Economically, this stands to even things out, which is more in line with the socialist view and less in line with the capitalist view. Perhaps you should read the definitions of words before you use them?
As India and other countries start taking more jobs, and gaining ground in "mindshare" or respect or whatever you call it, and as their domain knowledge increases, they will raise rates to reflect their power in the marketplace. As India as a country grows economically, inflation will settle in compounding it further.
As soon as the cost-effectiveness of doing work overseas starts to fade, less work will be considered to go there - in reality, at that point it will be a battle of labor laws and how much control the outsourcing company has over the employees under those laws.
I think you can figure out the last part, which mostly revolves around the word "union". People who work hard to obtain the knowledge they have do not like to get stepped on.
Of course, any good socialist already knows that capitalists are just tools to help socialism grow anyways.
If you want a concrete example, try to make money on rentacoder.com - good luck with that.
Actually, and I know this is going to get some people to bellow loudly, Word and vi have more in common than Word and Emacs. Like I said, those who compare the two are completely missing the point.
Word and vi are editors that have capabilities to automate certain editing features.
Emacs on the other hand is a lisp environment with editing features.
It's a small and mostly semantic difference but an important one - the reason people have email and IRC clients in emacs and not in the former two have a lot to do with this difference, and understanding WHY it's important is key to understanding why so many people choose to develop applications like this in emacs. Seriously, it's less about fandom and more about flexibility.
Writing an equivalent Common Lisp environment for your mail client, or your IRC client is a lot more complex. Emacs provides the framework and allows for all of that.
I have no use for an IRC client that has that much power, but if VM had better IMAP support, I wouldn't be swearing at Mail.app all the time.
Jebediah, Bob, they're on our turf! Break out the shotguns, bourbon and banjos!
Seriously folks, don't you realize that competition historically has been a/good/ thing as long as the playing field is fair? This doesn't sound unfair to me in the slightest.
I mean, it's too much to ask to read the article, it's too much to read the posts, is it too much to ask these days that the context is at least right before you reply with idiocy?
Heh. I think the last MM I actually beat was 3 as well.
Capcom is really good when they get it right - they are also really bad when they get it wrong, Not unlike Konami.
I/wanted/ to like viewtiful joe, I just couldn't get past the fact that I was just playing yet another MegaMan with a different character, and the game was 4 times as hard for no apparent reason.
Heh oh yeah, 9, 8, and X-2 are the only ones I haven't beaten, and the latter two I played for quite some time before I quit in disgust.
If you haven't played the early series, go out and find roms of them - I think considering what you found interesting, even though the graphics suck by today's standards, the story is well worth it.
I think you're a little more flexible than most, but I can understand your viewpoint.
FF7, as I'm sure you can attest to, was less running of scene to scene and more about exploration. The original series pre-7 was the same way. They did start to cross over at that point and I do agree, 7 is not great but I think it's better than where they have gone with it.
FFX had an awesome story, and even a great way to extend the job system in 5 further, but I just couldn't get past the fact that I really had no room to explore - pointing the airship at destinations with a menu just doesn't feel the same as flying it around, and that's a good portion of my dilemma.
X-2 is such a deviation off the radar that I can't help but think that people who like it are merely fans of the name and not the actual content of the series, and more or less, it's original formula. While the content has always differed, the formula took quite a bit of time to morph into what it is now.
Frankly, the online FFXI seems to follow the original formula much, much closer than anything released after it, or in close proximity prior.
Actually, I find that sequels, the longer they extend, cater to a different audience than the one that made them famous in the first place:
Final Fantasy X-2. Heck, anything after 7. It's crap. All crap. In 7, the cutscenes are minor, like in the older games, and drive the story in ways that can't be driven interactively.
8 and on seem like an exercise in doing what is necessary to get to the next cutscene. I couldn't force myself to digest all of X-2, after 8 and X, it was enough. I haven't played 9, but I'm tempted to considering all the people that like 8, X, and X-2 say it sucks. These people coincidentally hate 7 and haven't played anything earlier.
I remember buying FF1 the day it came out in the U.S. and being AWE STRICKEN by a stupid blue screen with text that faded in.... Between the music and the contents of that text, I was in love immediately.
Now I get something that probably took 5 months to render that amounts to a very well-done computer-generated britney spears video. Strangely, that stupid blue screen was much more interesting.
I hear ya... Like I said earlier, I liked Viewtiful Joe when it was.. Mega Man. Anyone who's familiar with Capcom's history knows that these games are the only thing they've ever been good at, and I wouldn't say that's a good thing... Realistically, SF2 and MM were the only successful genres, and the original releases (although I am partial to MM2) were the best.
I'm personally waiting for Viewtiful Joe: Hyper Super Extended Championship Edition vs SNK and Marvel in a Primal Rage Pit Fight while practicing the Art of Fighting to engage in Mortal Kombat.
Do people buy shrek 2? Likewise, do people buy copies of the movie Tomb Raider in their home? Was it a best seller at the box office?
LucasArts is probably the only exception to the rule in movie tie-ins, and that's because most of their games are actually good - X-Wing and Tie Fighter, even without branding would have been awesome flight sims. I never got into JK, but I could see the potential. I think it has a lot to do with the fact that they aren't blindly following the movie plot and rushed to release at the same time as the movie. That said, Pod Racer was a giant pile of poo and... did not follow their normal protocol. Since then it seems they have taken to farming off the good games to developers who still know how to release good games, like Bioware.
Those are just titles, which really detracts from your point.
Fable, of those, seems to be the closest to an "original" idea... Although I find myself stretching even there. You won't be finding a tempest or even a tetris in that crop. All the games you describe are more or less the equivalent of card games - even though you're playing something with different rules, you're playing with the same medium - cards.
This isn't always bad, but don't confuse titles with a new story with an original idea. Personally I think HL2 is a better game than all of them and it's the only one that I played all the way through, and I think only fable and the first one (which I honestly haven't heard of) are the only ones I haven't bought - because Molyneux hasn't completed anything since Populous 1, just broken visions - if he had taken his ideas from Fable, incorporated them into Black & White, we'd have something (since they are basically the same game but with different perspectives).. But then everyone would say the graphics blow as they'd be circa 1999 - and as we all know, graphics and cut scenes really enhance the gameplay, which is why you bought the damn thing in the first place... right? Molyneux needs to figure out if he wants to be Will Wright (mundane ideas that are great, easy to get out the door) or Sid Meier (complex ideas that take a while to get right - "when it's done")... Right now he's just a polite version of Derek Smart.
If I wanted to stare at pretty pictures, I'd buy... A movie. Frankly, Hoyle hasn't done anything to improve their gaming engine in 40 years or so.
4.1 is also fairly new and last I checked, the mysql team is only recently recommending it for production use. If the PHP native API takes advantage of it now, people need to convert the bales of old broken code to use it.
I apologize that my statement is out of date.
Then again, I have a mysql shirt that's proud of having transactions. I got this last june at OSCON, from the mysql team... it's a great piece of humor.
It's good to see they're finally doing what other database servers have been doing for almost (over?) 10 years now.
Still eagerly awaiting sequences that aren't attached to a table. Maybe then I can consider it more than a toy.
You know what shared memory is, right? I think the "shared" word might give you some clues.
Also as you note, you're saturating your server far beyond it's limits having over 4 times the amount of requests it can serve (I imagine that's actually serves and not theoretical B.S. - it takes longer to service a modem connection and will bog down your server doing so).. You might have considerable luck increasing the MinSpareServers directive in a forked model. The forking is killing you, not the memory usage, although it sure looks like that. The faster you're able to respond to those requests (thus getting rid of them), the less memory you need. Personally I think you should be forking at least 5 above your "average to heavy" RPS.
At a company I used to work for, heavy oracle traffic would cause this phenom and we called it "the oracle death spiral". It was bad application usage and much harder to fix.
Eh, the only issue with the language is that there isn't really any enforcement going on to ensure that user input is being checked. I don't think that's a problem with the language, although I do think the efforts that the PHP team have put toward it are not in the right direction.
Taint checking is nice but as a person who's worked with in perl before, no less secure - I've seen lots of code that goes out of it's way to get rid of the taint check as soon as possible whether the data in the variables have been validated or not - and since this could be done by a "friendly" CPAN module, kind of defeats the point.
What many web programmers don't realize is two major things:
1) HTTP is stateless, and no matter how hard you try, state checking can be defeated
2) The user can send you anything they want.
#2 is pretty much par for the course with software - #1 makes controlling #2 a lot more important.
The average authentication mechanism used in a HTTP query is a joke - cookies with a "secure" hash seeded with an IP address or whatever that is validated against the client - which means half of your damn hash is already known and easily manipulated. This is the same reason that you tightly control how NIS or SMB is transmitted over the network.
A secure, transmittable hash needs 3 (although preferably 4) parts - an identifier that helps defeat the most basic replays, a "shared secret", and a "secret" - you might see that X.509 certs share a similar spec in the hostname validation of the cert, the 'shared secret' being the key signature, and the 'secret' being the private key, if you leave the CA out of the picture (which really isn't an option for web sites). A nice touch is to have a constant key to muddy the hash, or to mix up the order regularily - the 'secret' should be variable (it does't have to be long with a strong hash like SHA1) per client or user. It could be something like a date + non-transmitted user ID, ie., data you already have - and it HAS to be validated the same way it was created. People often screw up that important step.
I'm/pretty sure/ that PHP's session tracking does this, but I haven't looked into it enough. The goal is that even with all the trickery the most you can compromise is 1 account for a limited amount of time.
Sanitizing text is something that most people don't pay much attention and at best can be used to make your site look unprofessional. In every case where it's possible, use a strict heuristic for defining data and use a library - not a inline method, for checking data. If you're going to display the content AT ANY TIME in the future to ANYONE over the web, normalize it - it should be the rule, not the exception, to strip anything that looks like a HTML tag.
At a place I used to work, one guy found that HTML was allowed in the name entry of the customer - so in their brilliance they started having fun with it and injecting pictures - what's to keep an attacker from injecting something malicious to compromise the users of the web administration interface (that won't be as clued in) to capture input regarding other accounts or credit card numbers? You know, like posting that information to a dummy account's "reviews" or something silly like that - it doesn't have to be directed at a place he controls - after all, he more or less owns the web application at that point.
Or worse - the CMS editing suite. Like your CC number on the front page of your website? I bet the business would like it less than you would.
If all database input was filtered from HTML by default, this would never be an issue.
Binding is *not* escaping. magic_quotes_gpc is a hack through and through.
Binding is separating the input data from the calling convention not unlike how you do with a function call. Please do not confuse the two.
See my earlier post regarding mysql's lack of a binding API for why you might be confused.
The problem is that PHP's database support is just a slight transliteration of the underlying database API, and has no centralized API for dealing with things - if they'd just write their own or adopt the mature PEAR solution that's already out there (so that it is installed with all PHP installations), I think many PHP applications would not be having nearly as many problems.
To clarify, the prepared statement in DBD::mysql and the PEAR equivalent are emulated portions of the API, where they are (at least in DBI), actually used in the case of DBD::Oracle and (I think) Pg.
It is better than doing it yourself - the code will sanitize your binds and is less mistake-prone, but the client is doing it, not the server - that's the major problem there. Also, a binding API allows type checking at each execution, which is not possible without it.
actually, mysql doesn't have a binding API and doesn't support prepared statements - that is, a statement that has it's validity checked before execution is possible.
Compare to Oracle, which will fail when you prepare the statement, not execute it.
Slashdot Mirror
Economically, this stands to even things out, which is more in line with the socialist view and less in line with the capitalist view. Perhaps you should read the definitions of words before you use them?
As India and other countries start taking more jobs, and gaining ground in "mindshare" or respect or whatever you call it, and as their domain knowledge increases, they will raise rates to reflect their power in the marketplace. As India as a country grows economically, inflation will settle in compounding it further.
As soon as the cost-effectiveness of doing work overseas starts to fade, less work will be considered to go there - in reality, at that point it will be a battle of labor laws and how much control the outsourcing company has over the employees under those laws.
I think you can figure out the last part, which mostly revolves around the word "union". People who work hard to obtain the knowledge they have do not like to get stepped on.
Of course, any good socialist already knows that capitalists are just tools to help socialism grow anyways.
If you want a concrete example, try to make money on rentacoder.com - good luck with that.
Actually, and I know this is going to get some people to bellow loudly, Word and vi have more in common than Word and Emacs. Like I said, those who compare the two are completely missing the point.
Word and vi are editors that have capabilities to automate certain editing features.
Emacs on the other hand is a lisp environment with editing features.
It's a small and mostly semantic difference but an important one - the reason people have email and IRC clients in emacs and not in the former two have a lot to do with this difference, and understanding WHY it's important is key to understanding why so many people choose to develop applications like this in emacs. Seriously, it's less about fandom and more about flexibility.
Writing an equivalent Common Lisp environment for your mail client, or your IRC client is a lot more complex. Emacs provides the framework and allows for all of that.
I have no use for an IRC client that has that much power, but if VM had better IMAP support, I wouldn't be swearing at Mail.app all the time.
No, that would require acting in iambic pentameter.
Anyone who compares Word to Emacs misses the entire point of how emacs works and where it's power lies.
Seriously, that's pretty damn ignorant.
But it's got fancy hotkeys and visual editing, right?
Jebediah, Bob, they're on our turf! Break out the shotguns, bourbon and banjos!
/good/ thing as long as the playing field is fair? This doesn't sound unfair to me in the slightest.
Seriously folks, don't you realize that competition historically has been a
You must not think foreign keys are useful, either. Right?
The 'R' is RDBMS means 'relational'. Without the ability to establish real, atomically produced relations between tables, you're just using a toy.
I was talking about the video game.
I mean, it's too much to ask to read the article, it's too much to read the posts, is it too much to ask these days that the context is at least right before you reply with idiocy?
Heh. I think the last MM I actually beat was 3 as well.
/wanted/ to like viewtiful joe, I just couldn't get past the fact that I was just playing yet another MegaMan with a different character, and the game was 4 times as hard for no apparent reason.
Capcom is really good when they get it right - they are also really bad when they get it wrong, Not unlike Konami.
I
Heh oh yeah, 9, 8, and X-2 are the only ones I haven't beaten, and the latter two I played for quite some time before I quit in disgust.
If you haven't played the early series, go out and find roms of them - I think considering what you found interesting, even though the graphics suck by today's standards, the story is well worth it.
I think you're a little more flexible than most, but I can understand your viewpoint.
FF7, as I'm sure you can attest to, was less running of scene to scene and more about exploration. The original series pre-7 was the same way. They did start to cross over at that point and I do agree, 7 is not great but I think it's better than where they have gone with it.
FFX had an awesome story, and even a great way to extend the job system in 5 further, but I just couldn't get past the fact that I really had no room to explore - pointing the airship at destinations with a menu just doesn't feel the same as flying it around, and that's a good portion of my dilemma.
X-2 is such a deviation off the radar that I can't help but think that people who like it are merely fans of the name and not the actual content of the series, and more or less, it's original formula. While the content has always differed, the formula took quite a bit of time to morph into what it is now.
Frankly, the online FFXI seems to follow the original formula much, much closer than anything released after it, or in close proximity prior.
Ah, slashdot. The last bastion of good old school gamer in-joke humor.
:)
Probably the only reason I still come back.
Does the virus wipe away giant blocks in the screen as you run the stylus over it? Be sure to follow the right pattern to avoid those holes!
I have to disagree. My cell phone is more than my phone.
You lost the argument right here, bud.
You realize that almost everything you said you qualified with "low quality", right?
If you need a simple computer, buy a simple computer. Don't buy a phone with a simple computer, buy a simple computer with a phone.
Sounds the same but they are very different things.
I was coming to post the same thing.
Seriously folks, it's a phone. If you want a computer, there are tons of other options that will do everything you need, and not have this problem.
Actually, I find that sequels, the longer they extend, cater to a different audience than the one that made them famous in the first place:
Final Fantasy X-2. Heck, anything after 7. It's crap. All crap. In 7, the cutscenes are minor, like in the older games, and drive the story in ways that can't be driven interactively.
8 and on seem like an exercise in doing what is necessary to get to the next cutscene. I couldn't force myself to digest all of X-2, after 8 and X, it was enough. I haven't played 9, but I'm tempted to considering all the people that like 8, X, and X-2 say it sucks. These people coincidentally hate 7 and haven't played anything earlier.
I remember buying FF1 the day it came out in the U.S. and being AWE STRICKEN by a stupid blue screen with text that faded in.... Between the music and the contents of that text, I was in love immediately.
Now I get something that probably took 5 months to render that amounts to a very well-done computer-generated britney spears video. Strangely, that stupid blue screen was much more interesting.
.. with a Killer Instinct.
I figure Virtua Fighter is inferred.
I hear ya... Like I said earlier, I liked Viewtiful Joe when it was.. Mega Man. Anyone who's familiar with Capcom's history knows that these games are the only thing they've ever been good at, and I wouldn't say that's a good thing... Realistically, SF2 and MM were the only successful genres, and the original releases (although I am partial to MM2) were the best.
I'm personally waiting for Viewtiful Joe: Hyper Super Extended Championship Edition vs SNK and Marvel in a Primal Rage Pit Fight while practicing the Art of Fighting to engage in Mortal Kombat.
Do people buy shrek 2? Likewise, do people buy copies of the movie Tomb Raider in their home? Was it a best seller at the box office?
... did not follow their normal protocol. Since then it seems they have taken to farming off the good games to developers who still know how to release good games, like Bioware.
LucasArts is probably the only exception to the rule in movie tie-ins, and that's because most of their games are actually good - X-Wing and Tie Fighter, even without branding would have been awesome flight sims. I never got into JK, but I could see the potential. I think it has a lot to do with the fact that they aren't blindly following the movie plot and rushed to release at the same time as the movie. That said, Pod Racer was a giant pile of poo and
I think you're creating a red herring.
Just to be clear, Viewtiful Joe was a lot more fun when it was Mega Man, when I was 8.
Those are just titles, which really detracts from your point.
... right? Molyneux needs to figure out if he wants to be Will Wright (mundane ideas that are great, easy to get out the door) or Sid Meier (complex ideas that take a while to get right - "when it's done")... Right now he's just a polite version of Derek Smart.
Fable, of those, seems to be the closest to an "original" idea... Although I find myself stretching even there. You won't be finding a tempest or even a tetris in that crop. All the games you describe are more or less the equivalent of card games - even though you're playing something with different rules, you're playing with the same medium - cards.
This isn't always bad, but don't confuse titles with a new story with an original idea. Personally I think HL2 is a better game than all of them and it's the only one that I played all the way through, and I think only fable and the first one (which I honestly haven't heard of) are the only ones I haven't bought - because Molyneux hasn't completed anything since Populous 1, just broken visions - if he had taken his ideas from Fable, incorporated them into Black & White, we'd have something (since they are basically the same game but with different perspectives).. But then everyone would say the graphics blow as they'd be circa 1999 - and as we all know, graphics and cut scenes really enhance the gameplay, which is why you bought the damn thing in the first place
If I wanted to stare at pretty pictures, I'd buy... A movie. Frankly, Hoyle hasn't done anything to improve their gaming engine in 40 years or so.
4.1 is also fairly new and last I checked, the mysql team is only recently recommending it for production use. If the PHP native API takes advantage of it now, people need to convert the bales of old broken code to use it.
I apologize that my statement is out of date.
Then again, I have a mysql shirt that's proud of having transactions. I got this last june at OSCON, from the mysql team... it's a great piece of humor.
It's good to see they're finally doing what other database servers have been doing for almost (over?) 10 years now.
Still eagerly awaiting sequences that aren't attached to a table. Maybe then I can consider it more than a toy.
You know what shared memory is, right? I think the "shared" word might give you some clues.
Also as you note, you're saturating your server far beyond it's limits having over 4 times the amount of requests it can serve (I imagine that's actually serves and not theoretical B.S. - it takes longer to service a modem connection and will bog down your server doing so).. You might have considerable luck increasing the MinSpareServers directive in a forked model. The forking is killing you, not the memory usage, although it sure looks like that. The faster you're able to respond to those requests (thus getting rid of them), the less memory you need. Personally I think you should be forking at least 5 above your "average to heavy" RPS.
At a company I used to work for, heavy oracle traffic would cause this phenom and we called it "the oracle death spiral". It was bad application usage and much harder to fix.
Eh, the only issue with the language is that there isn't really any enforcement going on to ensure that user input is being checked. I don't think that's a problem with the language, although I do think the efforts that the PHP team have put toward it are not in the right direction.
/pretty sure/ that PHP's session tracking does this, but I haven't looked into it enough. The goal is that even with all the trickery the most you can compromise is 1 account for a limited amount of time.
Taint checking is nice but as a person who's worked with in perl before, no less secure - I've seen lots of code that goes out of it's way to get rid of the taint check as soon as possible whether the data in the variables have been validated or not - and since this could be done by a "friendly" CPAN module, kind of defeats the point.
What many web programmers don't realize is two major things:
1) HTTP is stateless, and no matter how hard you try, state checking can be defeated
2) The user can send you anything they want.
#2 is pretty much par for the course with software - #1 makes controlling #2 a lot more important.
The average authentication mechanism used in a HTTP query is a joke - cookies with a "secure" hash seeded with an IP address or whatever that is validated against the client - which means half of your damn hash is already known and easily manipulated. This is the same reason that you tightly control how NIS or SMB is transmitted over the network.
A secure, transmittable hash needs 3 (although preferably 4) parts - an identifier that helps defeat the most basic replays, a "shared secret", and a "secret" - you might see that X.509 certs share a similar spec in the hostname validation of the cert, the 'shared secret' being the key signature, and the 'secret' being the private key, if you leave the CA out of the picture (which really isn't an option for web sites). A nice touch is to have a constant key to muddy the hash, or to mix up the order regularily - the 'secret' should be variable (it does't have to be long with a strong hash like SHA1) per client or user. It could be something like a date + non-transmitted user ID, ie., data you already have - and it HAS to be validated the same way it was created. People often screw up that important step.
I'm
Sanitizing text is something that most people don't pay much attention and at best can be used to make your site look unprofessional. In every case where it's possible, use a strict heuristic for defining data and use a library - not a inline method, for checking data. If you're going to display the content AT ANY TIME in the future to ANYONE over the web, normalize it - it should be the rule, not the exception, to strip anything that looks like a HTML tag.
At a place I used to work, one guy found that HTML was allowed in the name entry of the customer - so in their brilliance they started having fun with it and injecting pictures - what's to keep an attacker from injecting something malicious to compromise the users of the web administration interface (that won't be as clued in) to capture input regarding other accounts or credit card numbers? You know, like posting that information to a dummy account's "reviews" or something silly like that - it doesn't have to be directed at a place he controls - after all, he more or less owns the web application at that point.
Or worse - the CMS editing suite. Like your CC number on the front page of your website? I bet the business would like it less than you would.
If all database input was filtered from HTML by default, this would never be an issue.
Binding is *not* escaping. magic_quotes_gpc is a hack through and through.
Binding is separating the input data from the calling convention not unlike how you do with a function call. Please do not confuse the two.
See my earlier post regarding mysql's lack of a binding API for why you might be confused.
The problem is that PHP's database support is just a slight transliteration of the underlying database API, and has no centralized API for dealing with things - if they'd just write their own or adopt the mature PEAR solution that's already out there (so that it is installed with all PHP installations), I think many PHP applications would not be having nearly as many problems.
To clarify, the prepared statement in DBD::mysql and the PEAR equivalent are emulated portions of the API, where they are (at least in DBI), actually used in the case of DBD::Oracle and (I think) Pg.
It is better than doing it yourself - the code will sanitize your binds and is less mistake-prone, but the client is doing it, not the server - that's the major problem there. Also, a binding API allows type checking at each execution, which is not possible without it.
actually, mysql doesn't have a binding API and doesn't support prepared statements - that is, a statement that has it's validity checked before execution is possible.
Compare to Oracle, which will fail when you prepare the statement, not execute it.