> Just having toggles to enable/disable javascript and java near the address bar would be excellent. Same with disabling plugins like flash. Just so you don't have to go through menus and preference boxes to do it every time...
> Can anyone justify Redhat charging for something they're not able to legally charge for?
Bzzzzzzzzt! There is absolutely nothing in the licensing prohibiting you from selling it. If I wanted to, I could put out a Zontar-O-Matic UltraHypaSoopaDoopaLotsORichChocklityGoodnessLinux distro tomorrow, and charge $10,000 a copy, just so long as I made the source code available.
(Whether anybody would actually pay that much for it is another matter entirely.;-) )
> I seem to remember being bitten a while back by the way IE's JavaScript implementation allowed array access with (), while Netscape required the standard [].
Oops, I forgot about that.
(I learned JS first, so I never had that problem.)
There are some additional differences in ASP, like where you have to use enumerators and the item() method to loop through what are implemented in VBScript as arrays, too, IIRC.
> JavaScript is also an ECMA standard, and we all know how well it from browser to browser.
Bad example. Mozilla/Netscape let you access characters within strings using array notation and methods defined internally to classes as public methods, whereas Microsoft's implementation doesn't let you do these things. Those are the just about the only major differences in their JavaScript implementations. Otherwise, they both do an excellent job of supporting ECMA-262.
So... you use charAt() and remember to bind all class members that you want regarded as public to prototypes, and move on.
The various DOM implementations are a somewhat different story. However, there's now a higher degree of convergence between the version 6+ browsers in this regard than there's been since before NS 4 and MSIE 4 came out, and it's now possible to do quite a lot of interesting and/or useful stuff that works the same in NS, IE, and Opera without doing any branching whatsoever. (Particularly if you ignore NS 4 and MSIE 4, which I've been doing for about the last year and a half.) Some of the event-handling is a bit different, and so is retrieving computed style information. Hopefully MS will add support for the W3C ways of doing these things eventually.
At any rate, don't confuse JavaScript language implementation with object model support. That's like saying that GTK and QT implement C++ differently.
(BTW, JavaScript also supports lambda functions, and you can nest them as deep as you want, which IMO is one of the insanely cool things about it -- particularly where event handling is concerned.)
We now return you to your regularly-scheduled flamefest.
So you're a marketer. Your profession is to interfere in market transactions to persuade participants to make irrational decisions in your favor. That's defective and antisocial behavior.
That's bloody brilliant. Too bad you posted as AC -- I'd have added you to my Friends list.
Do you honestly think that if a particular brand of automobile actually caused more accidents, that people wouldn't avoid it on their own, if for no other reason than the fact that insurance premiums would be outlandish?
Opera's Javascript implementation has been good for years. The problem is more with doing actual scripting in JavaScript. Internet Explorer and Mozilla both have very different "API"s for DOM scripting. Opera 6 was pretty poor in that regard - didn't render much. Opera 7 renders about 90%, maybe more, of either Mozilla or Internet Explorer's JavaScript, depending on which browser string you send (identify as Internet Explorer and pretty much every IE-specific pages render perfectly)
When identifying as Opera, usually only the most IE-centric pages won't render.
Opera's JS console sucks ass. It's actually worse than MSIE's.
The problem, however, is not so much writing JS for Opera as it is the crap scripts that get written in general.
Read this next sentence very carefully: Clientside scripting should NOT depend on UA strings -- you should do object detection, which can't be fooled by UA spoofing in the first place. The browser either supports the required object/method/property, or it doesn't. ID'ing Mozilla as MSIE won't fix scripts that depend upon document.all, for example. Clueful scripters that actually think about what they're doing do NOT write "Mozilla JavaScript" or "MSIE JavaScript" -- they write to DOM in which case about 90% of cross-browser problems vanish. (Except for MSIE's crap event model and a few other things here and there, which Opera unfortunately tends to copy.)
These things being said, I happen to like Opera a great deal. It's fast and it's CSS support kicks butt. Not as good as Mozilla's, but very nearly so (although you'd think it'd be BETTER than anyone else's since Hakon Lie runs the company -- figure that one out, I can't).
My one big beef with it is that it doesn't support a big chunk of CSS-DOM -- no document.styleSheets collection. And neither getComputedStyle() nor the MSIE pseudo-equivalents seem to work, either. Working around these issues is a major pain in the ass for the kind of stuff I've been doing lately, which involves a lot of changing style rules on the fly. It means I have to restrict myself to working with style rules written inside STYLE elements and work with their contents using string methods and regexps -- and I can't use linked stylesheets at all if I want this stuff to work with Opera. This is pretty silly, since this defeats the raison d'etre for using stylesheets in the first bloody place.
Opera's still way ahead of MSIE in actual CSS support, tho.
(The preceding message brought to you by The Campaign to End User-Agent Spoofing and User-Agent Detection.)
He's using it to suck up some hits on an Amazon.com banner ad, and he's not even mirroring the damn story. So he's NOT just an idiot. He's a low-life. "Method to my madness," my arse.
Word up: never follow any "mirror" links that point to martin-studio.com.
> The UK, Switzerland and Japan are way more Socialist than the US,
True. Also Australia (and it's quite nicely done, too).
> North Korea is Communist.
True.
> China is also Communist...
False. China calls itself "Communist" but in reality its economic system is now more accurately described as State Capitalism.
> The US doesn't have a real free market: subsidies, unfair tax breaks, and tarrifs make a mockery of the system.
True. The US has a "free market" only in the sense that the wealthiest individuals and corporations there are free to screw everyone else over. (Which is why I now prefer to live and raise my child in Australia, where she gets guaranteed medical care and education, thanks very much.)
Forget India for a moment..Would you feel it was a betrayal of the "middle class" if your company in New York outsourced its system administration functions to EDS in Kansas?
Funny you should mention that possibility...
I worked for a US credit card company (a Dean Whitter subsidiary) that did just that -- they shut down heir offices on the coasts and built two new Ops Centers, one in Northeast Tennessee, and one in South Dakota. Then they replaced all their $50k/yr+bennies city dwellers with $7/hr college students working part-time.
I heard a couple years ago that they shut down the center in Tennessee where I'd worked and outsourced those jobs to a firm in India... Don't know if the one in Sioux Falls is still going or not.
Wonder where they'll ship the jobs when it's time for the Indians to get screwed like we did, and the New Yorkers before us?
Agreed. Some moderators need to look up the word "redundant" because it's been obvious to me that some of them don't know what it means. Those who do know what it means need to be a little less hard-arsed about applying it. It's not fair to label a similar post made 2 minutes after the first one "Redundant" -- not all of us are sitting atop T-1's or type 300 words a minute, after all! If it's the 3rd or 4th such post, or it's made an hour or two after the first one, that might qualify as "Redundant".
Moderators, look at the posting times and please try to be more reasonable about this, won't you? Thanks.
The whole SCO business just seems unbelievable. Maybe I've missed something somewhere but the basic story seems to be:
1. SCO helps out Linux development in a few small ways -- some hardware here, a few lines of code there
2. SCO turns around and says it owns the other 99% of Linux it had nothing to do with because of the 1% that it did
3. SCO is now trying to extort licensing fees from end users because it "owns" Linux
4. This is somewhat akin to Bob's Auto Parts Factory saying that, because the wholesaler from which my mechanic purchased a very small part he used in repairing my car part didn't pay them for the part, I now must pay Bob rent on my car -- and furthermore, they won't tell either my mechanic or me which part it was, because it's a "trade secret". Hell, they won't even tell my mechanic which wholesaler he supposedly bought the part from!
Am I wrong? Did I miss something somewhere? Please feel free to correct me if I have. ----- (*I Am Not A Kernel Developer)
> Even if you could document.body.blah.svg.line1 = "blah" via JS wouldn't document.draw.circle(x,y,r,color) be easier? At the very least circle, line, rect, bitblt and such should be in the JS api.
Nup. That's what DOM is for, and it's language-neutral. Why would you want to encumber JavaScript with additional objects or methods that are peculiar to a specific host environment?
Of course, you're free to build up your own API in JavaScript (or any other language that implements DOM) using createElement(), setAttribute(), appendChild(), etc. if you wish.
Fine, just turn off JS then (which I'm willing to bet you already do).
But to pigeonhole SVG as "just another way to do animations" is to miss the point. SVG is NOT a plugin format in the sense that Flash, QT, WMP, etc. are. It's not binary. It's a markup language. You can do cool and useful real-time interactive stuff with it by styling it with CSS and scripting it with JavaScript/DOM. Don't want it to move or change? Turn off JS. Don't want to see the pretty colours or interesting shapes? Turn off CSS, or override it with your own user stylesheet.
Hell, use Lynx, if that's what you prefer -- the point is that SVG text content is still actual text. So it's not an all or nothing proposition like you have with the binary plugin formats.
> So maybe I'm just a web Luddite who wants plain old text and images, but if the Mozilla developers manage to put default SVG support in Firebird while keeping it small and fast it'll be a good thing, even if it's still a while before we see widespread use of SVG. As long as there's a runtime option to turn it off.;-)
So you'd rather have your bandwidth gobbled up by animated text gifs than to spare a couple of bytes for some actual text marked up with a few SVG tags + CSS?
> Unfortunately, you can't do this whilst serving your XHTML documents as text/html (which is the only way Internet Explorer will understand them properly).
And here I was thinking that IE was the browser that only paid attention to file extensions while blithely ignoring mimetypes. Silly me.
In any case, the proper way to handle this issue is by using namespaces, is it not?
Slashdot Mirror
> Just having toggles to enable/disable javascript and java near the address bar would be excellent. Same with disabling plugins like flash. Just so you don't have to go through menus and preference boxes to do it every time...
Mozilla + PrefBar = just that.
> All companies are the devil. Capitalism is the devil.
Unbridled, "we must must crush all others at any cost" capitalism is definitely unhealthy.
> You can actually download a *powertoy* from Microsoft...
:)
Try Virtual Dimension instead. It's GPL'ed and it's quite spiffy.
Cool. Have your girl call my girl and we'll do lunch. Don't forget your credit card.
> Can anyone justify Redhat charging for something they're not able to legally charge for?
x distro tomorrow, and charge $10,000 a copy, just so long as I made the source code available.
;-) )
Bzzzzzzzzt! There is absolutely nothing in the licensing prohibiting you from selling it. If I wanted to, I could put out a Zontar-O-Matic UltraHypaSoopaDoopaLotsORichChocklityGoodnessLinu
(Whether anybody would actually pay that much for it is another matter entirely.
About 90% of the posts on Mozilla and Bugzilla complaining about MNG being dropped appear to have come from just one person.
Now if you want to carp about SVG instead YARF*, I'm with ya 110%, mate.
-----
*(Yet Another Raster Format).
Ta very much for the interesting bit of archaeological info there, matey.
> I seem to remember being bitten a while back by the way IE's JavaScript implementation allowed array access with (), while Netscape required the standard [].
Oops, I forgot about that.
(I learned JS first, so I never had that problem.)
There are some additional differences in ASP, like where you have to use enumerators and the item() method to loop through what are implemented in VBScript as arrays, too, IIRC.
> JavaScript is also an ECMA standard, and we all know how well it from browser to browser.
Bad example. Mozilla/Netscape let you access characters within strings using array notation and methods defined internally to classes as public methods, whereas Microsoft's implementation doesn't let you do these things. Those are the just about the only major differences in their JavaScript implementations. Otherwise, they both do an excellent job of supporting ECMA-262.
So... you use charAt() and remember to bind all class members that you want regarded as public to prototypes, and move on.
The various DOM implementations are a somewhat different story. However, there's now a higher degree of convergence between the version 6+ browsers in this regard than there's been since before NS 4 and MSIE 4 came out, and it's now possible to do quite a lot of interesting and/or useful stuff that works the same in NS, IE, and Opera without doing any branching whatsoever. (Particularly if you ignore NS 4 and MSIE 4, which I've been doing for about the last year and a half.) Some of the event-handling is a bit different, and so is retrieving computed style information. Hopefully MS will add support for the W3C ways of doing these things eventually.
At any rate, don't confuse JavaScript language implementation with object model support. That's like saying that GTK and QT implement C++ differently.
(BTW, JavaScript also supports lambda functions, and you can nest them as deep as you want, which IMO is one of the insanely cool things about it -- particularly where event handling is concerned.)
We now return you to your regularly-scheduled flamefest.
Ever hear of something called an "SUV"?
Opera's JS console sucks ass. It's actually worse than MSIE's.
The problem, however, is not so much writing JS for Opera as it is the crap scripts that get written in general.
Read this next sentence very carefully: Clientside scripting should NOT depend on UA strings -- you should do object detection, which can't be fooled by UA spoofing in the first place. The browser either supports the required object/method/property, or it doesn't. ID'ing Mozilla as MSIE won't fix scripts that depend upon document.all, for example. Clueful scripters that actually think about what they're doing do NOT write "Mozilla JavaScript" or "MSIE JavaScript" -- they write to DOM in which case about 90% of cross-browser problems vanish. (Except for MSIE's crap event model and a few other things here and there, which Opera unfortunately tends to copy.)
These things being said, I happen to like Opera a great deal. It's fast and it's CSS support kicks butt. Not as good as Mozilla's, but very nearly so (although you'd think it'd be BETTER than anyone else's since Hakon Lie runs the company -- figure that one out, I can't).
My one big beef with it is that it doesn't support a big chunk of CSS-DOM -- no document.styleSheets collection. And neither getComputedStyle() nor the MSIE pseudo-equivalents seem to work, either. Working around these issues is a major pain in the ass for the kind of stuff I've been doing lately, which involves a lot of changing style rules on the fly. It means I have to restrict myself to working with style rules written inside STYLE elements and work with their contents using string methods and regexps -- and I can't use linked stylesheets at all if I want this stuff to work with Opera. This is pretty silly, since this defeats the raison d'etre for using stylesheets in the first bloody place.
Opera's still way ahead of MSIE in actual CSS support, tho.
(The preceding message brought to you by The Campaign to End User-Agent Spoofing and User-Agent Detection.)
He's using it to suck up some hits on an Amazon.com banner ad, and he's not even mirroring the damn story. So he's NOT just an idiot. He's a low-life. "Method to my madness," my arse.
Word up: never follow any "mirror" links that point to martin-studio.com.
Thank you for making my Foes list, dickhead.
> The UK, Switzerland and Japan are way more Socialist than the US,
True. Also Australia (and it's quite nicely done, too).
> North Korea is Communist.
True.
> China is also Communist...
False. China calls itself "Communist" but in reality its economic system is now more accurately described as State Capitalism.
> The US doesn't have a real free market: subsidies, unfair tax breaks, and tarrifs make a mockery of the system.
True. The US has a "free market" only in the sense that the wealthiest individuals and corporations there are free to screw everyone else over. (Which is why I now prefer to live and raise my child in Australia, where she gets guaranteed medical care and education, thanks very much.)
I worked for a US credit card company (a Dean Whitter subsidiary) that did just that -- they shut down heir offices on the coasts and built two new Ops Centers, one in Northeast Tennessee, and one in South Dakota. Then they replaced all their $50k/yr+bennies city dwellers with $7/hr college students working part-time.
I heard a couple years ago that they shut down the center in Tennessee where I'd worked and outsourced those jobs to a firm in India... Don't know if the one in Sioux Falls is still going or not.
Wonder where they'll ship the jobs when it's time for the Indians to get screwed like we did, and the New Yorkers before us?
Agreed. Some moderators need to look up the word "redundant" because it's been obvious to me that some of them don't know what it means. Those who do know what it means need to be a little less hard-arsed about applying it. It's not fair to label a similar post made 2 minutes after the first one "Redundant" -- not all of us are sitting atop T-1's or type 300 words a minute, after all! If it's the 3rd or 4th such post, or it's made an hour or two after the first one, that might qualify as "Redundant".
Moderators, look at the posting times and please try to be more reasonable about this, won't you? Thanks.
> I am not a Kraft Dinner???
;-)
Well, I'm not one of those, either... at least, I don't think so.
The whole SCO business just seems unbelievable. Maybe I've missed something somewhere but the basic story seems to be:
1. SCO helps out Linux development in a few small ways -- some hardware here, a few lines of code there
2. SCO turns around and says it owns the other 99% of Linux it had nothing to do with because of the 1% that it did
3. SCO is now trying to extort licensing fees from end users because it "owns" Linux
4. This is somewhat akin to Bob's Auto Parts Factory saying that, because the wholesaler from which my mechanic purchased a very small part he used in repairing my car part didn't pay them for the part, I now must pay Bob rent on my car -- and furthermore, they won't tell either my mechanic or me which part it was, because it's a "trade secret". Hell, they won't even tell my mechanic which wholesaler he supposedly bought the part from!
Am I wrong? Did I miss something somewhere? Please feel free to correct me if I have.
-----
(*I Am Not A Kernel Developer)
> Even if you could document.body.blah.svg.line1 = "blah" via JS wouldn't document.draw.circle(x,y,r,color) be easier? At the very least circle, line, rect, bitblt and such should be in the JS api.
Nup. That's what DOM is for, and it's language-neutral. Why would you want to encumber JavaScript with additional objects or methods that are peculiar to a specific host environment?
Of course, you're free to build up your own API in JavaScript (or any other language that implements DOM) using createElement(), setAttribute(), appendChild(), etc. if you wish.
> No, I don't want animated anything.
Fine, just turn off JS then (which I'm willing to bet you already do).
But to pigeonhole SVG as "just another way to do animations" is to miss the point. SVG is NOT a plugin format in the sense that Flash, QT, WMP, etc. are. It's not binary. It's a markup language. You can do cool and useful real-time interactive stuff with it by styling it with CSS and scripting it with JavaScript/DOM. Don't want it to move or change? Turn off JS. Don't want to see the pretty colours or interesting shapes? Turn off CSS, or override it with your own user stylesheet.
Hell, use Lynx, if that's what you prefer -- the point is that SVG text content is still actual text. So it's not an all or nothing proposition like you have with the binary plugin formats.
> ...wouldn't a series of graphic routines for javascript be far more useful?
No need.
XML-DOM + SVG namespace = pretty much what you're asking for.
> So maybe I'm just a web Luddite who wants plain old text and images, but if the Mozilla developers manage to put default SVG support in Firebird while keeping it small and fast it'll be a good thing, even if it's still a while before we see widespread use of SVG. As long as there's a runtime option to turn it off. ;-)
So you'd rather have your bandwidth gobbled up by animated text gifs than to spare a couple of bytes for some actual text marked up with a few SVG tags + CSS?
> SVG SUCKS for handicap access sites...
SVG markup is inherently more screen-reader-friendly than a binary format like SWF.
> Unfortunately, you can't do this whilst serving your XHTML documents as text/html (which is the only way Internet Explorer will understand them properly).
And here I was thinking that IE was the browser that only paid attention to file extensions while blithely ignoring mimetypes. Silly me.
In any case, the proper way to handle this issue is by using namespaces, is it not?