Well, maybe it's just impossible for nothing to exist:-)
Seriously. Try to imagine nothing existing. No other universes, no physics, no mathematics. No such thing as 1 + 1 = 2. Nothing to even concieve of the concept. No such thing as existence. It's impossible to stretch your mind around the concept.
I guess I can't really back this up with anything, but intuitively, to me, it just seems impossible for the universe to not be infinite. Because then what would be outside the universe? I mean before/after/all around it? I mean, what is the universe 'embedded' in?
Maybe it's only beause of the way our brains are constructed that we can't hold in our heads the idea that the universe is embedded in 'nothing'. But if you try to think about it, it's just counter-intuitive that 'nothing' could exist at all, seeing as how something other than nothing exists.
Either there's nothing, or there's everything, and there's not nothing.
So from the outside, you'd look like you were living a loooong time, but from your point of view, everything around you (if you had enough energy to detect it at all) would constantly be speeding up. And you wouldn't actually feel like you were living forever. At some point, you'd have to have a last thought. Like, "Wow. Thi s i s r e a l l y b o r i--" dead.
Tee-hee! Your code looked a little bit familiar to me;-)
Well this seems to happen all the time, already. For instance, in an OO language like Ruby, you might have an object with a certain method that you want to call, and you know that you want to call that partular method. But when you say "obj.amethod", the VM'll have to go through and find the method called "amethod", depending on what type of object it is, even if you know when you call it exactly which chunk of code you wanted to call, and this is what allows 'a_int.to_s' to behave differently than 'a_rock.to_s' (what's this called? a "virtual function"?). So this is just a higher level of abstraction. You didn't have that in C. I think what he's talking about is just abstracting more.
What would be nice would be if you knew exactly what you were doing, you could give the machine more explicit instructions to make it more efficient, but if you weren't, you could just throw something fuzzy at it. Like in C++ where you can specify if a function is virtual or not.
Beh. Well I agree with you. What I don't like is the lack of pass-by-reference. After working with Java, Python, and Ruby (yum, yum), I find a lot of PHP's 'features' a little bit annoying.
The problem isn't that they're not pretty. KDE and Gnome are slow and bloated, and having to interface through X doesn't help much. I'll take icewm or windowmaker over KDE/Gnome anyday.
You should accomplish your layout with DIVs. Tables are supposed to be meta-containers, not exact-layout positioning tools.
In a perfect world, this would be true. I tried. I mean I really tried to not use tables for non-tabular data, but if I used tables, my page showed up just like I wanted it to, and was more tolerant of different browsers than it was when I tried to use floating DIVs (looks best in links:-)
I might be able to use floating DIVs if they could be 'contained', and not overflow into sections of the page where I don't want them. But If I have something like:
<div> <div class="box" style="float:right"> Lots of text </div> Small amount of text </div> <div class="box"> text of box </div>
Then the box in the upper right overlaps the lower one, which is not what I want. But the table:
<table><tr><td> Small amount of text </td><td> Lots of text </td></tr></table>
Does exactly what I want. See my webpage if you're curious. (Or if you know how to do this right;-)
I don't like using tables for this kind of thing, but they do work better than anything else for what I want to do, as far as I can tell.
Maybe if there was a 'layout' flag, to tell the browser that the table isn't really displaying tabular data, but is only being used for layout purposes? <table layout>blah</table> Then all would be well.:-P
Re:These aren't really problems.
on
Yahoo Moving to PHP
·
· Score: 2, Insightful
It's also pretty annoying when you don't read the manual - and this one is NOT obscure to find. Section 8 - "Variables"
OK. I thought I looked at that page. I guess I only looked at the one about superglobal arrays, though. But you don't have to be a jerk about it. And it still doesn't explain why that is done, aside from saying that you need the slashes when doing database queries. Most of the time you're not doing database queries. It seems rather arbitrary to add slashes to user-inputted strings but nothing else. Why not add slashes when you read a line from a file, too? And I realise that you can turn this behavior off in the configuration, but then your scripts won't be portable, since it is the default behavior.
/*...This is very inefficient. */
Then design a better language yourself that has nifty things like variable variables (echo $$b[0]->foo;) and is still faster than most other languages
What's that got to do with re-parsing pages? Python can use variable variable names, and it always compiles included files.
Or just get a host that will use one of the many caching products available (zend, ioncube, etc)
If PHP itself was little more clever about caching compiled scripts, I wouldn't need to do that. Although I agree that this isn't really a problem in that you can't do it as much as because you have to do extra work.
/* 3) It's a real pain to include scripts that are in different directories which include other scripts, because they will try to open them relative to the location of the original page. */
Seems pretty damn logical to me. Of course, PHP will also look for files in the 'include_path' which is set in the php.ini file, so it's really looking in multiple places. And it wouldn't kill you to just use a predefined constant like DOCUMENT_ROOT and include files relative to that so your scripts would be portable and a bit easier to move around internally if you need to.
That's almost exactly what I do, but as I said, it seems like a hack. There should be a way to include files relative to the current file.
Anyway, whether you think these are really problems or not, they are annoyances for me (along with other issues such as no real pass by reference), and they didn't need to be there.
Well... I don't know if someone else has already pointed these things out, and I don't have time to go through and read all 6800 posts, so I'll just voice a few complaints about PHP.
I use PHP because that's all that's available on my web hosting service other than cgi. I could use the cgi-bin but then all my URLs would start with cgi-bin/, which would suck.
PHP is nice for doing database query stuff, and it's nice for simplifying your pages (for instance if there's a standard block of HTML you want in your pages and don't want to have to copy it all the time or re-compile a bunch of static HTML files), but as I've used it, I've come across a few problems.
1) This one is really annoying. Certain pre-defined variables (I'm not sure which, maybe only user-inputted data) are pre-slashed. So if a user inputs the string 'My name is "Jon"', you get it as 'My name is \"Jon\"'. WTF is up with that? That's not what he said! I can't find the reason for this or anything else about it in the documentation (maybe it's there somewhere, but I can't find it).
2) Every time a page is loaded, it has to be re-parsed, as well as any included scripts that it uses. This is very inefficient.
3) It's a real pain to include scripts that are in different directories which include other scripts, because they will try to open them relative to the location of the original page.
I (and it seems other people do, too, guessing from the user-submitted documentation on PHP.net)have to use nasty hacks to get around these problems all the time.
As far as alternatives go, right now I'm a big fan of Python.
It has lots of useful libraries, and the language is very elegant (sp?). The one thing I don't like about it is that every time you access a variable, it has to look up the name in a dictionary (I think).
Unfortunately I couldn't get mod_python to work. I'm writing my own web server (in Python), instead, so that I don't have to deal with all of the Apache/PHP problems I've had. Looked at Medusa, but that seemed a little too complicated for what I was doing.
I actually don't care much about the syntax of the language I use (as long as it's not too unreasonable). What's important to me is that I can use an elegant, efficient runtime in which I don't pay for things I don't use, and that I get a useful standard library. I think I'll be happier when I can use mod_parrot:-D
I'm still stuck with PHP on my real site, though (as opposed to my personal box that I play with).
(That's TSDF2, for you. Easier to read, write than XML, and uses less bandwidth. Parse it in 40 lines of PHP, or less if you're clever:-)
I think the problem is that XML is a markup language, whereas what people really want is a structured data language. XML is good for marking up text, but pretty lousy for what most people want to do with it.
So put a fat guy on the bike who needs the excersize anyway. Now you're using energy that would have been wasted. You seem to be forgetting that the humans are going to be breathing whether you burn extra fuel or not. Just goes to show the short-sightedness of you.
At my University (UW-Platteville) you're basically not allowed to do anything with the network except web, email, and IM. This is nothing. If they only told me I couldn't run napsterish stuff, that would be fine, but we're not even allowed to run Linux while connected to the network. That's why I had to get a second box - Linux runs on the old one, and I telnet to it from the Win98 box which is on their network and also shares a drive. It's actually a pretty cute setup, though...
Heh, that's interesting. I remember there was this one thing on Sesame Street where there was this guy out counting stars and he went up to fourty and then just said "and many more". For a long time I actually thought fourty was some sort of upper limit to counting...
Take an old popcorn bucket from the movie theater, pop the pocorn into it, then spray on some canola oil, salt, shake around, and spray/salt one or two more times.
It's hardly a 'recipe', but then nothing I eat ever comes from anything more complicated than that. (ramen, potato soup with bread and cheese, etc)
I'd think that if you want this to be a low-bandwidth interface, XML-RPC is about the last thing you'd want to use. What's wrong with good old fasioned URL-formatted parameters?
I mean,
<?xml version="blah"?>
<methodCall>
<methodName>eat_cheese</methodName>
<params>
<param>
<name>amount</name>
<value>lots</value>
</param>
</params>
</methodCall>
just seems like overkill to me when you could just do:
Slashdot Mirror
Well, maybe it's just impossible for nothing to exist :-)
Seriously. Try to imagine nothing existing. No other universes, no physics, no mathematics. No such thing as 1 + 1 = 2. Nothing to even concieve of the concept. No such thing as existence. It's impossible to stretch your mind around the concept.
I guess I can't really back this up with anything, but intuitively, to me, it just seems impossible for the universe to not be infinite. Because then what would be outside the universe? I mean before/after/all around it? I mean, what is the universe 'embedded' in?
Maybe it's only beause of the way our brains are constructed that we can't hold in our heads the idea that the universe is embedded in 'nothing'. But if you try to think about it, it's just counter-intuitive that 'nothing' could exist at all, seeing as how something other than nothing exists.
Either there's nothing, or there's everything, and there's not nothing.
Blah.
*brain implodes*
So from the outside, you'd look like you were living a loooong time, but from your point of view, everything around you (if you had enough energy to detect it at all) would constantly be speeding up. And you wouldn't actually feel like you were living forever. At some point, you'd have to have a last thought. Like, "Wow. Thi s i s r e a l l y b o r i--" dead.
// Why have a computer guess at it when we KNOW?
;-)
Tee-hee! Your code looked a little bit familiar to me
Well this seems to happen all the time, already. For instance, in an OO language like Ruby, you might have an object with a certain method that you want to call, and you know that you want to call that partular method. But when you say "obj.amethod", the VM'll have to go through and find the method called "amethod", depending on what type of object it is, even if you know when you call it exactly which chunk of code you wanted to call, and this is what allows 'a_int.to_s' to behave differently than 'a_rock.to_s' (what's this called? a "virtual function"?). So this is just a higher level of abstraction. You didn't have that in C. I think what he's talking about is just abstracting more.
What would be nice would be if you knew exactly what you were doing, you could give the machine more explicit instructions to make it more efficient, but if you weren't, you could just throw something fuzzy at it. Like in C++ where you can specify if a function is virtual or not.
I think Perl6 is supposed to address this issue.
Well they're probably not using water from the ocean. There's plenty of that. Fresh water is a bit harder to come by.
Beh. Well I agree with you.
What I don't like is the lack of
pass-by-reference. After working with
Java, Python, and Ruby (yum, yum),
I find a lot of PHP's 'features' a little bit annoying.
Bah. When I saw Content-Transfer I thought some idiot was trying to say he had a patent on an HTTP headers or some dumb thing. WTF?
Puppeteers rule :)
Then again, there is a poster with the name of John Booty (a lousy DB with the Jets and Eagles in the 90's - one of Rich Kotite's boys).
What do the Jets and Eagles use databases for?The problem isn't that they're not pretty. KDE and Gnome are slow and bloated, and having to interface through X doesn't help much. I'll take icewm or windowmaker over KDE/Gnome anyday.
But you're stuck in a tube, instead.
You should accomplish your layout with DIVs. Tables are supposed to be meta-containers, not exact-layout positioning tools.
:-)
;-)
:-P
In a perfect world, this would be true. I tried. I mean I really tried to not use tables for non-tabular data, but if I used tables, my page showed up just like I wanted it to, and was more tolerant of different browsers than it was when I tried to use floating DIVs (looks best in links
I might be able to use floating DIVs if they could be 'contained', and not overflow into sections of the page where I don't want them. But If I have something like:
<div>
<div class="box" style="float:right">
Lots of text
</div>
Small amount of text
</div>
<div class="box">
text of box
</div>
Then the box in the upper right overlaps the lower one, which is not what I want. But the table:
<table><tr><td>
Small amount of text
</td><td>
Lots of text
</td></tr></table>
Does exactly what I want. See my webpage if you're curious. (Or if you know how to do this right
I don't like using tables for this kind of thing, but they do work better than anything else for what I want to do, as far as I can tell.
Maybe if there was a 'layout' flag, to tell the browser that the table isn't really displaying tabular data, but is only being used for layout purposes? <table layout>blah</table> Then all would be well.
It's also pretty annoying when you don't read the manual - and this one is NOT obscure to find. Section 8 - "Variables"
OK. I thought I looked at that page. I guess I only looked at the one about superglobal arrays, though. But you don't have to be a jerk about it. And it still doesn't explain why that is done, aside from saying that you need the slashes when doing database queries. Most of the time you're not doing database queries. It seems rather arbitrary to add slashes to user-inputted strings but nothing else. Why not add slashes when you read a line from a file, too? And I realise that you can turn this behavior off in the configuration, but then your scripts won't be portable, since it is the default behavior.
Then design a better language yourself that has nifty things like variable variables (echo $$b[0]->foo;) and is still faster than most other languages
What's that got to do with re-parsing pages? Python can use variable variable names, and it always compiles included files.
Or just get a host that will use one of the many caching products available (zend, ioncube, etc)
If PHP itself was little more clever about caching compiled scripts, I wouldn't need to do that. Although I agree that this isn't really a problem in that you can't do it as much as because you have to do extra work.
Seems pretty damn logical to me. Of course, PHP will also look for files in the 'include_path' which is set in the php.ini file, so it's really looking in multiple places. And it wouldn't kill you to just use a predefined constant like DOCUMENT_ROOT and include files relative to that so your scripts would be portable and a bit easier to move around internally if you need to.
That's almost exactly what I do, but as I said, it seems like a hack. There should be a way to include files relative to the current file.
Anyway, whether you think these are really problems or not, they are annoyances for me (along with other issues such as no real pass by reference), and they didn't need to be there.
/* no exceptions */
Ah. You're a fan of using Perl, then...
Oh. Nevermind.
Well... I don't know if someone else has already pointed these things out, and I don't have time to go through and read all 6800 posts, so I'll just voice a few complaints about PHP.
I use PHP because that's all that's available on my web hosting service other than cgi. I could use the cgi-bin but then all my URLs would start with cgi-bin/, which would suck.
PHP is nice for doing database query stuff, and it's nice for simplifying your pages (for instance if there's a standard block of HTML you want in your pages and don't want to have to copy it all the time or re-compile a bunch of static HTML files), but as I've used it, I've come across a few problems.
1) This one is really annoying. Certain pre-defined variables (I'm not sure which, maybe only user-inputted data) are pre-slashed. So if a user inputs the string 'My name is "Jon"', you get it as 'My name is \"Jon\"'. WTF is up with that? That's not what he said! I can't find the reason for this or anything else about it in the documentation (maybe it's there somewhere, but I can't find it).
2) Every time a page is loaded, it has to be re-parsed, as well as any included scripts that it uses. This is very inefficient.
3) It's a real pain to include scripts that are in different directories which include other scripts, because they will try to open them relative to the location of the original page.
I (and it seems other people do, too, guessing from the user-submitted documentation on PHP.net)have to use nasty hacks to get around these problems all the time.
As far as alternatives go, right now I'm a big fan of Python. It has lots of useful libraries, and the language is very elegant (sp?). The one thing I don't like about it is that every time you access a variable, it has to look up the name in a dictionary (I think).
Unfortunately I couldn't get mod_python to work. I'm writing my own web server (in Python), instead, so that I don't have to deal with all of the Apache/PHP problems I've had. Looked at Medusa, but that seemed a little too complicated for what I was doing.
I actually don't care much about the syntax of the language I use (as long as it's not too unreasonable). What's important to me is that I can use an elegant, efficient runtime in which I don't pay for things I don't use, and that I get a useful standard library. I think I'll be happier when I can use mod_parrot :-D
I'm still stuck with PHP on my real site, though (as opposed to my personal box that I play with).
I think Kelvins are different because they start at zero, so the number of Kelvins is directly proportional to the temperature.
What's 2 farenheits? That's not an absolute temperature, just a relative one.
Amen. Here's a virtual "insightful" mod point for you. On with the XML bashing!
:-)
An easy-to-parse text-based format (non-xml) could be good also:
fruit: {
type: "apple"
density: 5
worms: [
"randy"
"susan"
"george"
]
}
(That's TSDF2, for you. Easier to read, write than XML, and uses less bandwidth. Parse it in 40 lines of PHP, or less if you're clever
I think the problem is that XML is a markup language, whereas what people really want is a structured data language. XML is good for marking up text, but pretty lousy for what most people want to do with it.
So put a fat guy on the bike who needs the excersize anyway. Now you're using energy that would have been wasted. You seem to be forgetting that the humans are going to be breathing whether you burn extra fuel or not. Just goes to show the short-sightedness of you.
Save it for when you need it, or
use it to make plastic or something.
At my University (UW-Platteville) you're basically not allowed to do anything with the network except web, email, and IM. This is nothing. If they only told me I couldn't run napsterish stuff, that would be fine, but we're not even allowed to run Linux while connected to the network. That's why I had to get a second box - Linux runs on the old one, and I telnet to it from the Win98 box which is on their network and also shares a drive. It's actually a pretty cute setup, though...
Heh, that's interesting. I remember there was this one thing on Sesame Street where there was this guy out counting stars and he went up to fourty and then just said "and many more". For a long time I actually thought fourty was some sort of upper limit to counting...
Oh, hah! I say .1 compared to .13 and .25 and thought .01. Weird.
Oh yeah, just throw a [soy] hot dog in the pan with the noodles and it makes a half-decent meal.
That sounded like I was being sarcastic. No. I do that all the time. I dont put any Cheese Whiz or Miracle Whip on it, tho.
Amen to that!
Take an old popcorn bucket from the movie theater, pop the pocorn into it, then spray on some canola oil, salt, shake around, and spray/salt one or two more times.
It's hardly a 'recipe', but then nothing I eat ever comes from anything more complicated than that. (ramen, potato soup with bread and cheese, etc)
And plenty of green tea and diluted coolaid!
I'd think that if you want this to be a low-bandwidth interface, XML-RPC is about the last thing you'd want to use. What's wrong with good old fasioned URL-formatted parameters?
I mean,
<?xml version="blah"?>
<methodCall>
<methodName>eat_cheese</methodName>
<params>
<param>
<name>amount</name>
<value>lots</value>
</param>
</params>
</methodCall>
just seems like overkill to me when you could just do:
action=eat_cheese&amount=lots
I guess I'm just behind the times...
This seems more a way to hook up a Gameboy to an MP3 player than the other way around, then.
What do you need the Gameboy for if you're just going to hook it up to your own speakers?