God *damn* it, I wish the IBM PR folks would get off their asses and get this article corrected if it's wrong.
Also, I agree wholeheartedly with you that Notes sucks. It has wonderful lock-in potential, and I'm sure it's smart from a business perspective, but good God, it's a royally awful piece of software.
You know, it's really silly that WorldShip isn't Java-based. It's basically nothing more than a front end that does a bit of network work. You couldn't *ask* for something more appropriate for Java.
If IBM *really* needs some features, it might be a useful long-term expense (since it also improves their strategic position) to add a developer or two to the Dia project or similar, but doing something from the ground up...ick.
I'd like to suggest that you consider a current copy of Dia, if you haven't already.
Visio probably has more features and pre-drawn symbols. However, the only thing I've needed to use the two for -- building UML project skeletons -- it was *much* faster to enter the data in Dia, which didn't require clicking through a number of fields and tabs for each element entered. I've also had Visio die on me a couple times, which was quite irritating. Also, Dia could generate template code from my diagrams, which was kind of convenient.
Note that Dia has a GIMP-style interface, which weirds out a number of people. However, give it a try and see what you think. I was quite pleased.
Note that I'm not sure how easy it is to automate Dia. GNOME/GTK+ apps tend to be rather weak on scriptability, frusterating on a platform where scriptability has tied the whole thing together for ages.
This is actually also a big deal because of folks like you.
If there are suddenly thousands of talented researchers and developers moved to Linux, there is a sudden influx of people who now can scratch any itch that bothers *them* during their daily work. Perhaps they don't like something about their web browser -- they can just whip up a patch and send it in. They don't have to put up with the problem, and everyone else benefits.
is that a lot of people seem to have enough spare time that they can spare some to go into a thread in which they have no interest and complain about the topic.
This is not an issue where the topic is simply not of interest to me. It's where answers are unlikely to help other people. A topic of "What would you do to identify unknown cards?" might have produced similar information, but been much more useful.
Now, you can play the "just ignore what you don't want" card, but you're oversimplifying things. Slashdot has editors for a reason. If we could simply ignore what we didn't want with no cost, the story submissions might as well immediately go up on the main page.
Ah. Well now I kinda feel bad. I hope you're not personally insulted by my comments. If you got the idea that I'm against ever using Gimp, then I apologize.
Well, I'm not currently a core GIMP developer -- I just have a bit of free time at the moment, and am working on adding some GIMP features (which aren't currently in Photoshop, FWIW).:-) Since GIMP is currently in version 2 prerelease freeze, everything's just sitting on my own system at the moment, and will probably wait for version 2.1 or 3 or something to be included. But hey, no fear -- honest criticism is the only way software gets better.
If I was going to criticize GIMP, there are a couple of other things I'd add:
* GIMP, unlike almost all other major UNIX projects, is not particularly modular and is a pain in the ass to use from your average bash script or as a batch processor -- unlike ImageMagick. GIMP is unfortunately rather tied to X11 -- many is the time that I'd like to do effects in GIMP from the commandline, but no dice.
* I believe Photoshop does a nicer job of scaling performance up to very large images with that custom VM system it has.
* Someday, maybe, just maybe, it'd be nice to have GIMP do what xRes once did.
* The GIMP doesn't have unlimited undos. Photoshop, from what I hear, currently has a very, very nice undo system that acts essentially as a tree. I know of no other app (unless perhaps other Adobe apps like Illustrator do) that compare.
* The GIMP has some experimental work being done on natural-media composition, but it's a long way from competing with Painter.
* Photoshop has a few lighting effects that the GIMP can't pull off. Synthesizing glass in GIMP is significantly more difficult.
My attachment to three (not two) button mice is due to the fact that I use X11 software, and would even on a Mac. I do also like contextual menus.
Control-clicking is not a reasonable solution. It is a good, understandable way to transition people to multiple-button mice, and it was what I expected Apple to be doing. However, Apple has ended up presenting control-clicking as a good long-term solution, which it is not. They have decided to go with contextual menus, but to force you to use both arms to bring one up, is silly. I do not deny that a computer can be used in such I way -- I do so when I'm using a Mac. However, it's decidedly irritating. On my Linux box, I can use either the mouse or the keyboard at any given time. I can even comfortably eat while web browsing. On the Mac, I'm forced to constantly use both arms.
Also, I have seen few folks arguing that Apple is doing the *right* thing, or the *superior* thing, but rather that they are doing something that is not as bad as people are making it out to be.
I've had the pleasure of being able to talk to some engineers working at Apple when OS X was in production, and happened to bring up the multiple button issue. Both grimaced, and laughed. Apparently, from what they told me, the One Mac Button is a decision that comes straight from the Top -- Jobs is firmly married to the vision of a simple, easy-to-use single button mouse. I feel that he's wrong here. His idea was somewhat justified two decades ago, when folks were not familiar with computers, much less multibutton mice. However, whether Jobs wants to admit it or not, just about everyone has run into Windows, or at the *very* least, an OS that uses multiple buttons. Multiple buttons are just not a foreign concept that average Joe cannot understand any more. There is no significant ease-of-use issue present any more, and there is a functionality difference -- and a lot of peeved folks. The time has come to make the switch.
Apple occasionally has a "We Do Things Our Own Way, Dammit" moment. They provided only SCSI interfaces for an awfully long time, for instance. They insisted on using those darn little eight-pin serial ports for ages. They won't change their single button mouse style. In the PC world, people that do this quickly go away, because people simply use a competitor's product. However, you just don't have such an option in the Mac world, where Apple is the only game in town. So, while Apple is a Pretty Good hardware provider, if they insist on maintaining an absolute monopoly, they have to be The Best hardware provider to compete with the PC world, where people can simply choose hardware to suit their tastes.
Nowhere near as many layer modes. Photoshop is more like working in a darkroom,
Really? I'm a bit out-of-date on Photoshop, but IIRC GIMP has all the layer modes that Photoshop did back when I was admining a lab of Photoshop machines (which was, admittedly, back in the day...)
Photoshop's MDI environment is much better in that respect.
True, but Photoshop would then suck in a multi-windowed environment. It drives me up a wall that the X11 version of Opera can't do multiple windows -- that you're stuck with MDI and a single desktop. You may be on to something with Mozilla's combined MDI/SDI approach -- more apps really could use it. GIMP does do some of this -- the palettes are tearable tabs in GIMP 2 (well, GIMP 2 prerelease, soon to be GIMP 2).
* One of the filters I tried to show my coworker didn't have a preview mode. (It was either blur or a sharpen filter I think...) I've read that earlier versions of Gimp had no preview modes to speak of in any of the features, but that may have been fixed in recent months. I can tell you that as of a week or two ago, we ended up having to use trial and error to figure out the right number to plug in to a filter.
This used to be more epidemic than it once was. A few plugins still do not have it -- the Gaussian blur plugins lack it. This is actually on my to-do list.:-)
* This is related to my taskbar complaint, but a nicely contained problem all on it's own. Since there's no 'background window' to Gimp, be careful about running it with other apps on the same desktop. Otherwise an accidental click will bring a background window to front. Obnoxious. Simple fix really. Set up a background window, and when it comes to focus, bring up the other related windows. Don't want that? set up a switch to turn it off.
Mmmf. I can see how this might be annoying on other platforms -- in a traditional X11 environment, it's par for the course, and most folks comfortable with X11 apps are pretty happy.
Some window managers (such as sawfish) can group windows, if you want the Mac OS-style transparent MDI.
I don't run into accidental clicks outside of windows, though.
Gimp has the potential to be on the same level as Photoshop. That day has to arrive before I dump Photoshop. Sorry. I've got work to do.
Sure, but you do print work, right?
For folks that only do digital output, it's not such a big deal -- GIMP does have a few lacks, but it works pretty well for digital output.
Instead of trying to prevent an insecure mail client from screwing you over by trying to block it in on all sides, wouldn't it be easier to just use a secure mail client?
The major things I've seen that Photoshop has that GIMP doesn't are:
* No neat duotone tool. I like duotones.
* No non indexed/RGB color model support. Very, very bad if you're doing output for professional printing.
* Not sure, but I suspect Photoshop has better color matching support.
* Photoshop has a nicer warping interface.
* There are more plugins available for Photoshop. They're often quite pricy, but if you're a professional designer (the sort of person that would care about four color work and hence want to use Photoshop instead of GIMP), you're probably going to make back the cost pretty quickly.
There are only a few things that I know of that GIMP can do that Photoshop can't. Among these are:
* Better support for many languages to write plugins in.
* Some researchy plugins that go well beyond what Photoshop can do; Resynthesizer is one.
Their hardware is priced higher than the cheapest possible x86 box you can buy, but it also doesn't suck and doesn't contain used parts.
Actually, Apple doesn't really use ultra-nice components, as they did back in the day. I'd say what they put into their cases is about what you'd get in an x86 box. On the other hand, Apple's prices have also come down a lot since the day, and their laptops are price competitive.
OS X supports all hardware going back to their various G3 models, which means pretty sizable number of processors, laptops, video cards, motherboards, USB devices, firewire devices, printers, audio hardware, etc. etc.
This is still awfully few compared to the number of devices for the number of things that Linux can talk to.
Also, USB (as long as the hardware is HID-compliant) support is free. The USB mouse is going to be supported -- the USB SmartHome X10 controller may not be.
Sure, Linux and Windows still support your 1996 video card, but maybe it's time to invest a wee bit more money in your hardware setup?
Dammit, it's exactly this kind of thinking that irritates me about Apple. No, I bought the thing, it works fine, I don't need more performance at the moment, and so I don't see why I should pay even more because Apple found it profitable to not support something.
Most of what is here is cutting but true. However, the article, unlike most Slashdot posters, does not claim this. He doesn't say that OS X is a "better Linux" -- he says that they're two different beasts.
"Despite the fact that Linux is just code and can't WANT to be anything, I truly believe that it'd love to be a single-vendor, single-platform, sluggish half-proprietary OS with dwindling market share. Linux would love to throw away its impressively growing corporate takeup for that."
True, but I don't believe OS X has dwindling market share.
"Apple hardware is for real computer lovers."
I don't think I've ever seen people say this.
"My non-techie friends drool over the transparency and scaling effects, even though UI research has shown that they add practically nothing to getting real work done. It feels like KDE 2 on a Pentium 200, and I can't change to a light and fast WM, but those drop-shadows must make me work so quickly!"
True. However, I think they may also be referring to the lower learning curve of much Mac OS software. Unless you're using software quite a bit, the learning curve plays a larger role than the total amount of functionality. I claim that it takes around three years of heavy use of emacs before you really start to get a lot more good out of it than its traditional Windows and Mac OS counterparts.
"OpenDarwin.org and its community of about 27 is surely not just a token gesture by Apple. Pretty much nobody uses pure Darwin, and all the crucial components of the system are closed and require me to spend money just to get major OS updates, but they're really helping the community somehow."
True. Apple does not "get it" WRT open source in anywhere near the same way that Red Hat and friends do. They produce a high-end, propriatary product. However, they are infinitely better than Microsoft (and to many people, Mac OS is a valid alternative to Windows...but Linux is not). Furthermore, even before the open source thing started up, Apple was much better about helping folks tinker around with internals than Microsoft was.
"My iBook was made by in Taiwan by AlphaTop and has design and build quality flaws (needing foam sheets jammed in to stop the common problem of the keyboard scratching the screen). But it's silvery and cost far more than an x86 laptop of better spec, so it must be much higher quality!"
I agree that many folks try very hard (and fail) to justify the amount of money spent on their Apple hardware. I find such claims pretty much futile on desktops. However, while they aren't perfect, many Apple laptops are fairly price competitive and pretty good compared to their PC counterparts. Yes, Apple has had a history of doors breaking off, of scratches, and of some flimsiness. But I've also seen countless x86 laptops with all kinds of problems as well. Apple may not be light years ahead here, but they sure aren't light years behind either.
"Although there's truth in PPC being more elegant than x86, it's crushing that the top-of-the-range 1.5 GHz chip is slaughtered by the equivalent 3 GHz Pentium 4. However, Steve Jobs showed some vague Photoshop filter benchmarks at the last MacWorld, so being a leprotard, I'm convinced."
Very true. Macs are (significantly) slower than x86 machines. It's simply true. Folks who are arguing that Macs are good should not waste their time trying to argue otherwise. They're much better off with the "Yes, but what are you actually *using* said cycles for? I'm getting drop shadows out of it -- you seem to be using about 2% of your CPU on average!"
if your buying a computer based on how many buttons its OEM mouse has, you have some major issues.
There is one really, really big issue. Apple is famous for their laptops. Apple's desktops are not (IMHO) particularly exceptional or cost-competitive, but their laptops have traditionally been near-PC price and well-built. Most people I know that want Apple hardware want a laptop.
However, if you purchase an Apple laptop, you cannot simple snap in a new trackpad. You are stuck with a single button. Yes, you can can purchase an external mouse, but then you're stuck using an external mouse with your laptop. This is a pain in the ass, and something that you can avoid on non-Apple laptops -- you can get nice three-button laptops elsewhere.
This is not something that Apple is unaware of or incapable of fixing. However, they have made a conscious (and much-protested) decision to not natively support multiple buttons in their hardware, even as an option. While I can respect their reasons for doing so, it does make their hardware much less appealing. The reason people get so bent out of shape about this is partly because Apple *insists* on forcing you to use their hardware to use their software, and *insists* on not providing an option for more buttons for the (many) folks that are unhappy with their default setup.
If this is not a problem for your uses, that's fine. For me, it would be a major issue -- having to find a flat surface and carry along a big clunky external device to use the thing *is* an issue. Please do not call this "nitpicking" -- it is an entirely justified criticism that Apple has chosen not to address.
Local variables may be declared at any point in the code.
Supported by ISO C99 and later.
Overloaded function names by the argument types. Requires name mangling, unfortunately.
I think you've pointed out why it's going to be tough to get this in -- previous ABI changes in C++ have really, really sucked for everyone involved. I agree that it would be nice, but it would mean significant changes to a *lot* of software, and would produce a couple of years of pain.
Single inheritance using the Plan9 C syntax: if you have a type Type and declare "Type;" (no variable name) in a structure, an automatic cast from that structure to Type is allowed. If this is put first in the structure you have normal C++ public inheritance.
Hmm...I dunno. This is certainly an interesting idea, but it seems to me that this could cause nasty problems. You wouldn't get a lot of use out of this with primitives, since automatic conversion between two things typedefed to, say, unsigned long is already performed. So I guess this only applies to structures. I'm assuming that the cast simply introduces code to allocate and produce a new structure of the proper size, and then does a field-by-field copy, to deal with size/alignment differences?
Some simple static "constructors" such as the ability to declare an initialization value for each field in a structure definition. This would actually be much better than C++ constructors for a lot of code.
Not as easy as it seems -- this is a gimme for statically allocated structures, but you couldn't pull this off with dynamically allocated structures (well, without a *lot* of fiddling behind the scenes). If you allow this only for statically allocated structures, you introduce a somewhat confusing inconsistency.
No, no. I'm aware of this -- it also does tilde expansion. The problem is that it still isn't as nice as it should be.
In zsh, I can simply whack the spacebar to separate the filename from the directory name (depending upon the options you have set, this may not be necessary) and tab complete the directory without destroying the default file name. This is a particularly big deal for saving files off the Web, where I generally want to keep the default name. I can't do this in the GTK+ file selector.
In zsh, I can tab-complete using ".." in my path to go up a level. I can't do this in the gtk+ file selector.
There are a horde of other little features that zsh has that the file selector doesn't have. It's just so frusterating that zsh has a good, solid tab completion engine and gtk+ can't reuse that design.
Can you please explain further? There is one standard drag-and-drop box.
My comment was with DnD in general, not with just opening and saving. This may not be a huge problem with opening and saving.
The problem is that it takes a lot of work to do DnD everywhere (so that I can simply see something, want to copy it, and drag it). Application authors haven't done that -- every list of items I see, for instance, isn't draggable, so I don't use DnD.
On the Acorn platform it's the _only_ way of doing things, and works in 100% of the applications.
Okay, fair enough. Except the Acorn. (I've never used an Acorn.)
Slashdot Mirror
God *damn* it, I wish the IBM PR folks would get off their asses and get this article corrected if it's wrong.
Also, I agree wholeheartedly with you that Notes sucks. It has wonderful lock-in potential, and I'm sure it's smart from a business perspective, but good God, it's a royally awful piece of software.
You know, it's really silly that WorldShip isn't Java-based. It's basically nothing more than a front end that does a bit of network work. You couldn't *ask* for something more appropriate for Java.
It'd be expensive, though.
If IBM *really* needs some features, it might be a useful long-term expense (since it also improves their strategic position) to add a developer or two to the Dia project or similar, but doing something from the ground up...ick.
I'd like to suggest that you consider a current copy of Dia, if you haven't already.
Visio probably has more features and pre-drawn symbols. However, the only thing I've needed to use the two for -- building UML project skeletons -- it was *much* faster to enter the data in Dia, which didn't require clicking through a number of fields and tabs for each element entered. I've also had Visio die on me a couple times, which was quite irritating. Also, Dia could generate template code from my diagrams, which was kind of convenient.
Note that Dia has a GIMP-style interface, which weirds out a number of people. However, give it a try and see what you think. I was quite pleased.
Note that I'm not sure how easy it is to automate Dia. GNOME/GTK+ apps tend to be rather weak on scriptability, frusterating on a platform where scriptability has tied the whole thing together for ages.
This is actually also a big deal because of folks like you.
If there are suddenly thousands of talented researchers and developers moved to Linux, there is a sudden influx of people who now can scratch any itch that bothers *them* during their daily work. Perhaps they don't like something about their web browser -- they can just whip up a patch and send it in. They don't have to put up with the problem, and everyone else benefits.
Open Source Software is a positive feedback loop.
is that a lot of people seem to have enough spare time that they can spare some to go into a thread in which they have no interest and complain about the topic.
This is not an issue where the topic is simply not of interest to me. It's where answers are unlikely to help other people. A topic of "What would you do to identify unknown cards?" might have produced similar information, but been much more useful.
Now, you can play the "just ignore what you don't want" card, but you're oversimplifying things. Slashdot has editors for a reason. If we could simply ignore what we didn't want with no cost, the story submissions might as well immediately go up on the main page.
Ah. Well now I kinda feel bad. I hope you're not personally insulted by my comments. If you got the idea that I'm against ever using Gimp, then I apologize.
:-) Since GIMP is currently in version 2 prerelease freeze, everything's just sitting on my own system at the moment, and will probably wait for version 2.1 or 3 or something to be included. But hey, no fear -- honest criticism is the only way software gets better.
Well, I'm not currently a core GIMP developer -- I just have a bit of free time at the moment, and am working on adding some GIMP features (which aren't currently in Photoshop, FWIW).
If I was going to criticize GIMP, there are a couple of other things I'd add:
* GIMP, unlike almost all other major UNIX projects, is not particularly modular and is a pain in the ass to use from your average bash script or as a batch processor -- unlike ImageMagick. GIMP is unfortunately rather tied to X11 -- many is the time that I'd like to do effects in GIMP from the commandline, but no dice.
* I believe Photoshop does a nicer job of scaling performance up to very large images with that custom VM system it has.
* Someday, maybe, just maybe, it'd be nice to have GIMP do what xRes once did.
* The GIMP doesn't have unlimited undos. Photoshop, from what I hear, currently has a very, very nice undo system that acts essentially as a tree. I know of no other app (unless perhaps other Adobe apps like Illustrator do) that compare.
* The GIMP has some experimental work being done on natural-media composition, but it's a long way from competing with Painter.
* Photoshop has a few lighting effects that the GIMP can't pull off. Synthesizing glass in GIMP is significantly more difficult.
My attachment to three (not two) button mice is due to the fact that I use X11 software, and would even on a Mac. I do also like contextual menus.
Control-clicking is not a reasonable solution. It is a good, understandable way to transition people to multiple-button mice, and it was what I expected Apple to be doing. However, Apple has ended up presenting control-clicking as a good long-term solution, which it is not. They have decided to go with contextual menus, but to force you to use both arms to bring one up, is silly. I do not deny that a computer can be used in such I way -- I do so when I'm using a Mac. However, it's decidedly irritating. On my Linux box, I can use either the mouse or the keyboard at any given time. I can even comfortably eat while web browsing. On the Mac, I'm forced to constantly use both arms.
Also, I have seen few folks arguing that Apple is doing the *right* thing, or the *superior* thing, but rather that they are doing something that is not as bad as people are making it out to be.
I've had the pleasure of being able to talk to some engineers working at Apple when OS X was in production, and happened to bring up the multiple button issue. Both grimaced, and laughed. Apparently, from what they told me, the One Mac Button is a decision that comes straight from the Top -- Jobs is firmly married to the vision of a simple, easy-to-use single button mouse. I feel that he's wrong here. His idea was somewhat justified two decades ago, when folks were not familiar with computers, much less multibutton mice. However, whether Jobs wants to admit it or not, just about everyone has run into Windows, or at the *very* least, an OS that uses multiple buttons. Multiple buttons are just not a foreign concept that average Joe cannot understand any more. There is no significant ease-of-use issue present any more, and there is a functionality difference -- and a lot of peeved folks. The time has come to make the switch.
Apple occasionally has a "We Do Things Our Own Way, Dammit" moment. They provided only SCSI interfaces for an awfully long time, for instance. They insisted on using those darn little eight-pin serial ports for ages. They won't change their single button mouse style. In the PC world, people that do this quickly go away, because people simply use a competitor's product. However, you just don't have such an option in the Mac world, where Apple is the only game in town. So, while Apple is a Pretty Good hardware provider, if they insist on maintaining an absolute monopoly, they have to be The Best hardware provider to compete with the PC world, where people can simply choose hardware to suit their tastes.
Nowhere near as many layer modes. Photoshop is more like working in a darkroom,
:-)
Really? I'm a bit out-of-date on Photoshop, but IIRC GIMP has all the layer modes that Photoshop did back when I was admining a lab of Photoshop machines (which was, admittedly, back in the day...)
Photoshop's MDI environment is much better in that respect.
True, but Photoshop would then suck in a multi-windowed environment. It drives me up a wall that the X11 version of Opera can't do multiple windows -- that you're stuck with MDI and a single desktop. You may be on to something with Mozilla's combined MDI/SDI approach -- more apps really could use it. GIMP does do some of this -- the palettes are tearable tabs in GIMP 2 (well, GIMP 2 prerelease, soon to be GIMP 2).
* One of the filters I tried to show my coworker didn't have a preview mode. (It was either blur or a sharpen filter I think...) I've read that earlier versions of Gimp had no preview modes to speak of in any of the features, but that may have been fixed in recent months. I can tell you that as of a week or two ago, we ended up having to use trial and error to figure out the right number to plug in to a filter.
This used to be more epidemic than it once was. A few plugins still do not have it -- the Gaussian blur plugins lack it. This is actually on my to-do list.
* This is related to my taskbar complaint, but a nicely contained problem all on it's own. Since there's no 'background window' to Gimp, be careful about running it with other apps on the same desktop. Otherwise an accidental click will bring a background window to front. Obnoxious. Simple fix really. Set up a background window, and when it comes to focus, bring up the other related windows. Don't want that? set up a switch to turn it off.
Mmmf. I can see how this might be annoying on other platforms -- in a traditional X11 environment, it's par for the course, and most folks comfortable with X11 apps are pretty happy.
Some window managers (such as sawfish) can group windows, if you want the Mac OS-style transparent MDI.
I don't run into accidental clicks outside of windows, though.
Gimp has the potential to be on the same level as Photoshop. That day has to arrive before I dump Photoshop. Sorry. I've got work to do.
Sure, but you do print work, right?
For folks that only do digital output, it's not such a big deal -- GIMP does have a few lacks, but it works pretty well for digital output.
Instead of trying to prevent an insecure mail client from screwing you over by trying to block it in on all sides, wouldn't it be easier to just use a secure mail client?
Rafaello,
RH is profitable *and* maintaining a user distribution.
Frankly, Halo on Mac OS sucks from a performance standpoint.
The major things I've seen that Photoshop has that GIMP doesn't are:
* No neat duotone tool. I like duotones.
* No non indexed/RGB color model support. Very, very bad if you're doing output for professional printing.
* Not sure, but I suspect Photoshop has better color matching support.
* Photoshop has a nicer warping interface.
* There are more plugins available for Photoshop. They're often quite pricy, but if you're a professional designer (the sort of person that would care about four color work and hence want to use Photoshop instead of GIMP), you're probably going to make back the cost pretty quickly.
There are only a few things that I know of that GIMP can do that Photoshop can't. Among these are:
* Better support for many languages to write plugins in.
* Some researchy plugins that go well beyond what Photoshop can do; Resynthesizer is one.
Oh, and by the way, you can get an all-in-one eMac for about $999. Doesn't sound too outrageously priced to me.
True, but to be honest, the eMac kind of sucks compared to what you can get in the PC world for $1K.
...of course, there's the obvious rejoinder that Mac folks would upgrade more frequently if it were more affordable...
Their hardware is priced higher than the cheapest possible x86 box you can buy, but it also doesn't suck and doesn't contain used parts.
Actually, Apple doesn't really use ultra-nice components, as they did back in the day. I'd say what they put into their cases is about what you'd get in an x86 box. On the other hand, Apple's prices have also come down a lot since the day, and their laptops are price competitive.
A mac is like a top of the line luxury sports car...if I have to pick one to sit down in front of and do work... that's where I'll go.
While I see what you're saying, that's an awfully funny metaphor.
OS X supports all hardware going back to their various G3 models, which means pretty sizable number of processors, laptops, video cards, motherboards, USB devices, firewire devices, printers, audio hardware, etc. etc.
This is still awfully few compared to the number of devices for the number of things that Linux can talk to.
Also, USB (as long as the hardware is HID-compliant) support is free. The USB mouse is going to be supported -- the USB SmartHome X10 controller may not be.
Sure, Linux and Windows still support your 1996 video card, but maybe it's time to invest a wee bit more money in your hardware setup?
Dammit, it's exactly this kind of thinking that irritates me about Apple. No, I bought the thing, it works fine, I don't need more performance at the moment, and so I don't see why I should pay even more because Apple found it profitable to not support something.
Most of what is here is cutting but true. However, the article, unlike most Slashdot posters, does not claim this. He doesn't say that OS X is a "better Linux" -- he says that they're two different beasts.
"Despite the fact that Linux is just code and can't WANT to be anything, I truly believe that it'd love to be a single-vendor, single-platform, sluggish half-proprietary OS with dwindling market share. Linux would love to throw away its impressively growing corporate takeup for that."
True, but I don't believe OS X has dwindling market share.
"Apple hardware is for real computer lovers."
I don't think I've ever seen people say this.
"My non-techie friends drool over the transparency and scaling effects, even though UI research has shown that they add practically nothing to getting real work done. It feels like KDE 2 on a Pentium 200, and I can't change to a light and fast WM, but those drop-shadows must make me work so quickly!"
True. However, I think they may also be referring to the lower learning curve of much Mac OS software. Unless you're using software quite a bit, the learning curve plays a larger role than the total amount of functionality. I claim that it takes around three years of heavy use of emacs before you really start to get a lot more good out of it than its traditional Windows and Mac OS counterparts.
"OpenDarwin.org and its community of about 27 is surely not just a token gesture by Apple. Pretty much nobody uses pure Darwin, and all the crucial components of the system are closed and require me to spend money just to get major OS updates, but they're really helping the community somehow."
True. Apple does not "get it" WRT open source in anywhere near the same way that Red Hat and friends do. They produce a high-end, propriatary product. However, they are infinitely better than Microsoft (and to many people, Mac OS is a valid alternative to Windows...but Linux is not). Furthermore, even before the open source thing started up, Apple was much better about helping folks tinker around with internals than Microsoft was.
"My iBook was made by in Taiwan by AlphaTop and has design and build quality flaws (needing foam sheets jammed in to stop the common problem of the keyboard scratching the screen). But it's silvery and cost far more than an x86 laptop of better spec, so it must be much higher quality!"
I agree that many folks try very hard (and fail) to justify the amount of money spent on their Apple hardware. I find such claims pretty much futile on desktops. However, while they aren't perfect, many Apple laptops are fairly price competitive and pretty good compared to their PC counterparts. Yes, Apple has had a history of doors breaking off, of scratches, and of some flimsiness. But I've also seen countless x86 laptops with all kinds of problems as well. Apple may not be light years ahead here, but they sure aren't light years behind either.
"Although there's truth in PPC being more elegant than x86, it's crushing that the top-of-the-range 1.5 GHz chip is slaughtered by the equivalent 3 GHz Pentium 4. However, Steve Jobs showed some vague Photoshop filter benchmarks at the last MacWorld, so being a leprotard, I'm convinced."
Very true. Macs are (significantly) slower than x86 machines. It's simply true. Folks who are arguing that Macs are good should not waste their time trying to argue otherwise. They're much better off with the "Yes, but what are you actually *using* said cycles for? I'm getting drop shadows out of it -- you seem to be using about 2% of your CPU on average!"
if your buying a computer based on how many buttons its OEM mouse has, you have some major issues.
There is one really, really big issue. Apple is famous for their laptops. Apple's desktops are not (IMHO) particularly exceptional or cost-competitive, but their laptops have traditionally been near-PC price and well-built. Most people I know that want Apple hardware want a laptop.
However, if you purchase an Apple laptop, you cannot simple snap in a new trackpad. You are stuck with a single button. Yes, you can can purchase an external mouse, but then you're stuck using an external mouse with your laptop. This is a pain in the ass, and something that you can avoid on non-Apple laptops -- you can get nice three-button laptops elsewhere.
This is not something that Apple is unaware of or incapable of fixing. However, they have made a conscious (and much-protested) decision to not natively support multiple buttons in their hardware, even as an option. While I can respect their reasons for doing so, it does make their hardware much less appealing. The reason people get so bent out of shape about this is partly because Apple *insists* on forcing you to use their hardware to use their software, and *insists* on not providing an option for more buttons for the (many) folks that are unhappy with their default setup.
If this is not a problem for your uses, that's fine. For me, it would be a major issue -- having to find a flat surface and carry along a big clunky external device to use the thing *is* an issue. Please do not call this "nitpicking" -- it is an entirely justified criticism that Apple has chosen not to address.
Hmm. Are you telling me that I can run out, download a Mac OS app, and run it on my Linux box?
// comments
Supported by ISO C99 and later.
Local variables may be declared at any point in the code.
Supported by ISO C99 and later.
Overloaded function names by the argument types. Requires name mangling, unfortunately.
I think you've pointed out why it's going to be tough to get this in -- previous ABI changes in C++ have really, really sucked for everyone involved. I agree that it would be nice, but it would mean significant changes to a *lot* of software, and would produce a couple of years of pain.
Single inheritance using the Plan9 C syntax: if you have a type Type and declare "Type;" (no variable name) in a structure, an automatic cast from that structure to Type is allowed. If this is put first in the structure you have normal C++ public inheritance.
Hmm...I dunno. This is certainly an interesting idea, but it seems to me that this could cause nasty problems. You wouldn't get a lot of use out of this with primitives, since automatic conversion between two things typedefed to, say, unsigned long is already performed. So I guess this only applies to structures. I'm assuming that the cast simply introduces code to allocate and produce a new structure of the proper size, and then does a field-by-field copy, to deal with size/alignment differences?
Some simple static "constructors" such as the ability to declare an initialization value for each field in a structure definition. This would actually be much better than C++ constructors for a lot of code.
Not as easy as it seems -- this is a gimme for statically allocated structures, but you couldn't pull this off with dynamically allocated structures (well, without a *lot* of fiddling behind the scenes). If you allow this only for statically allocated structures, you introduce a somewhat confusing inconsistency.
No, no. I'm aware of this -- it also does tilde expansion. The problem is that it still isn't as nice as it should be.
In zsh, I can simply whack the spacebar to separate the filename from the directory name (depending upon the options you have set, this may not be necessary) and tab complete the directory without destroying the default file name. This is a particularly big deal for saving files off the Web, where I generally want to keep the default name. I can't do this in the GTK+ file selector.
In zsh, I can tab-complete using ".." in my path to go up a level. I can't do this in the gtk+ file selector.
There are a horde of other little features that zsh has that the file selector doesn't have. It's just so frusterating that zsh has a good, solid tab completion engine and gtk+ can't reuse that design.
Can you please explain further? There is one standard drag-and-drop box.
My comment was with DnD in general, not with just opening and saving. This may not be a huge problem with opening and saving.
The problem is that it takes a lot of work to do DnD everywhere (so that I can simply see something, want to copy it, and drag it). Application authors haven't done that -- every list of items I see, for instance, isn't draggable, so I don't use DnD.
On the Acorn platform it's the _only_ way of doing things, and works in 100% of the applications.
Okay, fair enough. Except the Acorn. (I've never used an Acorn.)
You guys, most window managers are capable of setting said file selector windows to whatever size you want. Medium, mazimized, whatever.
I can use sawfish to do this. However, the somewhat simplified WMs that both KDE and GNOME use by default may not be able to do this.