Sadly, ALSA has made the rotten design decision to put compatibility over functionality with its OSS/dev/dsp emulation.
The specific problem with a Soundblaster16 in Linux (as opposed to Windows(r)) is that it'll only let one program play sound at a time. Others are blocked most inelegantly (a nonrefreshing X11 window is often your only clue that XMMS must be stopped before Mozilla can continue).
ALSA has decided to continue the same bad behavior, so that programs using/dev/dsp will face the same problem as if OSS had been driving the sound. They say "To perfectly emulate OSS, we've got to include the limitations!". Bug-for-bug compatibility...
The 2.6 switch to ALSA is a good thing, but it'll take many years to really help. Only once 3rd party developers (Macromedia, I'm talking to you) stop using OSS interfaces will ALSA truely help against this problem.
Fine, I'll give you a hug. They say vat-born adolscent albino clones need love more than anyone...
(Oh, I shouldn't have respond to that. I really hope slashdot isn't becomng the kind of forum where users deluge each other with HUGS and "Have a COOKIE!". At least we can't embed dancing smilies in animated GIFs)
Yes, that's flaky. But you discovered a workaround by plugging it back into the case. My IO-killing mouse terminates Linux systems no matter how it's attached.
It's a legal requirement to ensure I'm non-infringing on intellectual property.
In KDE, go to the control panel, select Desktop->Size & Orientation and pick your screen size and refresh rate from the pick lists.
KDE has nothing called a "control panel". There is a Control Center, but that has nothing called "Size & Orientation". I'm sure I could get it working with some effort at configuring XFree86 extensions, but RandR is new and not yet something likely to be found on a typical Linux workstation.
Until we have firewire (or similar) video cards/monitors, this seems even less of an issue than changing color depth (which is also pointless
It's not. This problem happens, I've seen it needed on Windows(r) machines. Here's some ways this occurs:
When multiple-monitor support is first activated, the user will want to rapidly experiment with a variety of different settings until something comfortable is achieved. She'd rather leave the important applications running, instead of "vi/etc/XF86Config", Ctrl-Alt-Backspace.
Once multihead support is working, you may still want to "add & remove" displays on the fly. You don't need firewire; you don't even need to be plugging and unplugging hardware. You just need to turn them on and off. Sit at a workstation with only one monitor, power one of them down because you don't need it, and you'll need to ensure that software doesn't think the dark screen is still usable. It must prevent windows from opening up there, and your mouse from wandering over into never-never-land when bouncing off the edge of the screen.
Plus, if you get away from desktops towards laptops, then users are constantly plugging and deplugging from standalone CRTs and projection systems
don't you just set it at 24 or 32-bit and leave it alone?
No, there are many places this can be useful. Once again, there is the initial selection of good settings, which would be nice to accomplish from something like Control Center, without constantly killing X11. (This harkens to the Linux folk's stereotypical obsession with uptime). There are pro-artist applications that can benefit from changing bitdepth. But most importantly, changable bitdepth empowers reheadability. That would allow a proxy X server to move an application from one $DISPLAY to another, without imposing a long-term performance hit on either. It saddens me to see people deploy VNC just so they can reconnect to a running X11 application from multiple X11 servers.
And the multi-session capability of KDE
That's hardly a feature of KDE. More like an ancient XFree86 capability that KDE just recently added a widget to activate. Some more integration from KDE would be nice, such as providing a GUI list of other activate sessions, rather than requiring a user to guess around about Ctrl-Alt-F8, etc. The static CRT-snap which happens even going between sessions of identical monitor configuration is minorly irritating, too.
All I am saying (and really it was only a throwaway pedantic comment) is that from a dictionary definition of the word, I don't believe MS has a monopoly.
Interpeting it as you have, no company has ever held a monopoly. Absolute, 100% "exclusiveness" can never happen in the real world.
The simple way of looking at the "Innovator's Dilemma" is to think about the danger of self-cannibalization. (Oops, that sounds equally obtuse!)
What it means is that the people best able to come up with a new/better/cheaper solution to a problem are often those people who work professionally on old/worse/expensive methods. Beginning to sell the improved product may have you competing with yourself. Like supposedly General Motors can make superior cars that never wear out, but chooses not to so they can earn money from the current ones.
The Innovator's Dilemma is that companies faced with a new idea have a disincentive to bring it to market if it will undercut a currently profitable market. But- they've got to market it sometimes, because if anyone is going to undercut their sales, it should be them. So there's a game of holding back new ideas until just before you think someone else will sell them.
The ID describes how capitalism- the free market that supposedly drives corporations to serve consumer needs- can sometimes encourage them to hold back.
An example of the effect specific to Operating System vendors (and the Linux Distribution market in particular) is the question of support fees. RedHat's biggest revenue source (even moreso, with the recent retreat from retail selling) is customer support contracts. But if the usability of the software itself were improved, then the customer's need to buy support would go down.
So Linux vendors are in the awkward position of profiting by not improving their product. They still have a drive to improve, because having the "most usable" distrib will suck up market share from the others. But improving software quality is not an inarguably healthy idea for them.
Digital Cameras/Scanners... you know it's going to work if you don't buy something that's a cheap piece of shit, which, believe it or not, will suck A WHOLE LOT compared to something 25% more expensive.
Unfortunately, the 25% more expensive figure is off the mark. I'd like to have a scanner around for emergencies, but it's tough to find a USB model in the SANE list for less than $175. However, each December, Staples and BestBuy put the flimsy no-name scanners on sale for $19.95 each ("$9.95 if you come in before 8 AM!")
I'd buy one, if it'd work with Linux (it won't). And I'd actually prefer a small, light machine over a heavy metal box.
Cameras - If you get one with a REAL optical zoom, almost guaranteed it speaks PTP or is a USB mass storage device in disguise, so there's no issue.
And really, savy users will just buy a $12 USB-Flashcard adapter, to preserve battery life when transferring the photos (faster too).
But Joe Dummy would hardly risk installing hardware himself, would he?
Yes he will. He'll install the hardware following the guidebook step by step, and then Windows(r) will "Detecting new hardware" on startup, and take him through 9 dialog boxes, ask for the system disks, and fill the Device Manager with red exclamation points.
Then he'll get on the phone to the hardware manufacturer, and the nice guy on the other end will tell him precisely what mouse clicks will make it work. If he hadn't been using the same OS as 96% of all customers, that recourse wouldn't have been there to save him.
But if someone had set me up with Linux when I was a novice I am sure I would find it easier to install hardware in Linux now.
Not really. There's a whole separate category of problems Linux users can run into, which simply don't happen to people using a majority OS. When buying hardware, it's completely safe to assume that everything in CompUsa is compatible with a Windows/Intel platform, unless the box specifically says "Macintosh". And if you need confirmation, the clerk will absolutely be able to tell if the product is supposed to be Windows compatible.
With Linux, there's no guarrantees like that, so even choosing new hardware is tricky. Most packaging says nothing about whether it's Linux compatible, although much hardware is. Rarely labels announcing Linux compatiblity can be seen, but they sometimes refer only to one specific 2-year old RedHat distribution.
So there's the whole difficult problem of "Can it work?", before you even start to answer "How can it work?". And there are many products that have no hope of Linux compat. Winmodems (except for one Lucent chipset), the SoundBlaster 16 (and compatible) audio cards, Intel's Centrino wireless networking... the list goes on and on.
Just look at me: I'm using a USB mouse at my Linux workstation now. This year I bought a second USB mouse. It works fine in Microsoft(tm) Windows(r) systems. But plug it into Linux, and all USB devices stop responding until you reboot. Apparently there are subtle differences between mice that can really trip up an OS. Who knew?
PS. Technically the Soundblaster 16 can be made to output sounds under Linux. But the restrictions imposed by the driver design render it painfully less capable than the same card in Windows.
Which of the above deficiencies do you think will be corrected first? My money's on XFree86.
Pay up. Although Windows(r) installers always ask the user to reboot, they don't require it. And some self-contained programs work just fine without the reboot.
I can certainly find ONE example of a Windows program installing without a reboot. Can you show me more than ZERO ways to change the depth of an XFree86 display without killing the running applications?
(For that matter, I'd love to see someone even change the dimensions of a live X11 server. Supposedly there's an extension that makes this possible, but I could never get it to work. And would it allow you to add or remove monitors on the fly? We can only hope)
(And no, Control-Alt-Plus doesn't change the server's dimensions...)
The bold font is to make sure respondants like you read at least the most important parts of what I said, but I see from the reply it didn't work.
I'm sorry but you are displaying your total ignorance of the design of X and making completely unjustified and incorrect criticisms in
I'm displaying an ability to make logical inferences. If the authors of StarOffice, OpenOffice, Mozilla, KDE, Gnumeric, and Gimp were all unable to comprehend the X mechanism for clipboard exchange, then it's safe to say that mechanism is either too difficult or too weak to use in a major application. Or are you claiming that the developers of all those projects are stupid and too lazy to make a compatible product?
For a rebuttal, just point me to me one or two famous X11 applications which make correct use of this powerful clipboard, because I can't seem to find any my own.
Please read the design documentation for X before you post about X again.
Frequently X11 defenders fall back to this line, failing to understand that an inability to be comprehensibly documented is itself a design flaw. I have in fact read hundreds of pages of X11R6 documentation, a painful ordeal I have no stomach to repeat soon.
is really as simple as following the very clear guidelines in the excellent explanation by Jamie Zawinski written many years ago.
That's offtopic. The only thing JWZ explains is how to avoid confusing users about "Am I merely selecting text? Or copying it to the clipboard, overwriting previous clipboard contents?" That is not the issue I am complaining about. He only mentions the issue of supporting non-ASCII datatypes in a brief "extra credit" section on content negotiation, where he provides no explanation beyond a reference to some emacs source code. (Humorously, another poster in this thread has held up emacs alongside xterm as examples of major X11 programs that implement the clipboard wrongly)
PS. "Shigoto" means paid work. If someone were paying me, I'd be willing to slog through the ICCCM. For your jibe, I think gekimu would be more appropriate. Although it's attractive to consider arubaito, which (to non-Japanese) brings with it the hyperbolic image of toiling in a death camp.
the almost infinite configurability of X is a strength, not a weakness.
Too bad it's not configurable enough to change color depth while programs are running. A feature Microsoft(tm) has supported since 1996 or so, and which Macs probably had much earlier.
No, it's not the App's fault. An Operating System (which is anything that provides services so programs can run on top of it, and that includes X11) can be justifiably blamed if all of the applications built on top of it share a problem. In fact, it's the only reasonable thing TO blame.
If a user complains that Windows95 always crashes, do we allow Microsoft to get away with blaming the application writers? No! If all/most of the apps on a platform exhibit a problem, then the platform is flawed.
Now, what exactly is the flaw with the X11 clipboard that lead the multiple smart people from Gnome, KDE, Qt, and Motif to be unable to use it correctly? I don't really know. Maybe it's simply too hard to use. That's a valid flaw. I don't have time to completely research the interface, but the JWZ article you indicate contains a familiar keyword: "The content negotiation mechansim isvery powerful"
Whenever a programmer says his system is "very powerful", that's often a euphenism for "excessively generic". It can mean the developer went to the point of supplying enough functionality so that any required task was technically possible, but didn't spend any time making it straightforward to use.
, it's not enough to pass on objects from application to application, unless both applications are fully aware of the type of information the object contains.
And in the case of 24-bit 2d images, each application is fully aware of how to handle data like that. I verfied by using each program to load a JPG. OLE is not relevant there. In the case of the spreadsheet, or just formatted text, it's not needed either.
Weird. Is it really "ray tracing" when all the rays are co-linear and striking surfaces perpendicular to them? Sounds related to a software z-buffer implementation.
This approach eliminates the need to compute intersections, unions, etc..
IMO, computation of intersections and unions is typically an optimization attempted to accelerate the otherwise slow performance of per-pixel depth checking.
Scalability concerns: in the future, display resolutions will increase so that exponentially more pixels exist (and more rays are needed). Also, the complexity of onscreen graphics can be expected to increase so there are more primitives out there. RT methods will be really hurt by the pixel increase, while higher-level geometric calculations only care about the number of primitives. And, more primitives hurts RT too, as each ray has a larger set of potentially intersecting objects to check for.
The relocatability of X11 applications has been academically studied for a long time, and implementations like xmove are easy to download. But their shortcomings are numerous, as you'll quickly notice by giving it a try. The existence of the pseudo-server is the most obvious shortcoming, as it means there will be a performance hit for the entire life of the application, even though you'll only rarely want to move an X11 app from one display to another.
If X11 supported re-heading deeper in the protocol, then not only would transferring operation be cheaper, but we also might be able to restart XFree86 on a Linux system without killing all the applications connected to it.
We have one, it works and all new applications use it.
That is false. So completely false, I don't know how to begin. It's apparently based on the erroneous premise that ASCII text is the only thing that matters on a clipboard.
I just conducted a 20 minute study with recent versions of six high-profile X11 applications, as found in Debian Linux. I looked at two categories of data: 3 cells from a table ("spreadsheet"), and a 24bit photo imported from JPEG.
The applications I used were OpenOffice, Kword, Kspread, Abiword, Gnumeric, Gimp, and Mozilla. (Not every application made sense for each part of the test).
For the table cells, no pair of spreadsheet apps was able to communicate over the clipboard. Gnumeric could detect that there was data on the clipboard placed by someone else, but couldn't read it. The others just silently ignored it.
For the bitmapped image, no pair of applications could transfer it over the clipboard between them (even though the recipent was able to capable of displaying the image if I loaded it from disk). OpenOffice never acknowledged that external apps had filled the clipboard at all. Gnumeric, once again, offered to insert the graphic as hexadecimal junk. Abiword would occasionally crash on the Paste command, but normally just did nothing. Surprisingly, not even Kword and Kspread were compatible in this way.
The only time it appeared that an image was transferred on the clipboard was from Mozilla to OpenOffice. However, that turned out to be an illusion, because Mozilla supplied a URL that OpenOffice automatically loaded. The only data exchanged on the clipboard was still pure ASCII.
A lot of apps only support ASCII, but that's not X11's problem.
Almost any shortcoming in X11 can be fended off by saying "That's not an X11 problem, it's outside the scope of the design". That's not a valid defense. Complaints like that are a sign that although the design may have been implemented acceptably, the design itself is flawed; the scope was too narrow to account for what's needed by a GUI system.
The practical proof of the quality of an arch is the applications that run on it. And practically, non-ASCII clipboard contents in X11 suck. I've just pulled several major desktop apps off Debian Linux and tested their clipboard compatibility.
You can't paste spreadsheet rows between OpenOffice and Gnumeric. You can't copy an image from Gimp and paste it into OpenOffice, Abiword, Kword, or Gnumeric. (Of those 3, only Gnumeric will recognize that the clipboard has anything in it, but it reads it as ASCII junk). An image will copy from Kword to Kword, but not into Gimp, OpenOffice, or Abiword. For every pair of apps I've tried in the past 10 minutes, none could paste an image between each other. Even copying a picture from Kword to Kspread doesn't work! (You'd think that those applications were written by the same team, and would share clipboard protocols)
The only successful copying of images between different applications that I discovered was, rather bizarrely, from Mozilla into OpenOffice. That one surprised me. (Mozilla to Kword does nothing. Mozilla to Abiword crashes)
Your definition of "available" is provincial. 640x400+ Zauruses have been sold in Japan for almost a year. (And they're not difficult to import into the US)
wouldn't you have to run some sort of firewall on each individual machine, rather than just the gateway/router?
No. The questions of whether computers on a LAN have their own IP addresses and whether they are firewalled by a dedicated box are independent. Even if each machine has an IP address by which it is publically addressable, you can still have a system which protects it by blocking known-dangerous ports.
The advantage of a situation like that, for instance, would be that you could have the firewall block file-sharing/RPC ports, while still allowing port 80 inbound so the individual machines can run webservers. With a NAT, only one local system could have a webserver, and you'd have to configure which one got it on the firewall.
No. That is wrong. Completely, utterly, dangerously, stupidly, flat out WRONG.
No, you are wrong. If you want to be formal about it, go find a good philosophy book and you can learn that there are at least 5 variantions of the ad hominem technique, not all of them fallacious.
Ad hominem is a fallacy. It can never be a logical complaint.
It's only a fallacy in the restrictive world of facts that are so simple all interested parties can trivially understand them.
In the real world, there is more complexity out there than one human brain can handle. We must rely on others to help us think through problems we can't handle ourselves. And when deciding whether one person's opinion should be trusted, the rightness of his other opinions can be a valid predictive factor.
* Vegetable has 2 common meanings. One of them includes tomatoes. When a person claims that is incorrect, he is wrong.
* Fruit has a common meanings (culinary) and an uncommon one (botanical). The latter includes tomatoes, the former does not.
So, pray tell good sir, if both botanists and the common layman believe fruits and vegetables to be distinct, why should they be anything but?
The fact of the matter is that in the common use of English, tomatoes are not fruits. So pray tell, if the common layman and the skilled chef alike believe tomatoes to be vegetables, why should they be anything but?
Actually no. My definition is from a psychologist, whose field was the first that studied "intelligence". Because "intelligence" was such a controversial thing to define, they choose a minimal, inarguable definition so that research results could be unambiguously compared.
The Turing test is not used by many (any?) scientists yet, not even CS or AI types. And it's not intended to measure intelligence either... all it could really present is a boolean guess.
A real intelligence test is conducted by a dedicated psychologist, and has a greater range of values the scan-o-matic forms you may be used to. It's capable of measuring, supposedly, low intelligences like a dog (15 IQ) or venus flytrap (1 IQ). (And yes, there are some humans who score that low) As part of the process, the examiner will converse with the subject, thus subsuming the activities of a Turing test.
Slashdot Mirror
Sadly, ALSA has made the rotten design decision to put compatibility over functionality with its OSS /dev/dsp emulation.
/dev/dsp will face the same problem as if OSS had been driving the sound. They say "To perfectly emulate OSS, we've got to include the limitations!". Bug-for-bug compatibility...
The specific problem with a Soundblaster16 in Linux (as opposed to Windows(r)) is that it'll only let one program play sound at a time. Others are blocked most inelegantly (a nonrefreshing X11 window is often your only clue that XMMS must be stopped before Mozilla can continue).
ALSA has decided to continue the same bad behavior, so that programs using
The 2.6 switch to ALSA is a good thing, but it'll take many years to really help. Only once 3rd party developers (Macromedia, I'm talking to you) stop using OSS interfaces will ALSA truely help against this problem.
Fine, I'll give you a hug. They say vat-born adolscent albino clones need love more than anyone...
(Oh, I shouldn't have respond to that. I really hope slashdot isn't becomng the kind of forum where users deluge each other with HUGS and "Have a COOKIE!". At least we can't embed dancing smilies in animated GIFs)
Yes, that's flaky. But you discovered a workaround by plugging it back into the case. My IO-killing mouse terminates Linux systems no matter how it's attached.
It's a legal requirement to ensure I'm non-infringing on intellectual property.
In KDE, go to the control panel, select Desktop->Size & Orientation and pick your screen size and refresh rate from the pick lists.
KDE has nothing called a "control panel". There is a Control Center, but that has nothing called "Size & Orientation". I'm sure I could get it working with some effort at configuring XFree86 extensions, but RandR is new and not yet something likely to be found on a typical Linux workstation.
Until we have firewire (or similar) video cards/monitors, this seems even less of an issue than changing color depth (which is also pointless
It's not. This problem happens, I've seen it needed on Windows(r) machines. Here's some ways this occurs:
don't you just set it at 24 or 32-bit and leave it alone?
No, there are many places this can be useful. Once again, there is the initial selection of good settings, which would be nice to accomplish from something like Control Center, without constantly killing X11. (This harkens to the Linux folk's stereotypical obsession with uptime). There are pro-artist applications that can benefit from changing bitdepth. But most importantly, changable bitdepth empowers reheadability. That would allow a proxy X server to move an application from one $DISPLAY to another, without imposing a long-term performance hit on either. It saddens me to see people deploy VNC just so they can reconnect to a running X11 application from multiple X11 servers.
And the multi-session capability of KDE
That's hardly a feature of KDE. More like an ancient XFree86 capability that KDE just recently added a widget to activate. Some more integration from KDE would be nice, such as providing a GUI list of other activate sessions, rather than requiring a user to guess around about Ctrl-Alt-F8, etc. The static CRT-snap which happens even going between sessions of identical monitor configuration is minorly irritating, too.
All I am saying (and really it was only a throwaway pedantic comment) is that from a dictionary definition of the word, I don't believe MS has a monopoly.
Interpeting it as you have, no company has ever held a monopoly. Absolute, 100% "exclusiveness" can never happen in the real world.
Bttz bzzt. My employer got a G4 15 months ago, and we had to go order them separately... (or download ISOs). I'll believe my own eyes over you.
The simple way of looking at the "Innovator's Dilemma" is to think about the danger of self-cannibalization. (Oops, that sounds equally obtuse!)
What it means is that the people best able to come up with a new/better/cheaper solution to a problem are often those people who work professionally on old/worse/expensive methods. Beginning to sell the improved product may have you competing with yourself. Like supposedly General Motors can make superior cars that never wear out, but chooses not to so they can earn money from the current ones.
The Innovator's Dilemma is that companies faced with a new idea have a disincentive to bring it to market if it will undercut a currently profitable market. But- they've got to market it sometimes, because if anyone is going to undercut their sales, it should be them. So there's a game of holding back new ideas until just before you think someone else will sell them.
The ID describes how capitalism- the free market that supposedly drives corporations to serve consumer needs- can sometimes encourage them to hold back.
An example of the effect specific to Operating System vendors (and the Linux Distribution market in particular) is the question of support fees. RedHat's biggest revenue source (even moreso, with the recent retreat from retail selling) is customer support contracts. But if the usability of the software itself were improved, then the customer's need to buy support would go down.
So Linux vendors are in the awkward position of profiting by not improving their product. They still have a drive to improve, because having the "most usable" distrib will suck up market share from the others. But improving software quality is not an inarguably healthy idea for them.
Digital Cameras/Scanners... you know it's going to work if you don't buy something that's a cheap piece of shit, which, believe it or not, will suck A WHOLE LOT compared to something 25% more expensive.
Unfortunately, the 25% more expensive figure is off the mark. I'd like to have a scanner around for emergencies, but it's tough to find a USB model in the SANE list for less than $175. However, each December, Staples and BestBuy put the flimsy no-name scanners on sale for $19.95 each ("$9.95 if you come in before 8 AM!")
I'd buy one, if it'd work with Linux (it won't). And I'd actually prefer a small, light machine over a heavy metal box.
Cameras - If you get one with a REAL optical zoom, almost guaranteed it speaks PTP or is a USB mass storage device in disguise, so there's no issue.
And really, savy users will just buy a $12 USB-Flashcard adapter, to preserve battery life when transferring the photos (faster too).
They just don't serve the $400 market.
Ergo, they are more expensive. And remember that the needs of 80% of computer users can be handled by the $250 market.
Macs come with a full unix development environment.
That started less than 2 months ago. OS X 10.2 didn't ship with compilers, unless you ordered them specially.
But Joe Dummy would hardly risk installing hardware himself, would he?
Yes he will. He'll install the hardware following the guidebook step by step, and then Windows(r) will "Detecting new hardware" on startup, and take him through 9 dialog boxes, ask for the system disks, and fill the Device Manager with red exclamation points.
Then he'll get on the phone to the hardware manufacturer, and the nice guy on the other end will tell him precisely what mouse clicks will make it work. If he hadn't been using the same OS as 96% of all customers, that recourse wouldn't have been there to save him.
But if someone had set me up with Linux when I was a novice I am sure I would find it easier to install hardware in Linux now.
Not really. There's a whole separate category of problems Linux users can run into, which simply don't happen to people using a majority OS. When buying hardware, it's completely safe to assume that everything in CompUsa is compatible with a Windows/Intel platform, unless the box specifically says "Macintosh". And if you need confirmation, the clerk will absolutely be able to tell if the product is supposed to be Windows compatible.
With Linux, there's no guarrantees like that, so even choosing new hardware is tricky. Most packaging says nothing about whether it's Linux compatible, although much hardware is. Rarely labels announcing Linux compatiblity can be seen, but they sometimes refer only to one specific 2-year old RedHat distribution.
So there's the whole difficult problem of "Can it work?", before you even start to answer "How can it work?". And there are many products that have no hope of Linux compat. Winmodems (except for one Lucent chipset), the SoundBlaster 16 (and compatible) audio cards, Intel's Centrino wireless networking... the list goes on and on.
Just look at me: I'm using a USB mouse at my Linux workstation now. This year I bought a second USB mouse. It works fine in Microsoft(tm) Windows(r) systems. But plug it into Linux, and all USB devices stop responding until you reboot. Apparently there are subtle differences between mice that can really trip up an OS. Who knew?
PS. Technically the Soundblaster 16 can be made to output sounds under Linux. But the restrictions imposed by the driver design render it painfully less capable than the same card in Windows.
Which of the above deficiencies do you think will be corrected first? My money's on XFree86.
Pay up. Although Windows(r) installers always ask the user to reboot, they don't require it. And some self-contained programs work just fine without the reboot.
I can certainly find ONE example of a Windows program installing without a reboot. Can you show me more than ZERO ways to change the depth of an XFree86 display without killing the running applications?
(For that matter, I'd love to see someone even change the dimensions of a live X11 server. Supposedly there's an extension that makes this possible, but I could never get it to work. And would it allow you to add or remove monitors on the fly? We can only hope)
(And no, Control-Alt-Plus doesn't change the server's dimensions...)
The bold font is to make sure respondants like you read at least the most important parts of what I said, but I see from the reply it didn't work.
I'm sorry but you are displaying your total ignorance of the design of X and making completely unjustified and incorrect criticisms in
I'm displaying an ability to make logical inferences. If the authors of StarOffice, OpenOffice, Mozilla, KDE, Gnumeric, and Gimp were all unable to comprehend the X mechanism for clipboard exchange, then it's safe to say that mechanism is either too difficult or too weak to use in a major application. Or are you claiming that the developers of all those projects are stupid and too lazy to make a compatible product?
For a rebuttal, just point me to me one or two famous X11 applications which make correct use of this powerful clipboard, because I can't seem to find any my own.
Please read the design documentation for X before you post about X again.
Frequently X11 defenders fall back to this line, failing to understand that an inability to be comprehensibly documented is itself a design flaw. I have in fact read hundreds of pages of X11R6 documentation, a painful ordeal I have no stomach to repeat soon.
is really as simple as following the very clear guidelines in the excellent explanation by Jamie Zawinski written many years ago.
That's offtopic. The only thing JWZ explains is how to avoid confusing users about "Am I merely selecting text? Or copying it to the clipboard, overwriting previous clipboard contents?" That is not the issue I am complaining about. He only mentions the issue of supporting non-ASCII datatypes in a brief "extra credit" section on content negotiation, where he provides no explanation beyond a reference to some emacs source code. (Humorously, another poster in this thread has held up emacs alongside xterm as examples of major X11 programs that implement the clipboard wrongly)
PS. "Shigoto" means paid work. If someone were paying me, I'd be willing to slog through the ICCCM. For your jibe, I think gekimu would be more appropriate. Although it's attractive to consider arubaito, which (to non-Japanese) brings with it the hyperbolic image of toiling in a death camp.
making X completely irrelevant
Irrelevant? Oh, I wish. Then it would go away, and a new, better display system would be created.
But no, X will stay relevant, and it won't go away. It'll stay because although imperfect, it is good enough
the almost infinite configurability of X is a strength, not a weakness.
Too bad it's not configurable enough to change color depth while programs are running. A feature Microsoft(tm) has supported since 1996 or so, and which Macs probably had much earlier.
No, it's not the App's fault. An Operating System (which is anything that provides services so programs can run on top of it, and that includes X11) can be justifiably blamed if all of the applications built on top of it share a problem. In fact, it's the only reasonable thing TO blame.
If a user complains that Windows95 always crashes, do we allow Microsoft to get away with blaming the application writers? No! If all/most of the apps on a platform exhibit a problem, then the platform is flawed.
Now, what exactly is the flaw with the X11 clipboard that lead the multiple smart people from Gnome, KDE, Qt, and Motif to be unable to use it correctly? I don't really know. Maybe it's simply too hard to use. That's a valid flaw. I don't have time to completely research the interface, but the JWZ article you indicate contains a familiar keyword: "The content negotiation mechansim is very powerful"
Whenever a programmer says his system is "very powerful", that's often a euphenism for "excessively generic". It can mean the developer went to the point of supplying enough functionality so that any required task was technically possible, but didn't spend any time making it straightforward to use.
, it's not enough to pass on objects from application to application, unless both applications are fully aware of the type of information the object contains.
And in the case of 24-bit 2d images, each application is fully aware of how to handle data like that. I verfied by using each program to load a JPG. OLE is not relevant there. In the case of the spreadsheet, or just formatted text, it's not needed either.
experiments with raytracing 2D scene graphs.
Weird. Is it really "ray tracing" when all the rays are co-linear and striking surfaces perpendicular to them? Sounds related to a software z-buffer implementation.
This approach eliminates the need to compute intersections, unions, etc..
IMO, computation of intersections and unions is typically an optimization attempted to accelerate the otherwise slow performance of per-pixel depth checking.
Scalability concerns: in the future, display resolutions will increase so that exponentially more pixels exist (and more rays are needed). Also, the complexity of onscreen graphics can be expected to increase so there are more primitives out there. RT methods will be really hurt by the pixel increase, while higher-level geometric calculations only care about the number of primitives. And, more primitives hurts RT too, as each ray has a larger set of potentially intersecting objects to check for.
but I am not aware of any such solution.
xmove
The relocatability of X11 applications has been academically studied for a long time, and implementations like xmove are easy to download. But their shortcomings are numerous, as you'll quickly notice by giving it a try. The existence of the pseudo-server is the most obvious shortcoming, as it means there will be a performance hit for the entire life of the application, even though you'll only rarely want to move an X11 app from one display to another.
If X11 supported re-heading deeper in the protocol, then not only would transferring operation be cheaper, but we also might be able to restart XFree86 on a Linux system without killing all the applications connected to it.
We have one, it works and all new applications use it.
That is false. So completely false, I don't know how to begin. It's apparently based on the erroneous premise that ASCII text is the only thing that matters on a clipboard.
I just conducted a 20 minute study with recent versions of six high-profile X11 applications, as found in Debian Linux. I looked at two categories of data: 3 cells from a table ("spreadsheet"), and a 24bit photo imported from JPEG.
The applications I used were OpenOffice, Kword, Kspread, Abiword, Gnumeric, Gimp, and Mozilla. (Not every application made sense for each part of the test).
For the table cells, no pair of spreadsheet apps was able to communicate over the clipboard. Gnumeric could detect that there was data on the clipboard placed by someone else, but couldn't read it. The others just silently ignored it.
For the bitmapped image, no pair of applications could transfer it over the clipboard between them (even though the recipent was able to capable of displaying the image if I loaded it from disk). OpenOffice never acknowledged that external apps had filled the clipboard at all. Gnumeric, once again, offered to insert the graphic as hexadecimal junk. Abiword would occasionally crash on the Paste command, but normally just did nothing. Surprisingly, not even Kword and Kspread were compatible in this way.
The only time it appeared that an image was transferred on the clipboard was from Mozilla to OpenOffice. However, that turned out to be an illusion, because Mozilla supplied a URL that OpenOffice automatically loaded. The only data exchanged on the clipboard was still pure ASCII.
Conclusion: the X11 clipboard is broken.
A lot of apps only support ASCII, but that's not X11's problem.
Almost any shortcoming in X11 can be fended off by saying "That's not an X11 problem, it's outside the scope of the design". That's not a valid defense. Complaints like that are a sign that although the design may have been implemented acceptably, the design itself is flawed; the scope was too narrow to account for what's needed by a GUI system.
The practical proof of the quality of an arch is the applications that run on it. And practically, non-ASCII clipboard contents in X11 suck. I've just pulled several major desktop apps off Debian Linux and tested their clipboard compatibility.
You can't paste spreadsheet rows between OpenOffice and Gnumeric. You can't copy an image from Gimp and paste it into OpenOffice, Abiword, Kword, or Gnumeric. (Of those 3, only Gnumeric will recognize that the clipboard has anything in it, but it reads it as ASCII junk). An image will copy from Kword to Kword, but not into Gimp, OpenOffice, or Abiword. For every pair of apps I've tried in the past 10 minutes, none could paste an image between each other. Even copying a picture from Kword to Kspread doesn't work! (You'd think that those applications were written by the same team, and would share clipboard protocols)
The only successful copying of images between different applications that I discovered was, rather bizarrely, from Mozilla into OpenOffice. That one surprised me. (Mozilla to Kword does nothing. Mozilla to Abiword crashes)
Your definition of "available" is provincial. 640x400+ Zauruses have been sold in Japan for almost a year. (And they're not difficult to import into the US)
wouldn't you have to run some sort of firewall on each individual machine, rather than just the gateway/router?
No. The questions of whether computers on a LAN have their own IP addresses and whether they are firewalled by a dedicated box are independent. Even if each machine has an IP address by which it is publically addressable, you can still have a system which protects it by blocking known-dangerous ports.
The advantage of a situation like that, for instance, would be that you could have the firewall block file-sharing/RPC ports, while still allowing port 80 inbound so the individual machines can run webservers. With a NAT, only one local system could have a webserver, and you'd have to configure which one got it on the firewall.
No. That is wrong. Completely, utterly, dangerously, stupidly, flat out WRONG.
No, you are wrong. If you want to be formal about it, go find a good philosophy book and you can learn that there are at least 5 variantions of the ad hominem technique, not all of them fallacious.
Ad hominem is a fallacy. It can never be a logical complaint.
It's only a fallacy in the restrictive world of facts that are so simple all interested parties can trivially understand them.
In the real world, there is more complexity out there than one human brain can handle. We must rely on others to help us think through problems we can't handle ourselves. And when deciding whether one person's opinion should be trusted, the rightness of his other opinions can be a valid predictive factor.
* Vegetable has 2 common meanings. One of them includes tomatoes. When a person claims that is incorrect, he is wrong.
* Fruit has a common meanings (culinary) and an uncommon one (botanical). The latter includes tomatoes, the former does not.
So, pray tell good sir, if both botanists and the common layman believe fruits and vegetables to be distinct, why should they be anything but?
The fact of the matter is that in the common use of English, tomatoes are not fruits. So pray tell, if the common layman and the skilled chef alike believe tomatoes to be vegetables, why should they be anything but?
Actually no. My definition is from a psychologist, whose field was the first that studied "intelligence". Because "intelligence" was such a controversial thing to define, they choose a minimal, inarguable definition so that research results could be unambiguously compared.
The Turing test is not used by many (any?) scientists yet, not even CS or AI types. And it's not intended to measure intelligence either... all it could really present is a boolean guess.
A real intelligence test is conducted by a dedicated psychologist, and has a greater range of values the scan-o-matic forms you may be used to. It's capable of measuring, supposedly, low intelligences like a dog (15 IQ) or venus flytrap (1 IQ). (And yes, there are some humans who score that low) As part of the process, the examiner will converse with the subject, thus subsuming the activities of a Turing test.