I second this. I was thinking something like Moz (in my books Phoenix is Mozilla that has been chopped off a tail and a leg but it's growing something else instead) or uMoz (microMoz).
Unfortunately, gila.org is already taken but I wouldn't be surprised if the owner is slahdot member... I think it's important to be able to tell simple domain name to newbies when they ask about an alternate browser. In any case, gila.sf.net wouldn't be that bad either.
...if the line is OK just before a critical operation. But will testing the quality of the line now give enough assurance that this quality will still be met in the middle of the operation
I was thinking the same thing. Imagine some city workers digging somewhere along the line and hitting your fiber with something like excavator. How's any testing that will be done before operation going to provide any safety against something like that? Does it help if this system can for sure tell that "connection is lost" during the critical operation?
First thing to do is to prevent mechanical failure, but I don't see a way to do that. All you can do is to have multiple backups. Like in addition to that fiber connection, you need enough bandwith on some low latency radio system for the full length of communication line or something. And that can provide only one level of redundancy.
Are you suggesting that: x = (new List).Add(1).Add(2).Add(3).Add(4).max()
is a better representation of the concept in your code than: x = max(1, 2, 3, 4)
If you have code like max(1, 2, 3, 4), a simple bugfix would be replacing it with simple 4. Usually, if you have to get max() of a group (unordered) or list you first generate the group or list somehow. The concept here is that the list has the required method for finding the maximum value instead of some random procedure somewhere else.
So the code becomes: /* thelist is of type list and is initialized somewhere */
m = thelist.max()
If you hardcode the list's contents you already know what the maximum value will be. No need to compute it runtime.
Fantastic. I write home banking software for a living, and this is just going to drive us up a wall. Not to get on a high horse (our code is far from perfect, for a variety of reasons, but I'm going to say this anyway), but this version of Mozilla should not have been released with this bug.
If you read comments in bugs #171235 and #172097 it should be clear that as long as you don't have JS support for cookies disabled, cookies should work fine. Unless, of course, the server is b0rken like the one at meine.deutsche-bank.de which expects browser to treat paths "/mod/WebObjects/dbpbc" and "/mod/WebObjects/dbpbc.woa" as equal. Interestingly, IE does this and Mozilla had a similar bug in previous versions which has been fixed in this release (due potential security problems, see comment 23/bug #171235). Clearly the problem is in the server end and not in the Mozilla.
If you're speaking about some different bug that causes problems please specify.
...and try looking at screenshots for a game. Very often, they'll use Javascript to pop open a new window... Mozilla should have an option trap such things and open the gallery in a new tab.
Yes, such a option could be nice. Another one that comes to my mind is that even though I middle click link with href="javascript:..." which normally loads link in the background in a new tab it would do the same as clicking the link with default mouse button. This would be because usually that javascript code does only work in the page context the link was found.
But really, the problem isn't that Mozilla doesn't have such an option but that pages are doing stuff like that. If some sites used to display all images mirrored should mozilla have an option to mirror back all images? Web sites should be fixed instead. It's that thing of fixing the cause, not symptoms.
Is there any real use for window.open() anyway? The only reason my copy of Mozilla supports it even a little bit is because of all those b0rken sites. (Javascript code calling window.open when I haven't clicked anything will receive an exception. Check Tools-Web Development-Javascript Console in your copy of Mozilla.)
I read through that page but I fail to see how concept programming should be any better than OO programming. Sure, with C++ you cannot easily specify methods that take lists but it's not because of OO but because of C++ limitations. IMO, max() should be method of a list or a group and C++ can do that fine with templates. Does anybody grok the Zen of Concept Programming? Is it really any better than OO programming?
the cost to develop an app will always stand before cost of quality of the app
I agree partially. The key concept here is the cost. Sure, I'd love to have source for all the programs I ever use. I still understand that creating that software has taken very much effort and generally companies are selling software at very cheap price. And they're able to do that simply because they can count on selling something more later on. As long as I use commercial software I'll rather use somewhat usable software that costs C bucks instead of somewhat usable software with crippled source (read the article) that costs ten times more.
Yes, it sucks that they implement some trivial changes and sell the result to their customers for full price as version N+1. I still don't understand how this differs from, for example, car manufactures: cars have had gasoline engine, four wheels and a steering wheel for god knows how long. Newer cars may have a radio with cd player, a little quieter sound, somewhat different looks and in best case the engine needs a little less fuel for the same mileage. I'd say those changes are trivial but still people aren't complaining. Yep, I'm aware that manufacturing a car requires material and stuff and should therefore cost full price even though the product is almost similar to previous one. When you buy a car, you pay for manufacturing it, but when you buy a piece of software, you pay for the design. What's the difference after all?
Replying to my own post... I looked around for headphone amplifiers and it seems that be the best amplifier for about $200 range is Musical Fidelity X-can V2 (class A tube headphone amplifier!). Be sure to get V2 because there're some problems with channel separation and noise in the V1. Your husband cannot listen with the amplifier only but at least I would be much happier with a high quality amplifier without headphones than a cheap one with cheap headphones.
With a budget of $200 or less the best choice would be a good pair of headphones. You can get pretty nice sound from some small speakers but with that small amount of money spent to two or more speakers and an amplifier isn't going to result in high quality sound. With any luck, $200 should get you pretty good headphones and an amplifier for those. And be cheap about the headphones and use more money on the amplifier because it'll probably last longer anyway. Read some reviews and remember that if the subject is going to wear glasses some cans are going to cause problems (for example, I cannot use any sennheiser setup but I have Sony MDR-F1 instead. Be warned that MDR-F1 is a bit hard for the amplifier and you can ruin the result with a cheap one.
The screenshots of Scale2x look really promising. Looking at two last examples makes me wonder if the same algorithm could be used for pretty much any texture map to get higher quality output from games. Because algorithm needs only a few neighbor pixels to decide correct value for the output pixel it could be implemented in the rendering hardware pretty easily. Using this method with compressed textures should allow pretty nice texturing without using that much memory.
Also, by looking the algorithm on the page it seems to me that this algorithm decides which pixel value to use from left and right only. Running the result through a sligthly modified algorithm could perhaps provide 4x scaling with pretty nice image quality. Simple rotate the table with letters from A to I 90 degrees clock-wise and you should get an algorithm which selects best pixel value from above or below. It might be possible to join those algorithms for a single pass one but I'm afraid the result needs that many conditional jumps that it isn't usable for real time processing. Plus you usually don't need 4x scaling for video.
Scaling animated movie 4x with this algorithm and outputting it through hardware scaler to reduce pixel boundaries should provide pretty nice video quality...
Progressive or interlaced has nothing to do with refresh rate[1]. A display be could called progressive even if its refresh rate were 1 frame per second[2]. In the same way, an interlaced display could display 200 fields per second and therefore be able to display 100 frames per second but it would still be interlaced. That's because term "interlaced" means that every other line is refreshed at once (all the odd lines or all the even lines), usually from the top to bottom and then the other lines are refreshed in the same way. Interlaced does not mean that only pixels that don't change are refreshed. With slow enough phosphor interlaced displays look good with static images but they are inferior to progressive displays when there's any movement. In addition, slow phosphor decreases image quality during movement (like extra motion blur effect without an option to toggle it off).
If specs of plasma or LCD display say that it supports progressive inputs it doesn't mean that it can display the full progressive stream but simply that it can display some kind of image. For example, many new LCD projectors allow progressive VGA signals with 90kHz horizontal sync and 85Hz vertical sync. You can be pretty sure that LCD display isn't able to refresh its pixels at that rate but the spec is only saying that it can sync to the input data. If the display cannot keep up with the data some intermediate pixel values will be skipped and the end result can be pretty close to slow phosphor effects - in the best case.
That being said, plasma displays should be plenty fast for progressive HDTV signals but that's only because HDTV image really isn't that high quality. Any 19" CRT monitor can display much higher resolutions with higher refresh rates than HDTV setup. Don't expect to be able to get high quality image from PC. I suggest previewing the cheapest plasma displays before buying, though. They might have used low quality components to reduce price.
[1] Except that with low quality components it's easier to do a viewable interlaced display than a progressive display.
[2] For example, a slide projector is progressive display.
[...]but every LCD that we've gotten looks
good at ONE resolution[...]
Yes, that's how LCDs work. It has a
"native resolution"[...]
But that's a problem only as long as the native resolution of LCD is too low. Give me an LCD which has native resolution of 2000x1500 or more and I don't mind if it has to interpolate some pixels to display 1024x768. In fact, if the graphics adapters wouldn't suck, everybody would be running maximum resolution their monitor can output all the time and let the adapter handle the scaling. That's practically the same way CRTs display any resolution--it's not like there's infinite number of holes in the shadow mask.
Easier on the eyes? It's called 100Hz refresh rate...
I too think that if somebody prefers LCD because it flickers less is running her CRT monitor with a too low refresh rate.
LCD running it's native resolution is easier to the eyes because all the pixel borders are sharp which makes it easier for eyes to focus. Ever used a blurry CRT monitor for a long time?
If this works with lossless image formats, here's a simple way to achieve the same. I haven't read the paper so this could be practically same they're doing. For simplicity I describe a way to embed 8 bits of data in every pixel in a 24bit image.
Extract some LSBs from every pixel from all RGB components (3 from R, 2 from G, 3 from B) and generate a stream from those bits.
Append your secret message in the stream generated in the previous step.
Compress the stream with any algorithm (e.g. bzip2) and make sure you get size in bytes less than or equal to number of pixels in the original image[1]. Pad with zeros if size of compressed stream is less than number of pixels in the image.
Combine the stream with the image inserting bits from the stream to positions where we extracted bits in the first step and save the image any lossless way to want. The image quality will be reduced roughly to the same quality as 16 bit version would be but the original image can be restored from the compressed substream. During decompression, first there'll equally many bytes of image data as there're pixels in the original image and all the remaining data is the embedded message.
[1] This is possible because all natural images have very little information in the LSBs of every pixel and those should compress well. If the image is truly random down to LSB there's no way any algorithm can embed extra information in those pixels.
This will be probably patented. At least this is a bit more complicated than sideways swinging.
The easy way to show that faster sorts exist is to demonstrate absurd limit case of a tradeoff of space for speed. Consider you have an unlimited amount of memory available for your sort results, and that you are sorting a finite number of keys N for which a mapping M(n) exists to the positive integers. Then, since there can be at most N duplicates of any given key, scanning the data once and placing each key n(i) in memory address N*M(n(i))+i sorts all the data. This is O(N), and pretty much optimal.
Yeah, the sorting is O(N) but reading the result is O(N^2) and without knowing the result the sorting itself doesn't do much good. That's because you need N^2 of memory and in worst case you have to do linear scan through the whole memory. If you think about sorting it should be pretty clear that O(n log n) is best you can get if you have no information about the data and it's uniformly distributed. IIRC worst case for heap sort is O(n log n) and worst case for quicksort is O(n^2) so if you don't need in-place sorting you should use for example heap sorting. If you don't have many items to sort just do insertion sort because it has least overhead.
In Opera 6, you can [...] browse in tabs (long before Mozilla did it)
That's true, but Opera 6 allows only either MDI or SDI for all the windows, not both. Mozilla allows you to open multiple windows (CTRL+N) and open any number of tabs inside those. In addition to that, I have middle mouse button configured so that clicking a link with it will open the link in the background in a new tab in the same window. This allows me to separate different tasks I'm doing with my browser into separate windows: for example, I'm reading slashdot and open all the interesting stories with middle mouse button which results all those stories being loaded in the background in the same window (but none of those will be displayed immediatly). If a friend calls and asks me to check about a movie, I open new window and perhaps google for it and open another tab for imdb. The best part is that when I'm done I simply close the window I used to get information about the movie and I'm left with the slashdot task I was middle of.
And I'm using 10Mbps connection. Fetching multiple pages in the background in paraller increases total throughput of web browser even more if you're using something like modem or GPRS (via cell phone).
The only things Opera does better than Mozilla for me is having a shortcut key for User Stylesheet and option to disable plugins at will by pressing F12. I'm using a lousy javascript hacks instead.
Well, it depends... You really cannot use native widgets for any part of the viewport (the area of a visual browser that displays the page content) if you ever hope supporting full CSS. That's because CSS provides ways to style everything inside viewport. If you think that scrollbars should be themeable by web site like in IE/Win32 then those too cannot be native. So what remains? Menu, address bar, status bar and a couple of buttons (like back, reload etc). Mozilla has XUL to describe those basic interface items in a platform independant way. Isn't that what all the developers are asking for?
XUL doesn't prevent native implementations of those widgets just like GTK doesn't prevent that either. The issue is that doing the actual implementation for a single desktop environment for a single OS is huge work and probably only win32 and MacOS 9 have currently enough users to make that worthwhile. As for the others, we just have to stick with the skins. And remember that skinned UI can be just as fast as native UI as long as both user interfaces look the same (e.g. blitting texture is slower than drawing solid single color rectangles on most hardware) simply because all the stuff the OS's native UI or kernel is doing can be done in the application code too. At least if the OS has required APIs--for example, if you need alpha channel for image composition you have to emulate it under plain X. If the same feature was implemented in the X (making it "native") it could be faster. Current win32 API allows alpha blending and lots of other stuff so you can do all the effects the native widget set does from application code. I'm not saying that win32 is a good API (it's not) but that it exposes much more features than, for example, X/Xlib.
The remaining issue is Feel and if your OS/UI isn't fully supported by some skin that defines Feel or there's native port for you UI then you're out of luck. I haven't seen an operating system that had consistent UI. MacOS is the best in this regard but it's far from perfect.
If the XBox does not scare you, perhaps you should consider a future where all PCs are designed by Redmond.
PCs will be PCs in the future too. It might be that you cannot run Windows on a PC anymore in the future. I really don't understand the difference between this new Microsoft XBox and normal x86 PC if they really decided to offer Windows for XBox only. XBox, even in its current form, has all the processor power the casual office worker needs and if companies can get their boxes for less than 250 euros then they should go for it. But some workers doing 3D stuff and simulations do need more processing power than that XBox can offer. So MS needs to offer multiple versions of this new XBox because companies are not going to pay for features they aren't using--at least I hope so. If this new XBox does have replaceable Xcpu, Xmemory, Xmotherboard and Xpci cards then how it's different from a reqular PC? It's not like we have a single identical bus between first x86 PC and todays PC either and we still consider those as the same architecture.
The only thing to fear is that general use PCs could have really high price tag because all the normal people buy the XBox New Technology (pun intended) instead.
Is it just me, or do seeing the words Exchange and Email in the same sentence make you shiver?
Me too. I first understood that the subject meant that it's possible to steal email addresses for spamming during some protocol handshake with a Exhance server.
10Mbps sounds like a plenty fast for just about anything. Some of you talk about copying DVD's and stuff but if you want to transfer that much information there're better ways. My internet connection has been 10Mbps for a couple of years and I'm still happy with it.
Motion vector de-interlacing: use MPEG motion vectors to determine what parts of the image to bob and what parts to weave. Gives the best results but is compute-intensive and requires the MPEG-2 decoder to output motion vectors.
You can get pretty good results with something as simple as simply comparing pixels near each other. If you do have motion vectors it simply decreases the CPU power required and possibly increases image quality in the high motion scenes. A nice example for windows machines is DScaler which deinterlaces analog TV signals in real time. AFAIK this kind of setup is usually called motion adaptive deinterlacing. I once did it before I knew about DScaler and you can easily do 720x576@25fps (PAL) with a single Duron 650 and even then the biggest problem was that my VIA based motherboard had bad PCI throughput. I think one could take in 720x576@50fields/s interlaced signal and output 720x576@50frames/s progressive with a normal 1GHz+ desktop processor. It all comes down to whether or not you're able to write all the frames to video memory: 720x576@50fps requires that at least 720x576x50x2 bytes are transferred per second (=40MB/s to video ram).
Most sound cards you use won't be able to [output 20 bit audio] either, and that includes ones with SPDIF output. Not to mention, once you encode to Ogg or MP3, all that extra quality goes out the window.
I agree with the output part and in addition not that many consumer amplifiers have inputs for 20 bit digital audio either. However, I think that using 20 bit original for Ogg or MP3 should increase the resulting quality because you have more information to select from to discard unneeded pieces during compression. If you had 20 bit output I'd guess that 20 bit Ogg or MP3 (assuming 20 bit format is supported) would sound better than uncompressed sound dithered to 16 bit.
But then, I don't have audio equipment or ears to hear even full 16 bits. Even NICAM sounds pretty nice to me.
but also it is the creative process, not just the specific tool [...] a pirated copy Word vs using a free tool like OpenOffice [...] If you can create something spectacular with GIMP, you can definately do the same with Photoshop.
IMO, those cases aren't really that close to each other. When you use OpenOffice or MS Word the real work you do is typing words via keyboard. When you use GIMP or Photoshop the UI is everything--any graphics manipulation program with ability to modify unique pixels can be used to make any given spectacular work but the difference between tools is how easy or hard making of it will be.
With OOffice and Word there might be a little different way to specify margin for a paragraph but the real creative work consists of hitting the keyboard which hopefully follows the same keymap for both of the applications. I'm sure you all agree that if you have to use keymap you're not accustomed to your performance suffers greatly.
In addition to different input methods between different graphic manipulation applications there's a change that you can do something in GIMP with a filter and you have to modify 1.5 million pixels by hand to achieve similar result in Photoshop. The difference can be quite huge.
Though, there's the always friendly Clippy in the MS Word that can do the writing for you, right?;-)
Mozilla needs some marketing oriented types instead of more nerds.
Ever heard of company called "Netscape"? Mozilla isn't meant for end users. Quote:
Mozilla is an open-source web browser and toolkit, designed for standards compliance, performance and portability. Mozilla.org provides binaries
for testing and feedback.
M$ is probably going to do this everytime they start a new production run. The end result however will be that they are going to end up with a whole series of slightly incompatable versions of the xbox.
Do you mean the same way Sony did with the PS2? I don't know about other areas but we have 4 different versions in Europe alone. They are called V3, V4, V5 and V6. In addition, I've understood that there might be two different flawors of V3. Latest versions are claimed to be the easiest to mod.
The kind of security you are talking about is just not possible. Consider that if there are dubbed versions to be made (as i the case with LOTR2), several dozen countries all over the world have to receive the movie several months in advance.
That's an interesting theory. I still think that if they wanted security they could send a crippled copy. If the copy is used for dubbing only, shifting colors, black and white image and/or big honking across across the image doesn't really prevent the work but makes sure that the copy will not be pirated--or at least nobody will ever think the pirated copy would be useful at all. The only place they should send full copy is the film reproduction labs or DVD-manufacturer and because in that case that is the only outsider having access to the full material they know who to sue. Once they distribute the first DVD disk there's no way to prevent the pirated copy from being distributed.
Slashdot Mirror
Unfortunately, gila.org is already taken but I wouldn't be surprised if the owner is slahdot member... I think it's important to be able to tell simple domain name to newbies when they ask about an alternate browser. In any case, gila.sf.net wouldn't be that bad either.
I was thinking the same thing. Imagine some city workers digging somewhere along the line and hitting your fiber with something like excavator. How's any testing that will be done before operation going to provide any safety against something like that? Does it help if this system can for sure tell that "connection is lost" during the critical operation?
First thing to do is to prevent mechanical failure, but I don't see a way to do that. All you can do is to have multiple backups. Like in addition to that fiber connection, you need enough bandwith on some low latency radio system for the full length of communication line or something. And that can provide only one level of redundancy.
x = (new List).Add(1).Add(2).Add(3).Add(4).max()
is a better representation of the concept in your code than:
x = max(1, 2, 3, 4)
If you have code like max(1, 2, 3, 4), a simple bugfix would be replacing it with simple 4. Usually, if you have to get max() of a group (unordered) or list you first generate the group or list somehow. The concept here is that the list has the required method for finding the maximum value instead of some random procedure somewhere else.
So the code becomes:
/* thelist is of type list and is initialized somewhere */
m = thelist.max()
If you hardcode the list's contents you already know what the maximum value will be. No need to compute it runtime.
If you read comments in bugs #171235 and #172097 it should be clear that as long as you don't have JS support for cookies disabled, cookies should work fine. Unless, of course, the server is b0rken like the one at meine.deutsche-bank.de which expects browser to treat paths "/mod/WebObjects/dbpbc" and "/mod/WebObjects/dbpbc.woa" as equal. Interestingly, IE does this and Mozilla had a similar bug in previous versions which has been fixed in this release (due potential security problems, see comment 23/bug #171235). Clearly the problem is in the server end and not in the Mozilla.
If you're speaking about some different bug that causes problems please specify.
Yes, such a option could be nice. Another one that comes to my mind is that even though I middle click link with href="javascript:..." which normally loads link in the background in a new tab it would do the same as clicking the link with default mouse button. This would be because usually that javascript code does only work in the page context the link was found.
But really, the problem isn't that Mozilla doesn't have such an option but that pages are doing stuff like that. If some sites used to display all images mirrored should mozilla have an option to mirror back all images? Web sites should be fixed instead. It's that thing of fixing the cause, not symptoms.
Is there any real use for window.open() anyway? The only reason my copy of Mozilla supports it even a little bit is because of all those b0rken sites. (Javascript code calling window.open when I haven't clicked anything will receive an exception. Check Tools-Web Development-Javascript Console in your copy of Mozilla.)
I read through that page but I fail to see how concept programming should be any better than OO programming. Sure, with C++ you cannot easily specify methods that take lists but it's not because of OO but because of C++ limitations. IMO, max() should be method of a list or a group and C++ can do that fine with templates. Does anybody grok the Zen of Concept Programming? Is it really any better than OO programming?
I agree partially. The key concept here is the cost. Sure, I'd love to have source for all the programs I ever use. I still understand that creating that software has taken very much effort and generally companies are selling software at very cheap price. And they're able to do that simply because they can count on selling something more later on. As long as I use commercial software I'll rather use somewhat usable software that costs C bucks instead of somewhat usable software with crippled source (read the article) that costs ten times more.
Yes, it sucks that they implement some trivial changes and sell the result to their customers for full price as version N+1. I still don't understand how this differs from, for example, car manufactures: cars have had gasoline engine, four wheels and a steering wheel for god knows how long. Newer cars may have a radio with cd player, a little quieter sound, somewhat different looks and in best case the engine needs a little less fuel for the same mileage. I'd say those changes are trivial but still people aren't complaining. Yep, I'm aware that manufacturing a car requires material and stuff and should therefore cost full price even though the product is almost similar to previous one. When you buy a car, you pay for manufacturing it, but when you buy a piece of software, you pay for the design. What's the difference after all?
Replying to my own post... I looked around for headphone amplifiers and it seems that be the best amplifier for about $200 range is Musical Fidelity X-can V2 (class A tube headphone amplifier!). Be sure to get V2 because there're some problems with channel separation and noise in the V1. Your husband cannot listen with the amplifier only but at least I would be much happier with a high quality amplifier without headphones than a cheap one with cheap headphones.
With a budget of $200 or less the best choice would be a good pair of headphones. You can get pretty nice sound from some small speakers but with that small amount of money spent to two or more speakers and an amplifier isn't going to result in high quality sound. With any luck, $200 should get you pretty good headphones and an amplifier for those. And be cheap about the headphones and use more money on the amplifier because it'll probably last longer anyway. Read some reviews and remember that if the subject is going to wear glasses some cans are going to cause problems (for example, I cannot use any sennheiser setup but I have Sony MDR-F1 instead. Be warned that MDR-F1 is a bit hard for the amplifier and you can ruin the result with a cheap one.
Also, by looking the algorithm on the page it seems to me that this algorithm decides which pixel value to use from left and right only. Running the result through a sligthly modified algorithm could perhaps provide 4x scaling with pretty nice image quality. Simple rotate the table with letters from A to I 90 degrees clock-wise and you should get an algorithm which selects best pixel value from above or below. It might be possible to join those algorithms for a single pass one but I'm afraid the result needs that many conditional jumps that it isn't usable for real time processing. Plus you usually don't need 4x scaling for video.
Scaling animated movie 4x with this algorithm and outputting it through hardware scaler to reduce pixel boundaries should provide pretty nice video quality...
Progressive or interlaced has nothing to do with refresh rate[1]. A display be could called progressive even if its refresh rate were 1 frame per second[2]. In the same way, an interlaced display could display 200 fields per second and therefore be able to display 100 frames per second but it would still be interlaced. That's because term "interlaced" means that every other line is refreshed at once (all the odd lines or all the even lines), usually from the top to bottom and then the other lines are refreshed in the same way. Interlaced does not mean that only pixels that don't change are refreshed. With slow enough phosphor interlaced displays look good with static images but they are inferior to progressive displays when there's any movement. In addition, slow phosphor decreases image quality during movement (like extra motion blur effect without an option to toggle it off).
If specs of plasma or LCD display say that it supports progressive inputs it doesn't mean that it can display the full progressive stream but simply that it can display some kind of image. For example, many new LCD projectors allow progressive VGA signals with 90kHz horizontal sync and 85Hz vertical sync. You can be pretty sure that LCD display isn't able to refresh its pixels at that rate but the spec is only saying that it can sync to the input data. If the display cannot keep up with the data some intermediate pixel values will be skipped and the end result can be pretty close to slow phosphor effects - in the best case.
That being said, plasma displays should be plenty fast for progressive HDTV signals but that's only because HDTV image really isn't that high quality. Any 19" CRT monitor can display much higher resolutions with higher refresh rates than HDTV setup. Don't expect to be able to get high quality image from PC. I suggest previewing the cheapest plasma displays before buying, though. They might have used low quality components to reduce price.
[1] Except that with low quality components it's easier to do a viewable interlaced display than a progressive display.
[2] For example, a slide projector is progressive display.
But that's a problem only as long as the native resolution of LCD is too low. Give me an LCD which has native resolution of 2000x1500 or more and I don't mind if it has to interpolate some pixels to display 1024x768. In fact, if the graphics adapters wouldn't suck, everybody would be running maximum resolution their monitor can output all the time and let the adapter handle the scaling. That's practically the same way CRTs display any resolution--it's not like there's infinite number of holes in the shadow mask.
I too think that if somebody prefers LCD because it flickers less is running her CRT monitor with a too low refresh rate.
LCD running it's native resolution is easier to the eyes because all the pixel borders are sharp which makes it easier for eyes to focus. Ever used a blurry CRT monitor for a long time?
[1] This is possible because all natural images have very little information in the LSBs of every pixel and those should compress well. If the image is truly random down to LSB there's no way any algorithm can embed extra information in those pixels.
This will be probably patented. At least this is a bit more complicated than sideways swinging.
Yeah, the sorting is O(N) but reading the result is O(N^2) and without knowing the result the sorting itself doesn't do much good. That's because you need N^2 of memory and in worst case you have to do linear scan through the whole memory. If you think about sorting it should be pretty clear that O(n log n) is best you can get if you have no information about the data and it's uniformly distributed. IIRC worst case for heap sort is O(n log n) and worst case for quicksort is O(n^2) so if you don't need in-place sorting you should use for example heap sorting. If you don't have many items to sort just do insertion sort because it has least overhead.
That's true, but Opera 6 allows only either MDI or SDI for all the windows, not both. Mozilla allows you to open multiple windows (CTRL+N) and open any number of tabs inside those. In addition to that, I have middle mouse button configured so that clicking a link with it will open the link in the background in a new tab in the same window. This allows me to separate different tasks I'm doing with my browser into separate windows: for example, I'm reading slashdot and open all the interesting stories with middle mouse button which results all those stories being loaded in the background in the same window (but none of those will be displayed immediatly). If a friend calls and asks me to check about a movie, I open new window and perhaps google for it and open another tab for imdb. The best part is that when I'm done I simply close the window I used to get information about the movie and I'm left with the slashdot task I was middle of.
And I'm using 10Mbps connection. Fetching multiple pages in the background in paraller increases total throughput of web browser even more if you're using something like modem or GPRS (via cell phone).
The only things Opera does better than Mozilla for me is having a shortcut key for User Stylesheet and option to disable plugins at will by pressing F12. I'm using a lousy javascript hacks instead.
Well, it depends... You really cannot use native widgets for any part of the viewport (the area of a visual browser that displays the page content) if you ever hope supporting full CSS. That's because CSS provides ways to style everything inside viewport. If you think that scrollbars should be themeable by web site like in IE/Win32 then those too cannot be native. So what remains? Menu, address bar, status bar and a couple of buttons (like back, reload etc). Mozilla has XUL to describe those basic interface items in a platform independant way. Isn't that what all the developers are asking for?
XUL doesn't prevent native implementations of those widgets just like GTK doesn't prevent that either. The issue is that doing the actual implementation for a single desktop environment for a single OS is huge work and probably only win32 and MacOS 9 have currently enough users to make that worthwhile. As for the others, we just have to stick with the skins. And remember that skinned UI can be just as fast as native UI as long as both user interfaces look the same (e.g. blitting texture is slower than drawing solid single color rectangles on most hardware) simply because all the stuff the OS's native UI or kernel is doing can be done in the application code too. At least if the OS has required APIs--for example, if you need alpha channel for image composition you have to emulate it under plain X. If the same feature was implemented in the X (making it "native") it could be faster. Current win32 API allows alpha blending and lots of other stuff so you can do all the effects the native widget set does from application code. I'm not saying that win32 is a good API (it's not) but that it exposes much more features than, for example, X/Xlib.
The remaining issue is Feel and if your OS/UI isn't fully supported by some skin that defines Feel or there's native port for you UI then you're out of luck. I haven't seen an operating system that had consistent UI. MacOS is the best in this regard but it's far from perfect.
PCs will be PCs in the future too. It might be that you cannot run Windows on a PC anymore in the future. I really don't understand the difference between this new Microsoft XBox and normal x86 PC if they really decided to offer Windows for XBox only. XBox, even in its current form, has all the processor power the casual office worker needs and if companies can get their boxes for less than 250 euros then they should go for it. But some workers doing 3D stuff and simulations do need more processing power than that XBox can offer. So MS needs to offer multiple versions of this new XBox because companies are not going to pay for features they aren't using--at least I hope so. If this new XBox does have replaceable Xcpu, Xmemory, Xmotherboard and Xpci cards then how it's different from a reqular PC? It's not like we have a single identical bus between first x86 PC and todays PC either and we still consider those as the same architecture.
The only thing to fear is that general use PCs could have really high price tag because all the normal people buy the XBox New Technology (pun intended) instead.
Me too. I first understood that the subject meant that it's possible to steal email addresses for spamming during some protocol handshake with a Exhance server.
10Mbps sounds like a plenty fast for just about anything. Some of you talk about copying DVD's and stuff but if you want to transfer that much information there're better ways. My internet connection has been 10Mbps for a couple of years and I'm still happy with it.
You can get pretty good results with something as simple as simply comparing pixels near each other. If you do have motion vectors it simply decreases the CPU power required and possibly increases image quality in the high motion scenes. A nice example for windows machines is DScaler which deinterlaces analog TV signals in real time. AFAIK this kind of setup is usually called motion adaptive deinterlacing. I once did it before I knew about DScaler and you can easily do 720x576@25fps (PAL) with a single Duron 650 and even then the biggest problem was that my VIA based motherboard had bad PCI throughput. I think one could take in 720x576@50fields/s interlaced signal and output 720x576@50frames/s progressive with a normal 1GHz+ desktop processor. It all comes down to whether or not you're able to write all the frames to video memory: 720x576@50fps requires that at least 720x576x50x2 bytes are transferred per second (=40MB/s to video ram).
I agree with the output part and in addition not that many consumer amplifiers have inputs for 20 bit digital audio either. However, I think that using 20 bit original for Ogg or MP3 should increase the resulting quality because you have more information to select from to discard unneeded pieces during compression. If you had 20 bit output I'd guess that 20 bit Ogg or MP3 (assuming 20 bit format is supported) would sound better than uncompressed sound dithered to 16 bit.
But then, I don't have audio equipment or ears to hear even full 16 bits. Even NICAM sounds pretty nice to me.
IMO, those cases aren't really that close to each other. When you use OpenOffice or MS Word the real work you do is typing words via keyboard. When you use GIMP or Photoshop the UI is everything--any graphics manipulation program with ability to modify unique pixels can be used to make any given spectacular work but the difference between tools is how easy or hard making of it will be.
With OOffice and Word there might be a little different way to specify margin for a paragraph but the real creative work consists of hitting the keyboard which hopefully follows the same keymap for both of the applications. I'm sure you all agree that if you have to use keymap you're not accustomed to your performance suffers greatly.
In addition to different input methods between different graphic manipulation applications there's a change that you can do something in GIMP with a filter and you have to modify 1.5 million pixels by hand to achieve similar result in Photoshop. The difference can be quite huge.
Though, there's the always friendly Clippy in the MS Word that can do the writing for you, right? ;-)
Ever heard of company called "Netscape"? Mozilla isn't meant for end users. Quote:
(emphasis mine)Do you mean the same way Sony did with the PS2? I don't know about other areas but we have 4 different versions in Europe alone. They are called V3, V4, V5 and V6. In addition, I've understood that there might be two different flawors of V3. Latest versions are claimed to be the easiest to mod.
That's an interesting theory. I still think that if they wanted security they could send a crippled copy. If the copy is used for dubbing only, shifting colors, black and white image and/or big honking across across the image doesn't really prevent the work but makes sure that the copy will not be pirated--or at least nobody will ever think the pirated copy would be useful at all. The only place they should send full copy is the film reproduction labs or DVD-manufacturer and because in that case that is the only outsider having access to the full material they know who to sue. Once they distribute the first DVD disk there's no way to prevent the pirated copy from being distributed.