The interesting thing is that it's good that Firefox has such a fast release process because their quality control only catches little things like "images no longer working" after a release. And "reflects badly on Mozilla's testing efforts" is a formulation Slashdot users would not be generous enough to use with Apple or Microsoft products. Though the spirit emulates that from the posted article, which said the update fixes "a compatibility issue with some websites and extensions". The way I heard it, images weren't showing up: that's hardly a "compatibility issue" in a web browser.
All he was trying to do is see if an average business USER would be able to use the OS effectively.
And the average business user would naturally use X11 on the Mac. TFA has two and a half whole pages on X11 and its troubles, but that's hardly something that most users would even look for... you even have to install it separately.
He's right about the culture being more one that buys software rather than getting it for free -- but the quality of those tools is then also significantly higher than a lot of open-source software (with exceptions, of course). For example, CSSEdit or Fission are both easily worth the $20 or $25 you spend on and pay themselves back within days. On the other hand, he's wrong about the paucity of Freeware in general; it's just that the favorite tool tends to be commercial rather than freeware, as is often the case on Windows. On top of that, I've found the general quality of Freeware to be outstanding (Adium, Tunnelblick, TextWrangler, etc.), which is more than I can say for the download orgy that one often embarks on when searching for a tool on Windows.
Just because you, personally, don't happen to work on those projects, doesn't invalidate its usefulness for others who do.
Or, far more likely, "Just because you, personally, choose not to acknowledge that you work on projects that could benefit immensely from Design by Contract, doesn't invalidate its usefulness for others who do."
I've had to use GIMP (and its cousin, Inkscape, who's interface shows a close family resemblance) only once or twice. It feels like an application made by people who think of a UI as gratuitous -- a necessary evil for those too lazy or stupid to use a command line. When it starts with no documents open, there are at least 4 or 5 freely floating windows, including a main window with umpteen stacked toolbars.
And now I find that people have been using it for years without such basic tools as "Align". Aligning manually is like kerning your fonts by hand -- what's the point of software in this case? I bet the export as EPS function opens a text editor.
Wouldn't a user then expect some modicum of respect for a comment, regardless of how insipid?... after all, they paid good money to make that comment.
In the same way that Slashdot is in no way obliged to pay attention when I complain about their constant erosion (I don't consider spelling "lose" as "loose" and "balking" as "baulking" as advancing the language...:-/), sites which allow only for-fee commenting may have to relax some of their moderating practices.
Careful... while there is a high-quality mode, I believe there is also an "Ultra" quality mode, mentioned in PC-Gamer, which HardOCP doesn't even touch (I imagine id didn't bother benchmarking because it's too slow). According to PC-Gamer, you need a card with 512MB on it to even think about running that mode at reasonable levels.
I think the graphical browsing is very nice... but what happens when you can't remember where in the 1 meter x 1 meter space you should be looking?
You still need a find-as-you-type thing like Opera has, just to narrow down your selection. When you type, it could zoom out and highlight the 'found' pages, a la Expose. Then you can click on them to see if that's the thing you want.
I think it really needs an Expose-like view that shows the whole graph at once (within reason) and lets you display a set that matches a query.
The unreleased version of Opera M2 is free (with two small Google ads)
Yeah, but it's not available yet because it's a non-public Beta available only to users signed up as beta-testers with Opera. I know, because I'm a tester for the Mac version (6.03 is still the official version on that platform, but I've been using 7.x for months)
How can he justify reviewing an unreleased version of Opera M2, but then review an older version of Outlook because most Windows users don't have it yet?
Opera 7.50 is actually available for Mac and Linux as well (in beta-testing, as with Windows).
Also, as far as I know, pretty much every key can be rebound and configured in Opera.
When I read your comment, I thought I would quickly download a SLOC counter and write back that I had such an application. Sadly, I cannot do that. Over the last couple of years, I've slowly built a library in PHP to drive my web-site. It is almost exclusively composed of classes and follows OO typing rules in a fairly disciplined way.
I thought it was a big library, but it's only 22,000 lines of code. The change request/bug tracker I built on top of the core library can only add a paltry 10,000 more.
That's not even close to 100,000 lines, but perhaps that's also because I get a *lot* of re-use out of my class hierarchy.
I'm sure I speak for all C++ programmers when I say thanks for leaving this incredibly crucial rule entirely out of the language and entirely up to individual programmers to enforce with design guidelines.
As Bjarne notes, class invariants are incredibly important, as are routine pre- and post-conditions -- the tenets of Design-by-contract, but they are mysteriously never considered as actual language features in C++. Want a language that actually does help with class and code correctness (allowing developers to specify usage rules)?
Anyone remember that bit by Monty Python where those 4 guys are sitting around, one-upping each other on who had the toughest life?
"Right. I had to get up in the morning at ten o'clock at night half an hour before I went to bed, drink a cup of sulphuric acid, work twenty-nine hours a day down mill, and pay mill owner for permission to come to work, and when we got home, our Dad and our mother would kill us and dance about on our graves singing Hallelujah."
Kind of feels like that in here (did you catch the guy who runs 70 miles a week and pretends to still be answering the beer-drinking, gut-growing dude's question about simple exercise to try?)
But tomorrow, a new coder will add something that modifies the preconditions and suddenly that pointer can indeed be NULL.
That's why it would be lovely if more people would start using tools that allows preconditions, postconditions, etc. to be specified at the language level. Then, there are no more magically changing preconditions because they are explicitly specified in the code instead of implied through comments, or, worse, coding style. Changing a precondition then requires a concrete act of actually removing it from code.
I can't believe some other people finally beat me to the obligatory Eiffel plug here. That's quite refreshing actually. I've used Vision2 for some smaller test projects and found it to be an extremely well-written, abstracted library...pretty much like all the other Eiffel libraries.
To be fair, though, it only works on Unix/Windows...leaving Mac OS X out of the discussion when talking about cross-platform is problematic because most people kind of assume that's included in a cross-platform package (witness Mozilla or Opera, for example). Judging by Opera's results, I'd have to say QT does a great job as well.
Here comes the plug....if more people actually found out which features the Eiffel language and libraries offer, there would be a *lot* more people developing in Eiffel....or wishing they could.
I've scrolled through and seen the same comment over and over. Why not Gecko? Objectively, without knowing about standards-support details (because they are details...the sites I looked at, including mine, looked great in Safari), the browser is just *sick* fast. I've got a 15" iMac and I've got 7 browsers on it now (some just for testing.)
- Omniweb has bad standards support. - Mozilla/Netscape 6 are both really slow. They're tedious to browse with. - Opera has good standards support (outside of DOM missing...I know, flame away) - Chimera is faster than Mozilla, but still feels slow loading and rendering. - IE is faster than Chimera, but generally slower in scrolling than Opera.
And now, the seventh browser -- Safari.
It's wicked fast. Looks great. SCROLLS fast.
You see the killer app is not a web browser on OS X. It's a web browser that can scroll quickly. Go and use a Windows box for a while, then use any of the other browsers on an iMac and scroll the Slashdot site (try this page, for example). You'll want to kill yourself.
Remember. One word. Scrolling.
Finally, someone figured out how to do it quickly in OS X.
Slashdot Mirror
Nah, it looks like a professionally-done Doom 3 mod.
Thanks for clearing that up ... I was just as lazy in verifying my source (heise.de) as they were in reporting it.
The interesting thing is that it's good that Firefox has such a fast release process because their quality control only catches little things like "images no longer working" after a release. And "reflects badly on Mozilla's testing efforts" is a formulation Slashdot users would not be generous enough to use with Apple or Microsoft products. Though the spirit emulates that from the posted article, which said the update fixes "a compatibility issue with some websites and extensions". The way I heard it, images weren't showing up: that's hardly a "compatibility issue" in a web browser.
He's right about the culture being more one that buys software rather than getting it for free -- but the quality of those tools is then also significantly higher than a lot of open-source software (with exceptions, of course). For example, CSSEdit or Fission are both easily worth the $20 or $25 you spend on and pay themselves back within days. On the other hand, he's wrong about the paucity of Freeware in general; it's just that the favorite tool tends to be commercial rather than freeware, as is often the case on Windows. On top of that, I've found the general quality of Freeware to be outstanding (Adium, Tunnelblick, TextWrangler, etc.), which is more than I can say for the download orgy that one often embarks on when searching for a tool on Windows.
400mg of Capsaicin is suicide. That's about 1lb of the stuff...that's a hell of a "capsule".
I've had to use GIMP (and its cousin, Inkscape, who's interface shows a close family resemblance) only once or twice. It feels like an application made by people who think of a UI as gratuitous -- a necessary evil for those too lazy or stupid to use a command line. When it starts with no documents open, there are at least 4 or 5 freely floating windows, including a main window with umpteen stacked toolbars.
And now I find that people have been using it for years without such basic tools as "Align". Aligning manually is like kerning your fonts by hand -- what's the point of software in this case? I bet the export as EPS function opens a text editor.
Wouldn't a user then expect some modicum of respect for a comment, regardless of how insipid? ... after all, they paid good money to make that comment.
... :-/), sites which allow only for-fee commenting may have to relax some of their moderating practices.
In the same way that Slashdot is in no way obliged to pay attention when I complain about their constant erosion (I don't consider spelling "lose" as "loose" and "balking" as "baulking" as advancing the language
Careful ... while there is a high-quality mode, I believe there is also an "Ultra" quality mode, mentioned in PC-Gamer, which HardOCP doesn't even touch (I imagine id didn't bother benchmarking because it's too slow). According to PC-Gamer, you need a card with 512MB on it to even think about running that mode at reasonable levels.
Imagine if the entire IE market were to start using Firefox. How many of them do you think would use extensions?
For the average user, there's a huge advantage to having a product that was designed
I think the graphical browsing is very nice ... but what happens when you can't remember where in the 1 meter x 1 meter space you should be looking?
You still need a find-as-you-type thing like Opera has, just to narrow down your selection. When you type, it could zoom out and highlight the 'found' pages, a la Expose. Then you can click on them to see if that's the thing you want.
I think it really needs an Expose-like view that shows the whole graph at once (within reason) and lets you display a set that matches a query.
Thanks very much ... I've been wondering where the heck that was.
The unreleased version of Opera M2 is free (with two small Google ads)
Yeah, but it's not available yet because it's a non-public Beta available only to users signed up as beta-testers with Opera. I know, because I'm a tester for the Mac version (6.03 is still the official version on that platform, but I've been using 7.x for months)
How can he justify reviewing an unreleased version of Opera M2, but then review an older version of Outlook because most Windows users don't have it yet?
Opera 7.50 is actually available for Mac and Linux as well (in beta-testing, as with Windows).
Also, as far as I know, pretty much every key can be rebound and configured in Opera.
IE 5.x and higher only. With annoying popup preventing proper 'back' behavior. I'm enthralled.
When I read your comment, I thought I would quickly download a SLOC counter and write back that I had such an application. Sadly, I cannot do that. Over the last couple of years, I've slowly built a library in PHP to drive my web-site. It is almost exclusively composed of classes and follows OO typing rules in a fairly disciplined way.
... it's GPLd. The documentation was generated with PHPDoc.
I thought it was a big library, but it's only 22,000 lines of code. The change request/bug tracker I built on top of the core library can only add a paltry 10,000 more.
That's not even close to 100,000 lines, but perhaps that's also because I get a *lot* of re-use out of my class hierarchy.
Check it out at earthli Webcore if you're interested
"... as is every other programming langauge."
As is the English "langauge", eh?
As Bjarne notes, class invariants are incredibly important, as are routine pre- and post-conditions -- the tenets of Design-by-contract, but they are mysteriously never considered as actual language features in C++. Want a language that actually does help with class and code correctness (allowing developers to specify usage rules)?
Use Eiffel.
Kind of feels like that in here (did you catch the guy who runs 70 miles a week and pretends to still be answering the beer-drinking, gut-growing dude's question about simple exercise to try?)
That's why it would be lovely if more people would start using tools that allows preconditions, postconditions, etc. to be specified at the language level. Then, there are no more magically changing preconditions because they are explicitly specified in the code instead of implied through comments, or, worse, coding style. Changing a precondition then requires a concrete act of actually removing it from code.
I can't believe some other people finally beat me to the obligatory Eiffel plug here. That's quite refreshing actually. I've used Vision2 for some smaller test projects and found it to be an extremely well-written, abstracted library...pretty much like all the other Eiffel libraries.
To be fair, though, it only works on Unix/Windows...leaving Mac OS X out of the discussion when talking about cross-platform is problematic because most people kind of assume that's included in a cross-platform package (witness Mozilla or Opera, for example). Judging by Opera's results, I'd have to say QT does a great job as well.
Here comes the plug....if more people actually found out which features the Eiffel language and libraries offer, there would be a *lot* more people developing in Eiffel....or wishing they could.
I wouldn't worry too much about that. It'll be more like:
"Oh no! Did he say nukuler?!"
I've scrolled through and seen the same comment over and over. Why not Gecko? Objectively, without knowing about standards-support details (because they are details...the sites I looked at, including mine, looked great in Safari), the browser is just *sick* fast. I've got a 15" iMac and I've got 7 browsers on it now (some just for testing.)
- Omniweb has bad standards support.
- Mozilla/Netscape 6 are both really slow. They're tedious to browse with.
- Opera has good standards support (outside of DOM missing...I know, flame away)
- Chimera is faster than Mozilla, but still feels slow loading and rendering.
- IE is faster than Chimera, but generally slower in scrolling than Opera.
And now, the seventh browser -- Safari.
It's wicked fast. Looks great. SCROLLS fast.
You see the killer app is not a web browser on OS X. It's a web browser that can scroll quickly. Go and use a Windows box for a while, then use any of the other browsers on an iMac and scroll the Slashdot site (try this page, for example). You'll want to kill yourself.
Remember. One word. Scrolling.
Finally, someone figured out how to do it quickly in OS X.
Yay. *confetti*