T use to depretiate their hardware-costs over 40 years.. Intel doesn't have such a luxury.
I don't know if I agree with this.. These benchmarks show that Intel's 1.5 is roughly equivalent to AMD's 1.2, and we were told to expect this by the previewers.
I believe AMD uses a fully pipelined FPU (multiple ones at that). I'm sure that the P4 also fully pipelined their FPU, BUT, they added several additional stages to the basic instruction as well.
The addition of stages does two things: first it increases your max clock rate all-else-held-equal. And second it increases the latency for missed branch predictions. Again, all-things-being-equal, the missed branchs will tend to hurt more than the higher clock helps, except in a few special cases.
Intel, therefore included with those extra stages a highly advanced branch predictor that lives in the MIDDLE of the pipe. More-so, successfully predicted branches can skip the first several stages thanks to the branch cache. Thus, well-behaved code will get the benifit of heavier pipelining with fewer of the pit-falls. To make things even more tantelizing, they're using a 2x clocked Integer Unit. Thus they have a 3GHZ integer unit on these benchmarks. That means that they can further extend their pipelines with almost no visible penalty (even in branch-misses).
Unfortunately, they still seem to be plagued with branch misses (the only logical explanation for why AMD can still keep up or even surpase them). Obviously the memory played an important role in these benchmarks.. The KT133 v.s. AMD560 really only differ in memory speed, and that was enough to sway several percentage points. A more fair comparison would be between VIA's up and comming DDR-SDRAM P4 chipset.
But, as was pointed out; if Intel can get the P4 up to 2GHZ before AMD can (last rumor I heard was that AMD was going to hit.13u before Intel), then they can start pulling away.
Unfortunately, as several sites pointed out, buyers don't look at benchmarks, they look at CPU speed, so Intel should be able to wrongfully win people over on this synthetic basis. Thankfully, the only people that are going to be willing to buy P4's are people needing servers (or maybe even Q3). We'd have to see NT ASP/Sql Server and or Linux+Apache+PHP+Oracle, etc to determine who's king (including memory types). Unfortunately I rarely see benchmarks on these grounds.
Sooner or later AMD is going to come out with their 64bit proc. With Mustang gone, this is their only next-great-hope. An all new design - hopefully without a tremendous cost - that has started from scratch (as the P4 did). I'm sure it too will have a heavy pipeline, but several of it's new features (such as the flat-memory archetecture) should enhance the playing field.
By then, however, the P4 will have found a new chipset that handles DDR-SDRAM and will have enough volume MB's and cases that it'll be cheap enough for the hard core gamer and possibly even casual gamer to purchase. A 2 - 3GHZ processor running at.13u is going to be hard to beat. There are no benchmarks for AMD's sledgehammer, so there's no point in speculating about it.
I totally agree with another poster that said we should be rooting for both Intel AND AMD since competition is good.
I do not believe any "ThunderBirds" ran at 600MhZ.. Think they started at 700MHZ... Since you didn't say Duron, I assume that you have the good ole fashioned Athalon.. Which means that you have a slotted MB, and thus can not make use of the 1.2GHZ socked-A processor.
But you can get a new MB for $120, so it shouldn't be that big of a deal.. Except that you'll probably want to get PC133Memory (since you most likely had PC100 on that older board).
Really! I've been playing the latest 3d games with my matrox card for quite some time. Perhaps it's all been a hallucination!:(
I said keep up, not be able to produce a product. The 3D market is very volitile; one year, you make the fastest / cheapest card and you're topping the sales charts, then the next, you're merging with another company because you can't hold your own.
Matrox is in no danger of going under, 3D isn't their primary market. I respect Matrox since I've always thought of them as a quality company.
I don't forsee them even leaving the 3D market, but if you're a gamer, the Matrox card isn't the best buy for your money. I've never seen any review that has suggested otherwise.. The main things that might persuade a hard-core-gamer to choose the Matrox card is their Dual monitor or large monitor support.
That being said, any given business should probably opt for Matrox cards (due to quality, features and drivers for things other than games), so if you game-play at work, then Matrox is an obvious choice. But most of us don't get to choose the graphics cards at work (well, I do, but I bring my own machines to work).
On another subject, did you drop all the users from blacknova.net? My user (tolket) is gone....
Laugh, I'm not in charge of that, though I'm one of the OpenSource developers.. Yes, the DB was completely flushed... You need to create a new character.. Read the forums.
3Dfx burned a lot of bridges when they left the OEM market.. I'm sure they'll get back in, but they've lost much of their clout.. They're going to have to make a top notch card for people to promote 3Dfx over nVida.
Additionally, I don't think _anyone_ wants to go back to the add-on-card market.. With resolutions reaching 1600x1600, those pass-through cards are simply not acceptible. Back in the days of 1024x768 at 16bit color (in SLI mode), you could always put two monitors on your computer (since you obviously had money to burn).
Additionally, the only reason people went add-on card was because there was no all-in-one solution that could even compete.
It would be an extremely hard sell to say the least.. Though you might find a market back in the 200MHZ legacy computer market (with a full blown GPU).
Another SERIOUS issue is AGP.. Can't do SLI with AGP (unless you're Obsidian and obsessed), and you definately can't be an add on card and use AGP.
I realize you only meant this as a "worst case", but I wanted to debunk it as being profitable for them.
Matrox has never been able to keep up in the 3D market.. They've primarily been a professional 2D card company. I don't think they've ever topped the 3D charts (though you can put 1600x1600 monitors on them without a sweat).
They were innovative with their environmental bump mapping, and for a while there, they'd produce the cleanest pictures, but apparently this isnt' the case with their latest cards, and ATI (at least) has already caught up with them in the quality realm.
Competition is GOOD.. Winners that gain a monopoly in the 3D video card industry is BAAD.
ATI and nVida should be good competition with each other for a while. Both making state-of-the art cards (at almost the complete price range), but once one of them goes... That's about it..
The time to enter the video card market ended with nVida, who stepped up the competition until the then king couldn't compete. With the world consolidating towards DirectX and, on occasion, OpenGL, you can't depend on your proprietary drivers (a la Glide, and many other's that I've already forgotten). Most likely, the only way you can make a faster, more cost effective card is through propriety drivers (such as the failed infinite plains card (forgot the name of that too)).
Couple that with the fact that GPU's are now as (if not more) complex as CPU's, in addition to using state-of-the-art manufacturing, it's highly unlikely that a new contender can possibly dethrown the existing big boys.
In fact, the only thing I can image you could add is move the whole damn game into the GPU (a la MicroSoft's DirectGame or Direct3DShooter... ).
We're an offshoot of Dupont, and we've had flex-time for the longest time.. What we have are "core" hours from 9am till 3pm. You get to choose your 8hour day outside of that.
The advantage of shift-time's are that people can know when they can find you. Various activities such as network administration can be "paged" when needed, and coders are sent on week-long excursions which could even run into the weekend if necessary, so the exact hours that they work aren't always important. Secretary positions, however, tend to be more important and lend to shift-work (as would assembly line work).
For us, various benifits like a kitchen, flex hours, free soda, etc were in lue of higher pay, and for me, that was just fine. If I had to choose between higher pay and a less stressful working environment.. I'm all for living a higher quality life.
First, I started off with the words "Sadly, I don't know".. Therefore removing myself as an expert in that area. It was more a question than anything else.
As for the other languges, perl, tcl, etc, those can only be less optimial forms, and are more for procedural operations, which has nothing to do with over-comming the 16 join problem. (With the exception maybe of hashed-lookups, but that's dangerous)
It's great to hear about cached query plans (I'd forgotten the term).
As for the problem with speculation.. If nobody were to speculate, and those in the know don't impart their knowledge, then has anyone gained anything? 'Qualified' speculation does exactly as here; Inciting those knowing specific facts to quality errors. My speculation _was_ based on emperical analysis, however, so it wasn't as if I was blindly stateing something.
But now, as a question, is the compiled sql stored in any more optimal fashion? Or does postgres use an interpreter (which does little more than tokenize). To me, a compiled sql statement means that its making raw calls to table operations..
Simple benchmarks that I've done haven't shown any performance increases with using stored procedures, but that might just be because the compilation / interpretation stages were insignificant with respect to the overall operation. The other interpretation is that little is actually gained by the compilation of the sql statement (at least until the cached query plans come around).
As a followup idea.. 16 joins reminds me of a project that I did once where we used ID's for every table.. Thus every table stored meta data and a single piece of actual data.. Thus if you had 17 pieces of data to display, then you'd have to do 16 joins. There were three solutions to this:
First, cache all the smaller tables, and do the translation (from say user_id to user_name) in an external hash (I know perl can easily do this; don't know about php). For static fields such as user-names, email addresses, etc, this is easy. Unfortunately this doesn't help when trying to find all users with email addresses going to aol and souce code that's over-due, etc. But it does solve the summary report problem.
Second is to break up the query into multiple operations.. You need an incredible amount of memory to do certain types of joins, and sometimes the sql-processor guesses the wrong method. Even though postgres's "explain" command told me it was doing hash-lookups for a user-name summary, I found that it was faster to first get all the data with user-id's, then perform a second search with
"select user_name from users where user_id in (x,y,z,....)"
So long as you don't have more than a couple dozen entries, this works nicely.. ESPECIALLY if you have thousands of rows of real-data with only one or two unique user-names. This was a nice-short term fix, but "clean" sql should all be done on a single statement.. Soo onto the third
Denormalize back to form 4 or 3. Namely, whenever you have a user-name, don't use a user-id.. Sure you consume 8 more bytes of memory per row, but it greatly speeds up searches. The method is to ONLY use ID's when you can not uniquely represent a primary key with a single column.. (such as first-name, last-name,...).. But in our computer world, "user-name" tends to be unique. "project-name" should be unique, "working-branch" should be unique.. In which case, at the design level, it's worth the extra consumed space.. Now I know that Mysql advocates (including me a year ago), cringe at the idea of keying off variable length strings.. For me, it was the idea that you can compare integers much faster than strings.. BUT, if the string is an index, then your sorting is going to be MUCH faster, strings distribute much better than auto-incremented integers (which tend to want to be linear). Additionally, the fact that you won't have to perform a join at all removes almost all of the performance problems.
When you don't use meta-data as primary keys, then you find that you only perform joins when you actually have relational data... That joins based on summary data almost always goes away.
Additionally, if your DB ever get's corrupt (say you lose a table), then you don't have to fret over "who the hell was user-id 155?". Given that Mysql doesn't have roll-backs, the potential for data corruption with meta-data is enormous (when you do both reads/writes).
Essentially what that test showed was that mysql had to lock all those tables in order to do the join.. And that with Postgres's versioning system, it was a trivial matter.
I remember reading on the topic of excessive joins in a Sybase Admin book once (excelent book on "practical" database, since it shows you exactly how one company solved all the DB problems). It essentially said that determining join order was actually a very slow process.. Exponential in fact.. Especially if your DB tries to determine which indexes to used based on statistics... The selection of indexes increases the number of possible variations incredibly. Sybas's solution was to only compare a select few of the possible join methods - which is a trade-off.. Alternatively, if you were to create a stored procedure, then which-ever method was best at creation time will be used from there-on-out...
Stored procedures - in Sybase at least - reduce a great deal of the overhead, especially in a 16-way join. Sadly, I don't know if Postgres does ANY pre-compiling of stored procedures. They have two main methods: straight named 'sql' statements, and 'psql' (a true stored procedure 'language'). My guess is that they do little more than pre-parse the statment (if even that). If this is the case, then using a stored procedure might even be slower than raw SQL since a call requires the parsing of the function name, then loading of the function into memory from a table, then the parsing of that statement (two more steps than raw processing). It is, however, necessary for triggers and rules.
Postgres (and I believe mysql) has raw-c extensions, but I really dislike these for general DBA operations (not to mention I haven't spent the time learning them).
As for the 16-table join, a general web page probably shouldn't do this (especially a heavily used page), but a database SHOULD at least let the DBA perform the ad-hoc query when trying to extract information. The only real solution is summary information, which often tends to be more trouble then their worth... Another solution is de-normalization (which makes me feel dirty)... Basically like what Blacknova.net uses... If you have 1 to N data, then just pick some upper bound N and include that many extra columns directly in your main table. Fast and efficient, provided that you choose the right N.
Mysql uses multi-threading (each connection gets it's own thread), but I've read slashdot posted benchmarks that show Mysql failing after so many threads are spawned via concurrent connections(I think 60). Perhaps they're using some finite resources and the thread spawner was not using semephores or some other such resource control.
Postgres uses multi-processing (each connection gets it's own process) with shared memory segments and a controlling/monitoring process (postmaster).
Thus if Mysql could do a better job of managing resources under heavy loads, then it would waste less physical memory. Additionally, I don't know how good Linux is with multi-threading under heavy loads. Can it faithfullly handle hundreds of potentially thousands of concurrent/competing threads? I've never seen any tests to show this... I'm aware that the default max processes/threads was hard-coded in the kernel to 512 at one point (I believe it requires a recompile to change). This might have something to do with mysql's crashing.. You'd have to go to single-user mode to max this puppy out. Additionally I don't know if it's possible for an app to determine how many more threads can be spawned by the OS.. If a thread fails to spawn, then you can corrupt the entire process.. In MP, if you fail to spawn a process, then you simply lose that connection.
In general, I'd prefer MP on a Linux system. It's a similar story with Apache. You get a lot of reliability, and not _that_ much memory waste when compiled correctly. Additionally, MP has the potential to span computers (as in a slasdot article on a multi-hosted extension to Linux)
the method of dividing the power is completely incapable of increasing it. This is quite obvious.
I'm sorry, I don't understand what you're saying.. It isn't obvious to me at least. Perhaps instead of subjectively berating the system with unsophisticated language, you could provide evidence (you know.. being scientific and all).
The idea is that you want to reduce the power of the majority.. Geographically and ideologically. Power is a bi-stable system.. The minority has almost no power until just about the 50/50, then they suddenly are empowered almost instantaneously. After that, they grow exponentially until no-one else even has a say. This is not a fair system, and this is what the Federalist papers tried to prevent.. Mob rule.
NY and California are peeked.. Their voters will not achieve any more power, but that's fine.. The two of them alone carry a better part of the power required. NY is it's own little idiology (it's own melting pot), but California is rather liberal, while Texas is rather concervative. Should CA and NY alone determine the president, ignoring the Geographic differences in idiology? Granted, Texas has enough power to offset this, but then you have the remaining southern states, who population wise could not affect the presidential outcome much, but electorially they can.
Another issue of dividing power "increasing" it, is with his example of gerimandering(sp?). Ignoring the president for a moment, if you put all blacks in one district and gave them a black representative, then that senator would only have 1/5xxth of a voice. If, however, you spread those people out among several districts, then you could influence 5-30% of the vote for multiple districts (including sympathists), thereby having dozens of representatives in congress, a significantly greater margin.
Ahh.. I don't care to argue anymore.. If Hillary has her way, it'll be abolished anyway, and the math deficient among us will simply look at this election as all the proof they need.
HA HA HA HA!!! So that is why we haven't had a White president in the last 50 years!!!
Well, as confusing as this statement is, I'll assume that you're simply refuting my statement and saying that the electorial college has not prevented racists / radicals, and additionally has not allowed minorities to lead.
Well, let's use some emperical evidence.. IF the southern states had the majority of people in the country (thankfully they don't), and we didn't have an electorial college, then a former KKK person could actually be elected president. the north east corridor, however, would garuntee that such a thing could not happen (because they have a lot of people in NY and a lot of states overall). If you had a black radical who's primary goal was reperations for the mal treatment of blacks, then you'd have the entire south blocking them (even if there were 51% people represented by the liberal north). Likewise a religious zellot like Buchanan (poor guy.. with all his bad press, I almost feel sorry for him), should have little ability to win any state.. Though he may have 1 - 3% representation nation-wide, no single state is sufficiently right winged to carry him.. A similar case with Nader.
The one down side, is that it is unlikely that a totally new type of person could be elected.. So, for example, Jessie Jackson probably couldn't get elected.. BUT, colon Powwel might.. But for a different reason.. The president is supposed to be a national icon (similar to the Queen of England). The president is supposed to be our international representative, as well as making the most public descisions on which bills to pass, and selecting the types of people to run various government agencies. This president should be the conglomeration of all of it's citizens (the melding of a president). From this, war heros tend to be excelent choices.. They embody our pride and ideals, so other minor choices, such as issues are less important. Unless they strike nerves with different regions of the country.
As for the melting pot metaphore.. US citizens are not the only ones that coin that phrase.. Many famous Europeans have as well. If you were to go to my high school, you definately would have to agree that there is a merging of cultures. Yes there is still the ghetto, and the rich preppy development, but the middle class suburbia is becomming more and more diverse. The melting pot is like having different colored clumps meltable chocolate sitting on a frying pans slowly melting. The bulk of each piece is still solid and distinct from one another.. But over time, more and more of the liquid blends together to ultimately become indistinguishable.
You can not argue that there aren't elements of the middle class that are distinguishable.. I see this especially in high schools and colleges. Wealthy work along side the poor, black among the white... Granted, the extreme of the cultures (the eletists or the home-boys (of any race)) tend to keep to themselves, but there is a large growing group that sits happily in the middle.
Just think, not too long ago, the color of your hair or your european nationality was a big divider (Italians wouldnt mingle with the Irish, etc.). Now, I can barely notice someone's European nationality, though african/ asian is still obvious.. In another century or two, I predict that we'll be at a very uniform level (if we don't kill each other).
-Michael
Re:The problems are...
on
eLection '04
·
· Score: 2
Rumour has it that MI5 used to (during the cold war) check up on those who voted communist, though this is all denied, of course...
I would venture to guess that US history has had incidents with bullying people who have voted opposing parties in the past.. THe simplest thing I can think of is the good ole' boys, who's members include the police officers that protect the ballot boxes.. And much like the Amish, I doubt that any good ol' boys are reading this, so I feel safe in saying that.:) Other possibilities are the mafia who simply want to assure that their corrupt leaders are re-elected.. Though they probabily take hits on the vocal supporters of opposing candidates.
I am humbled buy our situation, but the problem, I think, would be likely no matter what voting system we had.. A race this close is impossible to resolve amicably - unless one side conceedes.. But I think Chivalry went out with Nixon; strangely enough.
As for the accuracy, In any analog system, you're going to have margins for error. paper-count measurement is based on the exact positioning of the card at the moment of reading.. Those partial punctures really make a difference (which is why I fully support the idea, that some senator brought up, in aboloshing paper voting in the US).
Sometimes you have to have something this severe to make a national change. You can bet that butterfly votes won't be seen here again (and possibly anywhere else in the world)
Re:The problems are...
on
eLection '04
·
· Score: 1
yuo are a whiney fucken BITCH!
Somebody has to be the scape goat? (aka, the voice of reason.. correct or otherwise)
-Michael
p.s. Sometimes I like being on top too ya know.
Re:Problems with the system
on
eLection '04
·
· Score: 2
The punch-card system (which is used in *many* places around the country) is known to not be accurate.
Well, I wasn't really supporting the existing system.. In fact, I think paper cards are diplorable. By I tried to argue that there are better ways than computers to do simple incremental arithmetic.. My state of Delaware, for example, uses a big electronic shmorgasboard.. More expensive than a simple Kiosk, I'm sure, but it has fewer components than a computer (namely each line of code).
A touch-screen system would be much more accurate for many reasons. It could give feedback to the voter (confirming the voter's choices) which would have eliminated this whole mess in Palm Beach County.
Again, I sorry if I sounded like I was supporting the existing system.. Delaware's system provides flashing lights as feed-back as well.. Plus there are physical push-buttons, which tend to make people happier.
You also seem to be confusing the cost of the *campaigns* with the cost of the *election process*. These are two completely different things.
Yes and no.. I did make a seperate reference to government and state-run election processes. I was blending the opinions of political expenditures which probably didn't belong. In Delaware, we're pretty rich per capita, so we could afford to buy our [supposedly] expensive machines. A poorer state with a sigificant population might simply not be able to afford it.. Who's to say? THE TAXPAYING CITIZENS. I'm sure that the Delaware senate voted on our election system.. Likewise, Florida will most likely vote on a new system (after such negative controversy with the existing system). They'll figure out what's best for them.. And who knows.. Maybe the'll like touch screens.
Sure, each piece of paper costs very little, but you actually have to *print* on them which costs more. Plus, you have to print *millions* of them. All paper ballots are counted by machines already, and those machines are, in many places, *really* old. They need to be replaced. Those machines aren't cheap.
Well, you only need a handfull of printers and readers, where-as you need hundreds of computer terminals. I can't imagine that printed paper is more expensive than hardware though.. Look at Junkmail! I'm really just playign the Devil's advocate here.. I really don't want anyone to use a paper system.. I'm just saying that I highly doubt that an electronic system will be more cost effective... As you pointed out, Cost isn't always a factor, but as I tried to suggest, sometimes it is.
As these old systems are replaced, they shoudl be replaced with easier to use, more reliable systems. This seems like a no-brainer to me
I totally agree with increasing reliability and ease of use (voting should be idiot simple and fool proof.. This is the ot's man!!) However, as I tried to point out.. A superficial look into the situation isn't sufficient to declare what's best for everybody. And thankfully because of state jurisdiction, no single organization will be able to be an imposition. Each state, and then each district will decide for themselves what's the most cost effective way to prevent Palm beach from happening again.
Personally, I'm biased towards hardware.. Use as little software as possible.. Fewer chances for problems. A mechanical drive train only has a few possible problems.. An electric 4 wheel drive vehicle with computer control could have any number of problems... Especially when you leave the air-conditioned server room environment, and hit the mud slapping -30 to 120 degree, shock-UNobsorbed real world..
Ok, I'm getting tired of repeating myself, but I know not everyone will read the entire article.
Read the following:
http://slashdot.org/article.pl?sid=00/11/09/1350 223&mode=nested
I had some bad luck posting it as a link, so I gave up.
It says that a popular election is only more fair if you have perfectly random people.. Or if they are biased (as is always the case), then only if they are perfectly distributed throughout the country.
Yes, our melting pot is working it's way towards uniformity, but it's not quite there yet.. The opposite extreme is Serbia. If one religious faction was the dominant, then their candidates would ALWAYS win, irrespective of the other religions.. In an electorial system, only candidates that appeal to a majority of RELIGIONS (not voters) will win. (especially since it's unlikely that the population is 100% evenly mixed throughout all districts)
Unfortunately, candidates in the US have learned that being moderate and accepting the popular opinions on a majority of issues will get them the most votes.. BUT THIS IS NOT BAD. If you have two candidates that are garunteed not to be radical (for fear of being de-elected), then you're less likely to have disasters. A libertarian would never be elected, because he is too radical (even if 51% of the people liked the idea of no taxation). The electorial college prevents racists and radicals from being elected by enhancing the voting power of minorities..
Again, the effect dimishes as you randomize the location of the population. But there are natural grouping forces that should resist this.. (such as the proxity of immagrant entry points, wealth clustering areas, or farming areas, etc).
Why shouldn't we all be forced to go to a protestant Church, eat at McD's, buy MicroSoft, surf with AOL, and buy Intel Chips, or shop at the Gap?
Offset the balance in the name of diversity. Life will thank you for it later.
In answer to your question however. That the same minority in Rhode Island has a greater percentage vote than a minority in NY is circumstantial. We've shuffled the population around the country so much that the original boundries no longer accurately represent the same geographical idiologies. But the point is less that a minority will help pick the election, and more that a minority that consists of 10% of some (but not all) districts, means that appeasing them has a greater effect than if they only counted as 3% of the total population. The electoral process ONLY fails when you have 100% purely distributed population. If _every_ state had an equal percentage of each race and idiology, then a general election would be more fair. But We don't have many Cuban farmers in Kentucky that I'm aware of.
I disagree. As other posters have said, the reasons your vote meant little might have had more to do with the heavy sway of the general population of your state. Additionally, your vote might have counted if there were sub-district votes. Sadly, my state of Delaware with only 3 votes could not have been split up very fairly (maybe 1 vote per county).
Sadly, as spoken by the mathematician featured in yesterday's slash-article, very close races don't fair too well for electorial races. Only the bigger states really matter at that point, and individual voting power actually decreases with respect to a purely popular vote. But it is very rare to have such a close election.
If you knew you only had 30% of the vote, however, you would be concerned with every last small state. George H Bush made a visit to Dover Delaware (my home town) during the 92 campaign, for example.
The best reason, however, still lies within the federalist papers.
-Michael
Re:Problems with the system
on
eLection '04
·
· Score: 2
Should there be a price tag on the democratic system?
Um.. I'd say yes.. We spend entirely too much on the election system as it is.. 60Million to win a NJ seat... I'd be curious to see the costs of the states and Fed for the entire process. Anybody have any numbers?
Anyway, throwing money at a problem doesn't necessarily make it better. Just like internet access in the schools.. Just because you can do a thing.. It doesn't mean that you should. (famous quote from where? Star Trek I think?) The point is that computers are not the best solution to every problem. It's an engineering problem, and a thinking engineer will look at many possible solutions and weigh them.
A touch screen has many problems.. Scratches, breakages... (If I didn't like Bush, I could have totally scratched out his name on the screen so nobody else could vote for him). People's general distrust of something that looks like a computer.. These are very real concerns. I'm not saying that a computer system can't solve the problem (other countries seem to think it can). But would you really trust your life to software at this point? Would you let an software company like MS drive your car or fly your plane? Or how about picking your president. I love computers, but even I'd say no.
(Note, dispite popular flaming, MS isn't a BAD software company.. They're just so market driven that they have bloat-ware which lends itself to errors - by sheer volume. It's a testiment that they've made such complex systems actually workable. )
As for the apparent added "fair"ness in a computerized system. Read any of my other articles on this topic. Computers add many weak links to the chain. It's just bad engineering for a life-supporting real-world system. Yes it can be MADE to work, but it's a bad foundation. The only thing you really get from computers is added functionality - but doesn't that sound like MS's point of view?
Lastly, the issue of marginal cost.. A sheet of paper costs less than a penny. A reliable and secure server (with support) costs thousands of dollars. Kiosks probably cost the same as existing stations. We're no where near the fictitious paperless office, so it's really a moot point.
Re:The problems are...
on
eLection '04
·
· Score: 2
Slashdot Mirror
The French have had a history in blocking American Air-waves; I see no reason why the internet should be any different.. If the French Government doesn't like what America (or any other country) has to say, then they can simply close their ears as they've done for decades.. And if the people don't like this sort of "law and order".. Well, they can revolt (civilly of course :)
Seriously. As a yongster, I was appauled that the KKK could hold a march in DC. But as I've gotten older, I understand better.. It is not right to supress opinions (so long as those opinions do not migrate into harmful actions). Those that adore the Nazi regieme are entitled to their opinions. It is this very sort of "Treaty of Vers..(sp??)" punishment and attempts at covering up ugly parts of the world or history that has exemplified trouble.
This is obviously a trivial (albeit sensationalized) case of sensorship (who's to say that Nazi imagry is less offensive or more dangerous than child-porn.. Personally, I don't understand the puritanism of anti-pornography to begin with.. It's just sex.. But that's just my opinion).
Lastly, if France really wants national sensorship then have THEM put firewalls/Filters at their borders! Sensor at the client, not at the server.
-Michael
If you have a 24/7 internet connection, then it's a fixed cost, and you can't factor in the "marginal cost" of the download of a distribution. This is especially true if you download it from work, where the fixed cost is paid by someone else.
Heck, I'm still at a University, where I don't even directly pay for the internet connection.. The fixed cost is bundled in with thousands of other costs. Being that I get a better part of a tax rebate on tuition costs, and my grants are never taxed, it's entirely possible to completely avoid taxation of free software alltogether.
This is obviously an extreme, and some people are plagued with pay-per-hour usage.. But even if you get a 50hr / month fixed rate, you can't economically consider that to be taxation of your individual online downloads.
I'm still waiting for the following scenereo: fiber to the home, but you pay for the BW at the switching station.. Charge people for BW just as you charge them for their electric bill. Most Americans don't consider their electric expense in their every-day lives, and I think the same would be true if we had pay-per-usage of the internet, so long as BW / cost is marginal. With this, governments could tax you just like they do your normal utilities, but this would still be such a marginally small amount, that it would be on par with electricity; once again.. The main advantage is that, just like electricity, you'd have a virtually unlimited potential, but you wouldn't want to abuse that potential often because of budgeting concerns.
-Michael
MB's can be very expensive.. Especially when they're in low availability. When the Athalon came out, the MB's were almost as expensive as the chips.. Throw in the fact that they were buggy (since it was all brand new.. including the EV6 connectors). The P4 requires an all new powering structure, so it's possible that we're going to find MB's that blow up the CPU's.. Thus you might want to wait a few rounds before considering these guys (this is usually a good rule of thumb anyway).
Given that it's brand new, you're also going to be at the mercy of the MB manufacturers in terms of features.. If you like ultra-SCSI-RAID on your MB, it might be a while. There is a similar argument to be made for your case.
Expect to pay premium dollars for this combo.
The problem with the head-sink is logistics and ergonomics.. Where do you put everything? Unless you have a monster tower case, you're not going to be able to fit too many full lenght boards.. You have a massively over-heated PAIR of RDRAM RIMMS, a MASSIVE CPU, a larger than normal power-supply.. And then you get to have your first AGP card. A Mid-Tower is probably too small for these guys.. That might not be a problem for some.. But personally, I like stacking my computers, so it's a problem for people like me.
The heat-issue also means real-estate for lots of cooling.. And an increased risk of heart-failure.. So to speak.
My guess is that you're looking at $300 - $400 for the MB + case alone. That's more than I usually spend on an entire bare-bones system. Throw in roughly $600 for each of CPU and memory, and you've got a rather large handi-cap for an almost miniscule performance gain.
As for the pricing of their CPUs... To be fair, they _have_ to depretiate their costs.. It's not cheap to design and entire CPU from scratch.. Remember, that this is their FIRST CPU redisign since the pentium Pro some 5 years ago!! Everything else (with the exception of the vapor-ware Italium) has been an add-on to their old archetecture.
I don't know what their margins are per CPU - I wouldn't be surprised if it was 100% - But when you take in the cost of multiple billions of dollars, it's not all money in the bank. AT&T use to depretiate their hardware-costs over 40 years.. Intel doesn't have such a luxury.
-Michael
I don't know if I agree with this.. These benchmarks show that Intel's 1.5 is roughly equivalent to AMD's 1.2, and we were told to expect this by the previewers.
.13u before Intel), then they can start pulling away.
.13u is going to be hard to beat. There are no benchmarks for AMD's sledgehammer, so there's no point in speculating about it.
I believe AMD uses a fully pipelined FPU (multiple ones at that). I'm sure that the P4 also fully pipelined their FPU, BUT, they added several additional stages to the basic instruction as well.
The addition of stages does two things: first it increases your max clock rate all-else-held-equal. And second it increases the latency for missed branch predictions. Again, all-things-being-equal, the missed branchs will tend to hurt more than the higher clock helps, except in a few special cases.
Intel, therefore included with those extra stages a highly advanced branch predictor that lives in the MIDDLE of the pipe. More-so, successfully predicted branches can skip the first several stages thanks to the branch cache. Thus, well-behaved code will get the benifit of heavier pipelining with fewer of the pit-falls. To make things even more tantelizing, they're using a 2x clocked Integer Unit. Thus they have a 3GHZ integer unit on these benchmarks. That means that they can further extend their pipelines with almost no visible penalty (even in branch-misses).
Unfortunately, they still seem to be plagued with branch misses (the only logical explanation for why AMD can still keep up or even surpase them). Obviously the memory played an important role in these benchmarks.. The KT133 v.s. AMD560 really only differ in memory speed, and that was enough to sway several percentage points. A more fair comparison would be between VIA's up and comming DDR-SDRAM P4 chipset.
But, as was pointed out; if Intel can get the P4 up to 2GHZ before AMD can (last rumor I heard was that AMD was going to hit
Unfortunately, as several sites pointed out, buyers don't look at benchmarks, they look at CPU speed, so Intel should be able to wrongfully win people over on this synthetic basis. Thankfully, the only people that are going to be willing to buy P4's are people needing servers (or maybe even Q3). We'd have to see NT ASP/Sql Server and or Linux+Apache+PHP+Oracle, etc to determine who's king (including memory types). Unfortunately I rarely see benchmarks on these grounds.
Sooner or later AMD is going to come out with their 64bit proc. With Mustang gone, this is their only next-great-hope. An all new design - hopefully without a tremendous cost - that has started from scratch (as the P4 did). I'm sure it too will have a heavy pipeline, but several of it's new features (such as the flat-memory archetecture) should enhance the playing field.
By then, however, the P4 will have found a new chipset that handles DDR-SDRAM and will have enough volume MB's and cases that it'll be cheap enough for the hard core gamer and possibly even casual gamer to purchase. A 2 - 3GHZ processor running at
I totally agree with another poster that said we should be rooting for both Intel AND AMD since competition is good.
-Michael
I do not believe any "ThunderBirds" ran at 600MhZ.. Think they started at 700MHZ... Since you didn't say Duron, I assume that you have the good ole fashioned Athalon.. Which means that you have a slotted MB, and thus can not make use of the 1.2GHZ socked-A processor.
But you can get a new MB for $120, so it shouldn't be that big of a deal.. Except that you'll probably want to get PC133Memory (since you most likely had PC100 on that older board).
-Michael
Really! I've been playing the latest 3d games with my matrox card for quite some time. Perhaps it's all been a hallucination! :(
I said keep up, not be able to produce a product. The 3D market is very volitile; one year, you make the fastest / cheapest card and you're topping the sales charts, then the next, you're merging with another company because you can't hold your own.
Matrox is in no danger of going under, 3D isn't their primary market. I respect Matrox since I've always thought of them as a quality company.
I don't forsee them even leaving the 3D market, but if you're a gamer, the Matrox card isn't the best buy for your money. I've never seen any review that has suggested otherwise.. The main things that might persuade a hard-core-gamer to choose the Matrox card is their Dual monitor or large monitor support.
That being said, any given business should probably opt for Matrox cards (due to quality, features and drivers for things other than games), so if you game-play at work, then Matrox is an obvious choice. But most of us don't get to choose the graphics cards at work (well, I do, but I bring my own machines to work).
-Michael
On another subject, did you drop all the users from blacknova.net? My user (tolket) is gone....
Laugh, I'm not in charge of that, though I'm one of the OpenSource developers.. Yes, the DB was completely flushed... You need to create a new character.. Read the forums.
-Michael
3Dfx burned a lot of bridges when they left the OEM market.. I'm sure they'll get back in, but they've lost much of their clout.. They're going to have to make a top notch card for people to promote 3Dfx over nVida.
Additionally, I don't think _anyone_ wants to go back to the add-on-card market.. With resolutions reaching 1600x1600, those pass-through cards are simply not acceptible. Back in the days of 1024x768 at 16bit color (in SLI mode), you could always put two monitors on your computer (since you obviously had money to burn).
Additionally, the only reason people went add-on card was because there was no all-in-one solution that could even compete.
It would be an extremely hard sell to say the least.. Though you might find a market back in the 200MHZ legacy computer market (with a full blown GPU).
Another SERIOUS issue is AGP.. Can't do SLI with AGP (unless you're Obsidian and obsessed), and you definately can't be an add on card and use AGP.
I realize you only meant this as a "worst case", but I wanted to debunk it as being profitable for them.
-Michael
Matrox has never been able to keep up in the 3D market.. They've primarily been a professional 2D card company. I don't think they've ever topped the 3D charts (though you can put 1600x1600 monitors on them without a sweat).
They were innovative with their environmental bump mapping, and for a while there, they'd produce the cleanest pictures, but apparently this isnt' the case with their latest cards, and ATI (at least) has already caught up with them in the quality realm.
-Michael
You're missing the point.
Competition is GOOD.. Winners that gain a monopoly in the 3D video card industry is BAAD.
ATI and nVida should be good competition with each other for a while. Both making state-of-the art cards (at almost the complete price range), but once one of them goes... That's about it..
The time to enter the video card market ended with nVida, who stepped up the competition until the then king couldn't compete. With the world consolidating towards DirectX and, on occasion, OpenGL, you can't depend on your proprietary drivers (a la Glide, and many other's that I've already forgotten). Most likely, the only way you can make a faster, more cost effective card is through propriety drivers (such as the failed infinite plains card (forgot the name of that too)).
Couple that with the fact that GPU's are now as (if not more) complex as CPU's, in addition to using state-of-the-art manufacturing, it's highly unlikely that a new contender can possibly dethrown the existing big boys.
In fact, the only thing I can image you could add is move the whole damn game into the GPU (a la MicroSoft's DirectGame or Direct3DShooter... ).
ASICs wired video games as the future?
-Michael
We're an offshoot of Dupont, and we've had flex-time for the longest time.. What we have are "core" hours from 9am till 3pm. You get to choose your 8hour day outside of that.
The advantage of shift-time's are that people can know when they can find you. Various activities such as network administration can be "paged" when needed, and coders are sent on week-long excursions which could even run into the weekend if necessary, so the exact hours that they work aren't always important. Secretary positions, however, tend to be more important and lend to shift-work (as would assembly line work).
For us, various benifits like a kitchen, flex hours, free soda, etc were in lue of higher pay, and for me, that was just fine. If I had to choose between higher pay and a less stressful working environment.. I'm all for living a higher quality life.
-Michael
First, I started off with the words "Sadly, I don't know".. Therefore removing myself as an expert in that area. It was more a question than anything else.
As for the other languges, perl, tcl, etc, those can only be less optimial forms, and are more for procedural operations, which has nothing to do with over-comming the 16 join problem. (With the exception maybe of hashed-lookups, but that's dangerous)
It's great to hear about cached query plans (I'd forgotten the term).
As for the problem with speculation.. If nobody were to speculate, and those in the know don't impart their knowledge, then has anyone gained anything? 'Qualified' speculation does exactly as here; Inciting those knowing specific facts to quality errors. My speculation _was_ based on emperical analysis, however, so it wasn't as if I was blindly stateing something.
But now, as a question, is the compiled sql stored in any more optimal fashion? Or does postgres use an interpreter (which does little more than tokenize). To me, a compiled sql statement means that its making raw calls to table operations..
Simple benchmarks that I've done haven't shown any performance increases with using stored procedures, but that might just be because the compilation / interpretation stages were insignificant with respect to the overall operation. The other interpretation is that little is actually gained by the compilation of the sql statement (at least until the cached query plans come around).
-Michael
As a followup idea.. 16 joins reminds me of a project that I did once where we used ID's for every table.. Thus every table stored meta data and a single piece of actual data.. Thus if you had 17 pieces of data to display, then you'd have to do 16 joins. There were three solutions to this:
First, cache all the smaller tables, and do the translation (from say user_id to user_name) in an external hash (I know perl can easily do this; don't know about php). For static fields such as user-names, email addresses, etc, this is easy. Unfortunately this doesn't help when trying to find all users with email addresses going to aol and souce code that's over-due, etc. But it does solve the summary report problem.
Second is to break up the query into multiple operations.. You need an incredible amount of memory to do certain types of joins, and sometimes the sql-processor guesses the wrong method. Even though postgres's "explain" command told me it was doing hash-lookups for a user-name summary, I found that it was faster to first get all the data with user-id's, then perform a second search with
"select user_name from users where user_id in (x,y,z,....)"
So long as you don't have more than a couple dozen entries, this works nicely.. ESPECIALLY if you have thousands of rows of real-data with only one or two unique user-names. This was a nice-short term fix, but "clean" sql should all be done on a single statement.. Soo onto the third
Denormalize back to form 4 or 3. Namely, whenever you have a user-name, don't use a user-id.. Sure you consume 8 more bytes of memory per row, but it greatly speeds up searches. The method is to ONLY use ID's when you can not uniquely represent a primary key with a single column.. (such as first-name, last-name,...).. But in our computer world, "user-name" tends to be unique. "project-name" should be unique, "working-branch" should be unique.. In which case, at the design level, it's worth the extra consumed space.. Now I know that Mysql advocates (including me a year ago), cringe at the idea of keying off variable length strings.. For me, it was the idea that you can compare integers much faster than strings.. BUT, if the string is an index, then your sorting is going to be MUCH faster, strings distribute much better than auto-incremented integers (which tend to want to be linear). Additionally, the fact that you won't have to perform a join at all removes almost all of the performance problems.
When you don't use meta-data as primary keys, then you find that you only perform joins when you actually have relational data... That joins based on summary data almost always goes away.
Additionally, if your DB ever get's corrupt (say you lose a table), then you don't have to fret over "who the hell was user-id 155?". Given that Mysql doesn't have roll-backs, the potential for data corruption with meta-data is enormous (when you do both reads/writes).
-Michael
Essentially what that test showed was that mysql had to lock all those tables in order to do the join.. And that with Postgres's versioning system, it was a trivial matter.
I remember reading on the topic of excessive joins in a Sybase Admin book once (excelent book on "practical" database, since it shows you exactly how one company solved all the DB problems). It essentially said that determining join order was actually a very slow process.. Exponential in fact.. Especially if your DB tries to determine which indexes to used based on statistics... The selection of indexes increases the number of possible variations incredibly. Sybas's solution was to only compare a select few of the possible join methods - which is a trade-off.. Alternatively, if you were to create a stored procedure, then which-ever method was best at creation time will be used from there-on-out...
Stored procedures - in Sybase at least - reduce a great deal of the overhead, especially in a 16-way join. Sadly, I don't know if Postgres does ANY pre-compiling of stored procedures. They have two main methods: straight named 'sql' statements, and 'psql' (a true stored procedure 'language'). My guess is that they do little more than pre-parse the statment (if even that). If this is the case, then using a stored procedure might even be slower than raw SQL since a call requires the parsing of the function name, then loading of the function into memory from a table, then the parsing of that statement (two more steps than raw processing). It is, however, necessary for triggers and rules.
Postgres (and I believe mysql) has raw-c extensions, but I really dislike these for general DBA operations (not to mention I haven't spent the time learning them).
As for the 16-table join, a general web page probably shouldn't do this (especially a heavily used page), but a database SHOULD at least let the DBA perform the ad-hoc query when trying to extract information. The only real solution is summary information, which often tends to be more trouble then their worth... Another solution is de-normalization (which makes me feel dirty)... Basically like what Blacknova.net uses... If you have 1 to N data, then just pick some upper bound N and include that many extra columns directly in your main table. Fast and efficient, provided that you choose the right N.
-Michael
Mysql uses multi-threading (each connection gets it's own thread), but I've read slashdot posted benchmarks that show Mysql failing after so many threads are spawned via concurrent connections(I think 60). Perhaps they're using some finite resources and the thread spawner was not using semephores or some other such resource control.
Postgres uses multi-processing (each connection gets it's own process) with shared memory segments and a controlling/monitoring process (postmaster).
Thus if Mysql could do a better job of managing resources under heavy loads, then it would waste less physical memory. Additionally, I don't know how good Linux is with multi-threading under heavy loads. Can it faithfullly handle hundreds of potentially thousands of concurrent/competing threads? I've never seen any tests to show this... I'm aware that the default max processes/threads was hard-coded in the kernel to 512 at one point (I believe it requires a recompile to change). This might have something to do with mysql's crashing.. You'd have to go to single-user mode to max this puppy out. Additionally I don't know if it's possible for an app to determine how many more threads can be spawned by the OS.. If a thread fails to spawn, then you can corrupt the entire process.. In MP, if you fail to spawn a process, then you simply lose that connection.
In general, I'd prefer MP on a Linux system. It's a similar story with Apache. You get a lot of reliability, and not _that_ much memory waste when compiled correctly. Additionally, MP has the potential to span computers (as in a slasdot article on a multi-hosted extension to Linux)
-Michael
the method of dividing the power is completely incapable of increasing it. This is quite obvious.
I'm sorry, I don't understand what you're saying.. It isn't obvious to me at least. Perhaps instead of subjectively berating the system with unsophisticated language, you could provide evidence (you know.. being scientific and all).
The idea is that you want to reduce the power of the majority.. Geographically and ideologically. Power is a bi-stable system.. The minority has almost no power until just about the 50/50, then they suddenly are empowered almost instantaneously. After that, they grow exponentially until no-one else even has a say. This is not a fair system, and this is what the Federalist papers tried to prevent.. Mob rule.
NY and California are peeked.. Their voters will not achieve any more power, but that's fine.. The two of them alone carry a better part of the power required. NY is it's own little idiology (it's own melting pot), but California is rather liberal, while Texas is rather concervative. Should CA and NY alone determine the president, ignoring the Geographic differences in idiology? Granted, Texas has enough power to offset this, but then you have the remaining southern states, who population wise could not affect the presidential outcome much, but electorially they can.
Another issue of dividing power "increasing" it, is with his example of gerimandering(sp?). Ignoring the president for a moment, if you put all blacks in one district and gave them a black representative, then that senator would only have 1/5xxth of a voice. If, however, you spread those people out among several districts, then you could influence 5-30% of the vote for multiple districts (including sympathists), thereby having dozens of representatives in congress, a significantly greater margin.
Ahh.. I don't care to argue anymore.. If Hillary has her way, it'll be abolished anyway, and the math deficient among us will simply look at this election as all the proof they need.
-Michael
HA HA HA HA!!! So that is why we haven't had a White president in the last 50 years!!!
Well, as confusing as this statement is, I'll assume that you're simply refuting my statement and saying that the electorial college has not prevented racists / radicals, and additionally has not allowed minorities to lead.
Well, let's use some emperical evidence.. IF the southern states had the majority of people in the country (thankfully they don't), and we didn't have an electorial college, then a former KKK person could actually be elected president. the north east corridor, however, would garuntee that such a thing could not happen (because they have a lot of people in NY and a lot of states overall). If you had a black radical who's primary goal was reperations for the mal treatment of blacks, then you'd have the entire south blocking them (even if there were 51% people represented by the liberal north). Likewise a religious zellot like Buchanan (poor guy.. with all his bad press, I almost feel sorry for him), should have little ability to win any state.. Though he may have 1 - 3% representation nation-wide, no single state is sufficiently right winged to carry him.. A similar case with Nader.
The one down side, is that it is unlikely that a totally new type of person could be elected.. So, for example, Jessie Jackson probably couldn't get elected.. BUT, colon Powwel might.. But for a different reason.. The president is supposed to be a national icon (similar to the Queen of England). The president is supposed to be our international representative, as well as making the most public descisions on which bills to pass, and selecting the types of people to run various government agencies. This president should be the conglomeration of all of it's citizens (the melding of a president). From this, war heros tend to be excelent choices.. They embody our pride and ideals, so other minor choices, such as issues are less important. Unless they strike nerves with different regions of the country.
As for the melting pot metaphore.. US citizens are not the only ones that coin that phrase.. Many famous Europeans have as well. If you were to go to my high school, you definately would have to agree that there is a merging of cultures. Yes there is still the ghetto, and the rich preppy development, but the middle class suburbia is becomming more and more diverse. The melting pot is like having different colored clumps meltable chocolate sitting on a frying pans slowly melting. The bulk of each piece is still solid and distinct from one another.. But over time, more and more of the liquid blends together to ultimately become indistinguishable.
You can not argue that there aren't elements of the middle class that are distinguishable.. I see this especially in high schools and colleges. Wealthy work along side the poor, black among the white... Granted, the extreme of the cultures (the eletists or the home-boys (of any race)) tend to keep to themselves, but there is a large growing group that sits happily in the middle.
Just think, not too long ago, the color of your hair or your european nationality was a big divider (Italians wouldnt mingle with the Irish, etc.). Now, I can barely notice someone's European nationality, though african/ asian is still obvious.. In another century or two, I predict that we'll be at a very uniform level (if we don't kill each other).
-Michael
Rumour has it that MI5 used to (during the cold war) check up on those who voted communist, though this is all denied, of course...
:) Other possibilities are the mafia who simply want to assure that their corrupt leaders are re-elected.. Though they probabily take hits on the vocal supporters of opposing candidates.
I would venture to guess that US history has had incidents with bullying people who have voted opposing parties in the past.. THe simplest thing I can think of is the good ole' boys, who's members include the police officers that protect the ballot boxes.. And much like the Amish, I doubt that any good ol' boys are reading this, so I feel safe in saying that.
I am humbled buy our situation, but the problem, I think, would be likely no matter what voting system we had.. A race this close is impossible to resolve amicably - unless one side conceedes.. But I think Chivalry went out with Nixon; strangely enough.
As for the accuracy, In any analog system, you're going to have margins for error. paper-count measurement is based on the exact positioning of the card at the moment of reading.. Those partial punctures really make a difference (which is why I fully support the idea, that some senator brought up, in aboloshing paper voting in the US).
Sometimes you have to have something this severe to make a national change. You can bet that butterfly votes won't be seen here again (and possibly anywhere else in the world)
yuo are a whiney fucken BITCH!
Somebody has to be the scape goat? (aka, the voice of reason.. correct or otherwise)
-Michael
p.s. Sometimes I like being on top too ya know.
The punch-card system (which is used in *many* places around the country) is known to not be accurate.
Well, I wasn't really supporting the existing system.. In fact, I think paper cards are diplorable. By I tried to argue that there are better ways than computers to do simple incremental arithmetic.. My state of Delaware, for example, uses a big electronic shmorgasboard.. More expensive than a simple Kiosk, I'm sure, but it has fewer components than a computer (namely each line of code).
A touch-screen system would be much more accurate for many reasons. It could give feedback to the voter (confirming the voter's choices) which would have eliminated this whole mess in Palm Beach County.
Again, I sorry if I sounded like I was supporting the existing system.. Delaware's system provides flashing lights as feed-back as well.. Plus there are physical push-buttons, which tend to make people happier.
You also seem to be confusing the cost of the *campaigns* with the cost of the *election process*. These are two completely different things.
Yes and no.. I did make a seperate reference to government and state-run election processes. I was blending the opinions of political expenditures which probably didn't belong. In Delaware, we're pretty rich per capita, so we could afford to buy our [supposedly] expensive machines. A poorer state with a sigificant population might simply not be able to afford it.. Who's to say? THE TAXPAYING CITIZENS. I'm sure that the Delaware senate voted on our election system.. Likewise, Florida will most likely vote on a new system (after such negative controversy with the existing system). They'll figure out what's best for them.. And who knows.. Maybe the'll like touch screens.
Sure, each piece of paper costs very little, but you actually have to *print* on them which costs more. Plus, you have to print *millions* of them. All paper ballots are counted by machines already, and those machines are, in many places, *really* old. They need to be replaced. Those machines aren't cheap.
Well, you only need a handfull of printers and readers, where-as you need hundreds of computer terminals. I can't imagine that printed paper is more expensive than hardware though.. Look at Junkmail! I'm really just playign the Devil's advocate here.. I really don't want anyone to use a paper system.. I'm just saying that I highly doubt that an electronic system will be more cost effective... As you pointed out, Cost isn't always a factor, but as I tried to suggest, sometimes it is.
As these old systems are replaced, they shoudl be replaced with easier to use, more reliable systems. This seems like a no-brainer to me
I totally agree with increasing reliability and ease of use (voting should be idiot simple and fool proof.. This is the ot's man!!) However, as I tried to point out.. A superficial look into the situation isn't sufficient to declare what's best for everybody. And thankfully because of state jurisdiction, no single organization will be able to be an imposition. Each state, and then each district will decide for themselves what's the most cost effective way to prevent Palm beach from happening again.
Personally, I'm biased towards hardware.. Use as little software as possible.. Fewer chances for problems. A mechanical drive train only has a few possible problems.. An electric 4 wheel drive vehicle with computer control could have any number of problems... Especially when you leave the air-conditioned server room environment, and hit the mud slapping -30 to 120 degree, shock-UNobsorbed real world..
-Michael
Ok, I'm getting tired of repeating myself, but I know not everyone will read the entire article.
0 223&mode=nested
Read the following:
http://slashdot.org/article.pl?sid=00/11/09/135
I had some bad luck posting it as a link, so I gave up.
It says that a popular election is only more fair if you have perfectly random people.. Or if they are biased (as is always the case), then only if they are perfectly distributed throughout the country.
Yes, our melting pot is working it's way towards uniformity, but it's not quite there yet.. The opposite extreme is Serbia. If one religious faction was the dominant, then their candidates would ALWAYS win, irrespective of the other religions.. In an electorial system, only candidates that appeal to a majority of RELIGIONS (not voters) will win. (especially since it's unlikely that the population is 100% evenly mixed throughout all districts)
Unfortunately, candidates in the US have learned that being moderate and accepting the popular opinions on a majority of issues will get them the most votes.. BUT THIS IS NOT BAD. If you have two candidates that are garunteed not to be radical (for fear of being de-elected), then you're less likely to have disasters. A libertarian would never be elected, because he is too radical (even if 51% of the people liked the idea of no taxation). The electorial college prevents racists and radicals from being elected by enhancing the voting power of minorities..
Again, the effect dimishes as you randomize the location of the population. But there are natural grouping forces that should resist this.. (such as the proxity of immagrant entry points, wealth clustering areas, or farming areas, etc).
-Michael
Why shouldn't we all be forced to go to a protestant Church, eat at McD's, buy MicroSoft, surf with AOL, and buy Intel Chips, or shop at the Gap?
Offset the balance in the name of diversity. Life will thank you for it later.
In answer to your question however. That the same minority in Rhode Island has a greater percentage vote than a minority in NY is circumstantial. We've shuffled the population around the country so much that the original boundries no longer accurately represent the same geographical idiologies. But the point is less that a minority will help pick the election, and more that a minority that consists of 10% of some (but not all) districts, means that appeasing them has a greater effect than if they only counted as 3% of the total population. The electoral process ONLY fails when you have 100% purely distributed population. If _every_ state had an equal percentage of each race and idiology, then a general election would be more fair. But We don't have many Cuban farmers in Kentucky that I'm aware of.
I disagree. As other posters have said, the reasons your vote meant little might have had more to do with the heavy sway of the general population of your state. Additionally, your vote might have counted if there were sub-district votes. Sadly, my state of Delaware with only 3 votes could not have been split up very fairly (maybe 1 vote per county).
Sadly, as spoken by the mathematician featured in yesterday's slash-article, very close races don't fair too well for electorial races. Only the bigger states really matter at that point, and individual voting power actually decreases with respect to a purely popular vote. But it is very rare to have such a close election.
If you knew you only had 30% of the vote, however, you would be concerned with every last small state. George H Bush made a visit to Dover Delaware (my home town) during the 92 campaign, for example.
The best reason, however, still lies within the federalist papers.
-Michael
Should there be a price tag on the democratic system?
Um.. I'd say yes.. We spend entirely too much on the election system as it is.. 60Million to win a NJ seat... I'd be curious to see the costs of the states and Fed for the entire process. Anybody have any numbers?
Anyway, throwing money at a problem doesn't necessarily make it better. Just like internet access in the schools.. Just because you can do a thing.. It doesn't mean that you should. (famous quote from where? Star Trek I think?) The point is that computers are not the best solution to every problem. It's an engineering problem, and a thinking engineer will look at many possible solutions and weigh them.
A touch screen has many problems.. Scratches, breakages... (If I didn't like Bush, I could have totally scratched out his name on the screen so nobody else could vote for him). People's general distrust of something that looks like a computer.. These are very real concerns. I'm not saying that a computer system can't solve the problem (other countries seem to think it can). But would you really trust your life to software at this point? Would you let an software company like MS drive your car or fly your plane? Or how about picking your president. I love computers, but even I'd say no.
(Note, dispite popular flaming, MS isn't a BAD software company.. They're just so market driven that they have bloat-ware which lends itself to errors - by sheer volume. It's a testiment that they've made such complex systems actually workable. )
As for the apparent added "fair"ness in a computerized system. Read any of my other articles on this topic. Computers add many weak links to the chain. It's just bad engineering for a life-supporting real-world system. Yes it can be MADE to work, but it's a bad foundation. The only thing you really get from computers is added functionality - but doesn't that sound like MS's point of view?
Lastly, the issue of marginal cost.. A sheet of paper costs less than a penny. A reliable and secure server (with support) costs thousands of dollars. Kiosks probably cost the same as existing stations. We're no where near the fictitious paperless office, so it's really a moot point.
Think of the cost.. What do we gain by it?