Personaly, I still think the correct name for it is Apple Macintosh System Software, as it used to called. True that was freaking hard to write out all of the time, sounds so 1983, and clone unfriendly, but still it's better then MacOS.
Finally, I prefer MacOS over Mac OS. And I like System Software over both of them.
"Slashdot -- The Place to Flame and be Flamed" --from linuxppc.org/links.html
Thanks,
Andrew B. Arthur aka AArthur arthur99@global2000.net AIM: Arthur998
I disagree... Most opensource authors (and users) try to insure the product is as secure as possible. Look at various security holes in Linux kernel for example -- they were fixed as soon as they were found. (That's were the expersion "Security in Internet time" comes in).
It would be hard to prove that Microsoft *doesn't* fix bugs/security holes quick enought. After all, they are commerical -- how long should it take them to recall a defective product?
Unless you can prove that either Microsoft or any Linux product spefically ignores a certain security holes, they can not be sued.
For example, they might be able to try to sue Microsoft over the Melissa virus, since Word Macros are a well known security flaw [they run with virtually no protection/sandboxing features (such as disabled file I/O), besides a macro warning box], which Microsoft should have done something about.
That's basically the same kind of like GM purposely (and knowing it) shipping defective seat belts, causing the death of 23 people in the past year. Of course GM would get sued for it, and the sueing party would likely win. Of course if GM didn't know about the problem -- it would be quite unlikely that any law suit against them would be successful.
It's not a price issue -- we already know that GPL'd software has to follow the law, eventhough it's GPL'd (which makes stealing parts of commerical code or reverse engineering illegal or infriging on trademarks). Recently the author of gaim, the gnome aol instant messager was required to remove "AOL" logos from his product. So he had to follow the law.
So basically the same laws apply to closed source and opensource projects.
One thing to note: OpenSource projects would be *MUCH* harder to sue, since you could easily claimed that the user:
1) didn't get the lastest updates / didn't read bugtraq
2) didn't fully inspect / test the code before installing (since he had everything the author had).
But still, the author would be responsble if the problem was blatly obvious (such as every Caldera employee knew that every 99th CD would distroy the user's Windows partition or the source code proved that this features was spefically coded in by a Caldera employee). This would be a definate lawsuit. However if this wasn't a known bug, the author would not be repsonble.
***So does Microsoft really know that Macros are a real security hole (or just a great easy to use feature that as a side effect makes viruses easy to write), and that they are ignoring to fix it?
***Well, we will leave it up to the lawyers to decide. Obviously, everybody has there own opinions.
Thanks,
Andrew B. Arthur aka AArthur arthur99@global2000.net AIM: aarthurppc
I really hope, this could lead to inexpensive PowerPC-based PCs, even cheaper then some of Apple's cheap machines-- (the iMac [at us$1299] and the iBook [at us$1599]).
Other PowerPC hardware is much more expensive.:-( This is too bad, since the PowerPC platform is well designed and put together, even if it has some faults.
Yes, that is really true. Both Mandrake and Rasterman have really fast machines since they need them--they compile (at least parts) of Enlightenment and other pieces of software daily.
Very few people, who have old slow machines update CVS daily, much less recompile E every few minutes. You really don't want to try this on your typical PC, it will drag it down too much.
If your work requires it, you can afford those nice high tech tools.
(Although, they do recommend that you test all of your software reguluarly on the minium supported configuration, it's stupid to waste time on those machines compiling it on them).
Well, actually for several years Apple has owned cinepack (they bought it in the early 1990s), but it isn't as strictly controlled like Sorsen.
You see, we really don't need the full quicktime program, just a binary codec. We have Xamin which seems to support Cinepack movies to a limited degree of success.
Apple Cinepack is still under a binary license, and I don't see the problem with them licensing them to us.
Actually, they told us they would port the full quicktime to Linux *if* we payed them several thousand dollars. Eventually, if they think they can benfit from Quicktime on Linux x86, they will do it.
Lets not forget the opensourced Mac-On-Linux program, that lets several PowerPC-OS actually run in a seprate frame-buffered console.
Personally, I have seen it boot both LinuxPPC (watching LinuxPPC using BootX boot on top of LinuxPPC is cool), MkLinux (it's sometimes helpful for testing stuff), and Mac OS 8.6. It emulates a 604e PowerPC on most machines, and it runs at close to native speeds (maybe 3 or 4% slower, but that it). Video speed is decent, if you use the MOL video driver. Ethernet works with some cards.
Things it still needs--serial port support, support for more Macs without using rom-images, real SCSI support and other.
And yes it is GPL'd, and you can use the source or download it for free. And it is avalible now. SheepShaver will be out this summer.
Shareware in my experience tends to be done either quickly or by people that have little talent at what they code. If these people had real talent, how come they haven't been hired to write software professionally?
I really don't want a 16 year old kid or somebody who has remedial coding skills to be writing closed source software, since we really don't know what or how he makes it work. Finally, much of the shareware software I have seen is quite buggy (such as ircle--it doesn't crash often on a good day).
Portablity:
Another, disadvantage to closed source software--if the author doesn't port it, you have no chance to get it on your archutechure. So if you are running something like Solaris or Linux/Alpha or Linux/PPC and you want to run it, you are out of luck.
Improving Coding Skills:
If you are going to showcase your coding skills to the world, it forces you to write good code, since everybody in the world will see it. It also makes you more competive to everybody else--you end up making better code.
Expensive:
Paying shareware fees can freaking add up quickly. $10 + $25 + more and more adds up to big bucks, trust me.
Linux Serial Numbers Guides:
If Linux gets flooded by bad shareware titles, how long to we get something like "Penguin Cracks" or "Linux Serial Numbers". Hell, I am sure their is them already, but I would think the guides are pretty small. And soon we will have warez sites. Lame.
Payback to Author:
It's a fact that less then 5% of downloads of shareware actually get payed for. People just run shareware unregistered or crack it (trust me, anybody with two months experience on a computer can crack a registration system on several shareware programs).
What does Shareware Make in Money:
Lets say a program super-shareware-draw pro is $10 for the program (a reasonable fee for the program). 1,000 people download if from either download.com or your homepage at kagi.com. Just for example, 1 out of 20 people register it (a reasonable number). Finally, about 4/10 of the people discard the program after downloading (that's also reasonable).
Do this on a caclualator:
10 * ((1000 *.4) / 20) = $ 200 (maybe)
If you get $200, you can consider your program to be a success. Of course, $200 to write a great application isn't really worth it in most cases -- you won't make $200 up front. It might take several months to get that full $200 in, your money will trickle in every one and a while at $10 a check.
Maybe shareware pays off for some, but going the gpl route would be in general a more rewarding route--if you get 1,000 downloads you know that like over 600 people are using your great piece of software to improve their lives. Also, if you write a good populuar piece of software, and people love it and enjoy it, that person who uses it could be future boss or something.
Shareware or GPL?
Depends on your situation. You want something you can make a little money off while you code, do it. However it most the time makes more sense to get a better paying job, such as working in a college libary or something.
Yes, I do rember when a 68040/40mhz could have preformance close to a 486/50-66mhz, but it would be pushing it to compare to any pentium.
The 68060s may have had pretty good preformance, but since I have never used or seen or seen benchmarks a 68060-based machine I will never know.
At any rate, having a lower clock rate, always looked bad compared to 486s. People would compare a 486/66 to 040/40, and think that the 486 would be like 1/3 faster or so. Of course that wasn't true, but it looked that way to stupid consumers and stupid sales staff.
This was one reason why Motorla pretty much droped the good old 68k out of the line of mainstream processors (not embedded), another was some of those 68040s and highers were kind of like the Pentium III -- power hogs and hot chips. And they were fairly big chips. Bad for both the embedded and PowerBook production.
And yes, the Pentium II and up does more work per cycle then the 68k chips, since they have been hacked up to preform decently.
hmm... those 68060s must be pretty warm and large chips similar to the Pentuim IIs. That was one complaint about the 68040 was they were just slow, hot and uneffectent chips...
The PowerPC did improve on most things, but early PowerPCs were so slow at 68k emulation, that some early PowerMacs actually had hardware working on 68k emulation (not quite a full processor, though)
Just a tadbit to tack on to that is that Linux has pretty good support (or will have) when altivec comes out. gcc will support altivec in the future, you can add support for it today, by getting the patch from motorola.
What does Altivec do? Well, as described by David McEvery, it is MMX implemented correctly. Basically its a set of additional instructions to the standard PowerPC RISC proccessor, adding support for better proccessing of graphical and related stuff. Sounds like MMX? Well it is, in a sense, basically what it has larger and faster processing of things, and implements cache correctly.
hmm... that kind of reminds me of the record changer, if any of you guys are old enought to remember what they do.
Basically, around the late 1970's record changers became unpopuluar (people were afraid they damaged the records and were sonically inferior to regular one disc players), yet I still own one of them and use it daily. The fact that you can not buy record changers anymore stop me from listening to one record after another on the changer, no.
The same thing with mp3s. All current hardware can play Mp3s, and it will be able to do so in the futuire (umm.. xmms-0.91 or later should continue working fine on my LinuxPPC R5 system (even 2 years from now), as long as I don't break it's depencies).
I can see that possibly hardware might have disablable Mp3 players, but on computers, their will always be hacks to get around other programs that might try to block the output from Mp3s.
So how do record from SDMI file to Mp3. Simple. You just copy the analog output of SDMI to your digital recorder. You feed that sound into Mp3 encoder... Yes their is a slight loss of sound quality, but if you use monster gold plated cable, it should be barely audiable if audiable at all.
What about analog watermarking?
From what I have read analog watermarking is total bullshit. Nobody is going to buy recording that have the sound quality audibly degraded just to keep the RIAA happy. People will rather fall back to Tapes and CDs to rip before using that DVD-Audio and SDIA.
What if DVD-Audio is a success with all it's anti-copy stuff?:
You got to be kidding. The first Audio-DVDs are two track with a higher sampling rate / frequency response. Maybe that will improve your dog's hearing of the music or that die hard audiofile that has a sound proofed room with dead slience, but for most people, no. Some DVD-Audio formats, actually degrade the sound, by adding analog (audiable) encyrption stuff.
Mulitchannel DVD-Audio discs. Hmm.. for one they are not comming out tommrow. Another problem with multichannel DVD's is they will not help out recordings recorded in stereo (skip the cheesey surrond effects, please!) Another problem, is most people's stereos are still only two channels, and going to 5.1 channels is expensive. Multichannel DVDs are looking, at least for now to be another RIAA pipe dream.
Actually Mac OS 8.5.x and better (and I think Windows 98), don't have to have jaggy icons. Mac OS 8.5.x supports alpha-chanel transperency and 32-bit icons, but they are a tadbit difficult to edit, since the tools suck to edit them (I use clip2icons and PhotoDeluxe [I cann't afford the real thing, and no I am not booting Linux/GIMP to edit a 32-bit icon!]).
They look nice if you know what you are doing with anti-aliased corners, and stuff.
The reason why NeXTstep does not have jaggys on the icons is that they use tiffs. Tiffs support 32-bit color natively and support alpha-transperency (I think).
Actually, somebody in like 1986 or early-1987 ported that demostation program to the Mac Plus.
It was a pretty cool demo. Some how it used a resolution hack (or at least made it look that way) so you got a high resolution (more then 72-dpi) using the standard Mac Plus video card. It's lines were smoother then smooth, but it seemed to take like a 1 1/2 minutes to initilize the screen at that resolution.
I have seen some pretty cool demos of Super3d on a MacPlus (I still have it), which looked cool, and you could make your own 3d images, but the resolution was far inferior (think 72 dpi) then that amiga clone demo.
If somebody could explain how that demo could create more then 72 dpi on a B&W mac plus, I would be a very happy person.
1) They had to call the trash can garbage. That really is kind of confusing, since your files are typically trashed not garbaged (sounds to much like garbaled, and plus you don't use english that way). If you really must, at least call it something that makes sense like 'Trash' or 'Recycle Bin', etc. Afraid of lawsuit, huh?
2) What's up with this white mouse cursor. Most normal systems use a black curser (X11 and Macs), the only reason to use a white curser is to avoid a lawsuit. At any rate, studies show that having a black cursor == less eye strain, easier to find (most work is done on white paper/white screen).
It's not that easy just to open the win32 API. It might be to the benifit of Linux x86, but it wouldn't go much farther.
Even with Win32 APIs out there, it would be a major challenge to develop a OS from scratch. Since the Win32 APIs don't make any programs up, you would have to write a whole new desktop enviroment from sratch.
For one, it would be usless in Mac OS X, since Mac OS X is big endian PowerPC OS [mainly], while the win32 API is are mainly oriented to little endian x86. (Yes there was a Windows NT 3.5.1 port to CHRP PowerPC [running in little endian mode] a few years back, but it failed in general, because x86 binary programs could not run on the PowerPC.
Would it kill posix?:
OF Course NOT. Posix is a set of APIs for *nix-like systems, designed for scalblity, power, and stablity. Win32 APIs are designed to bring Windows a stable set of 32-bit APIs. Most *nix-like OSs rely heavly on posix APIs, so they will be in use for year and years and years.
At any rate, the main benfit of releasing win32 APIs, Windows would be more stable, faster (since everybody knew about the APIs). Also it would greatly help out projects like WINE.
Cows and farm animals can not be easily made resistant to the e.coli virus. Cows can't be changed (yet) by humans to get rid of the e.coli virus. No vaccine exists yet.
This is different with Windows. Microsoft could have and can engineer Windows easily to be resistant against the Back Orifice attacks out there. Linux / Unix is resitant to these kind of types of attacks. Did it require advance sicence and gene engineering? No. It just required some simple engineering ideas.
I would be mad if my security system in the office could be broken into via. some back door / secret way in (and the manifacture convently forgot to tell me).
If that back door was announced on TV (rendering my security system useless), I would feel more secure, since I now know about this exploit.
Since I now know that my security system is useless (and the world also knows that), I feel compleded to fix my security system up.
CDC is doing the right thing with Back Orifice 2000, releasing the source code. Now I can know for sure how they are breaking in, and I can fix it (using some little hack onto Windows, since I can not directly modify the source code). (Microsoft fixing it soon... haha)
I disagree on what you call stupid features. Yes, that Windows sliding menu stuff might be a waste of CPU/memory resources, it looks pretty (at least to some people). Of course, if Microsoft was smart like the KDE 2.0 team, for example they would have made the menu sliding feature optional, and completely disablable if you hate it, using 0 system resources. And yes the also give you 7 other styles to choose from if you kind of sorta like it. But Microsoft forces you to use this feature, creating bloat, and slowing you down.
Is anybody making you use a GUI on your system? No you don't have to, to enjoy the the power of Linux (you must use a GUI to use Windows). (since Windows == GUI + big patch to DOS).
The nice thing about Linux is it extermely competive, you don't like blah blah feature in your program, you can either contact the author and suggest he makes it optional, if he refuses, you can always code fork (under most licenses like the GPL or BSD). Why else would have things like E been invented if some commerical company had made it so we all had to use TWM, no questions asked.
Re:It's not the features that bloat the ware
on
All Hail Bloatware
·
· Score: 1
Easter Eggs should not add bloat if added to a program correctly. Most good easter eggs are pretty well hidden in the program, and don't waste memory and CPU when you are not using them.
I got to say my two favorite ones are from the MacOS:
System 7.5.x - The Ignuana Flag, by Sal Slagon the AppleScript Guy.
Mac OS 8.5/6 - The interactive presentation of authors of the Mac OS.
I don't think they add too much bloat, although they may be part of the reason why the Mac OS has gotten so big so quickly.
Yes, making a OS moduluar has it problems, such as you create depencies. But that's life. You have to use depency every day in life (you need money before you can buy that car), the other way just doesn't work. The same thing is true with Windows, a program might require you have Windows 98 installed and some werid graphics lib before you can use it. The Mac isn't much different either, you often need shared libaries for programs to run (Word 6 on the Mac required like 21 shared libaries to just run! Forently, Word 98 doesn't require any, because of a tool that automantains the shared libaries on your system.)
Forently, packaging systems makes this easy to deal with, for example installing the qt package will tell you kindly, you must have X11, and this and that installed.
But not everybody needs/wants qt on their system. So build it as part of your X Server? That would add bloat.
Commerical Developing for Linux is right now a major risk, in the fact that something far better to your commerical program becomes avalible for Linux for free (under the GPL). Think about the risk. You would quickly lose user base, even though your product has special features not in the GPL version of the program.
This has been a problem with the MacOS and speech software. Companies are afraid to develop speech software for the Mac, because they are afraid Apple will make Speech software part of the Mac OS for free.
E and Window Maker are both Window Mangers, and one has to run at a time.
If you are suggesting intergrating the code bases into one source code, that would not make sense, since Window Maker aims to be something totally different then Enlightenment.
If you are talking about GNOME compatiblity, both E and Window Maker work great with GNOME. Window Maker also works pretty good with KDE, although E could use some intergration work.
Slashdot Mirror
Microsoft is using AOL's properity OSC protocol, which is copyrighted and may not be reversed engineered by anybody including Microsoft.
The Free clients (GAIM, FAIM, etc.) are using TOC, which is fully documented, and you are encouraged to create clients that are compatible with it.
Microsoft chose to reverse engineer OSC because they felt that OSC users get better service and better features.
It's too bad they couldn't have gone the TOC protocol method... it would have saved them alot of bad publicity.
Personaly, I still think the correct name for it is Apple Macintosh System Software, as it used to called. True that was freaking hard to write out all of the time, sounds so 1983, and clone unfriendly, but still it's better then MacOS.
Finally, I prefer MacOS over Mac OS. And I like System Software over both of them.
"Slashdot -- The Place to Flame and be Flamed" --from linuxppc.org/links.html
Thanks,
Andrew B. Arthur aka AArthur
arthur99@global2000.net
AIM: Arthur998
I disagree... Most opensource authors (and users) try to insure the product is as secure as possible. Look at various security holes in Linux kernel for example -- they were fixed as soon as they were found. (That's were the expersion "Security in Internet time" comes in).
It would be hard to prove that Microsoft *doesn't* fix bugs/security holes quick enought. After all, they are commerical -- how long should it take them to recall a defective product?
Unless you can prove that either Microsoft or any Linux product spefically ignores a certain security holes, they can not be sued.
For example, they might be able to try to sue Microsoft over the Melissa virus, since Word Macros are a well known security flaw [they run with virtually no protection/sandboxing features (such as disabled file I/O), besides a macro warning box], which Microsoft should have done something about.
That's basically the same kind of like GM purposely (and knowing it) shipping defective seat belts, causing the death of 23 people in the past year. Of course GM would get sued for it, and the sueing party would likely win. Of course if GM didn't know about the problem -- it would be quite unlikely that any law suit against them would be successful.
It's not a price issue -- we already know that GPL'd software has to follow the law, eventhough it's GPL'd (which makes stealing parts of commerical code or reverse engineering illegal or infriging on trademarks). Recently the author of gaim, the gnome aol instant messager was required to remove "AOL" logos from his product. So he had to follow the law.
So basically the same laws apply to closed source and opensource projects.
One thing to note: OpenSource projects would be *MUCH* harder to sue, since you could easily claimed that the user:
1) didn't get the lastest updates / didn't read bugtraq
2) didn't fully inspect / test the code before installing (since he had everything the author had).
But still, the author would be responsble if the problem was blatly obvious (such as every Caldera employee knew that every 99th CD would distroy the user's Windows partition or the source code proved that this features was spefically coded in by a Caldera employee). This would be a definate lawsuit. However if this wasn't a known bug, the author would not be repsonble.
***So does Microsoft really know that Macros are a real security hole (or just a great easy to use feature that as a side effect makes viruses easy to write), and that they are ignoring to fix it?
***Well, we will leave it up to the lawyers to decide. Obviously, everybody has there own opinions.
Thanks,
Andrew B. Arthur aka AArthur
arthur99@global2000.net
AIM: aarthurppc
I really hope, this could lead to inexpensive PowerPC-based PCs, even cheaper then some of Apple's cheap machines-- (the iMac [at us$1299] and the iBook [at us$1599]).
:-( This is too bad, since the PowerPC platform is well designed and put together, even if it has some faults.
Other PowerPC hardware is much more expensive.
Yes, that is really true. Both Mandrake and Rasterman have really fast machines since they need them--they compile (at least parts) of Enlightenment and other pieces of software daily.
Very few people, who have old slow machines update CVS daily, much less recompile E every few minutes. You really don't want to try this on your typical PC, it will drag it down too much.
If your work requires it, you can afford those nice high tech tools.
(Although, they do recommend that you test all of your software reguluarly on the minium supported configuration, it's stupid to waste time on those machines compiling it on them).
Well, actually for several years Apple has owned cinepack (they bought it in the early 1990s), but it isn't as strictly controlled like Sorsen.
You see, we really don't need the full quicktime program, just a binary codec. We have Xamin which seems to support Cinepack movies to a limited degree of success.
Apple Cinepack is still under a binary license, and I don't see the problem with them licensing them to us.
Actually, they told us they would port the full quicktime to Linux *if* we payed them several thousand dollars. Eventually, if they think they can benfit from Quicktime on Linux x86, they will do it.
Lets not forget the opensourced Mac-On-Linux program, that lets several PowerPC-OS actually run in a seprate frame-buffered console.
Personally, I have seen it boot both LinuxPPC (watching LinuxPPC using BootX boot on top of LinuxPPC is cool), MkLinux (it's sometimes helpful for testing stuff), and Mac OS 8.6. It emulates a 604e PowerPC on most machines, and it runs at close to native speeds (maybe 3 or 4% slower, but that it). Video speed is decent, if you use the MOL video driver. Ethernet works with some cards.
Things it still needs--serial port support, support for more Macs without using rom-images, real SCSI support and other.
And yes it is GPL'd, and you can use the source or download it for free. And it is avalible now. SheepShaver will be out this summer.
It's at http://ibrium.se/linux/
Poorly Done Shareware:
.4) / 20) = $ 200 (maybe)
Shareware in my experience tends to be done either quickly or by people that have little talent at what they code. If these people had real talent, how come they haven't been hired to write software professionally?
I really don't want a 16 year old kid or somebody who has remedial coding skills to be writing closed source software, since we really don't know what or how he makes it work. Finally, much of the shareware software I have seen is quite buggy (such as ircle--it doesn't crash often on a good day).
Portablity:
Another, disadvantage to closed source software--if the author doesn't port it, you have no chance to get it on your archutechure. So if you are running something like Solaris or Linux/Alpha or Linux/PPC and you want to run it, you are out of luck.
Improving Coding Skills:
If you are going to showcase your coding skills to the world, it forces you to write good code, since everybody in the world will see it. It also makes you more competive to everybody else--you end up making better code.
Expensive:
Paying shareware fees can freaking add up quickly. $10 + $25 + more and more adds up to big bucks, trust me.
Linux Serial Numbers Guides:
If Linux gets flooded by bad shareware titles, how long to we get something like "Penguin Cracks" or "Linux Serial Numbers". Hell, I am sure their is them already, but I would think the guides are pretty small. And soon we will have warez sites. Lame.
Payback to Author:
It's a fact that less then 5% of downloads of shareware actually get payed for. People just run shareware unregistered or crack it (trust me, anybody with two months experience on a computer can crack a registration system on several shareware programs).
What does Shareware Make in Money:
Lets say a program super-shareware-draw pro is $10 for the program (a reasonable fee for the program). 1,000 people download if from either download.com or your homepage at kagi.com. Just for example, 1 out of 20 people register it (a reasonable number). Finally, about 4/10 of the people discard the program after downloading (that's also reasonable).
Do this on a caclualator:
10 * ((1000 *
If you get $200, you can consider your program to be a success. Of course, $200 to write a great application isn't really worth it in most cases -- you won't make $200 up front. It might take several months to get that full $200 in, your money will trickle in every one and a while at $10 a check.
Maybe shareware pays off for some, but going the gpl route would be in general a more rewarding route--if you get 1,000 downloads you know that like over 600 people are using your great piece of software to improve their lives. Also, if you write a good populuar piece of software, and people love it and enjoy it, that person who uses it could be future boss or something.
Shareware or GPL?
Depends on your situation. You want something you can make a little money off while you code, do it. However it most the time makes more sense to get a better paying job, such as working in a college libary or something.
Yes, I do rember when a 68040/40mhz could have preformance close to a 486/50-66mhz, but it would be pushing it to compare to any pentium.
The 68060s may have had pretty good preformance, but since I have never used or seen or seen benchmarks a 68060-based machine I will never know.
At any rate, having a lower clock rate, always looked bad compared to 486s. People would compare a 486/66 to 040/40, and think that the 486 would be like 1/3 faster or so. Of course that wasn't true, but it looked that way to stupid consumers and stupid sales staff.
This was one reason why Motorla pretty much droped the good old 68k out of the line of mainstream processors (not embedded), another was some of those 68040s and highers were kind of like the Pentium III -- power hogs and hot chips. And they were fairly big chips. Bad for both the embedded and PowerBook production.
And yes, the Pentium II and up does more work per cycle then the 68k chips, since they have been hacked up to preform decently.
* Think Different
hmm... those 68060s must be pretty warm and large chips similar to the Pentuim IIs. That was one complaint about the 68040 was they were just slow, hot and uneffectent chips...
The PowerPC did improve on most things, but early PowerPCs were so slow at 68k emulation, that some early PowerMacs actually had hardware working on 68k emulation (not quite a full processor, though)
Just a tadbit to tack on to that is that Linux has pretty good support (or will have) when altivec comes out. gcc will support altivec in the future, you can add support for it today, by getting the patch from motorola.
What does Altivec do? Well, as described by David McEvery, it is MMX implemented correctly. Basically its a set of additional instructions to the standard PowerPC RISC proccessor, adding support for better proccessing of graphical and related stuff. Sounds like MMX? Well it is, in a sense, basically what it has larger and faster processing of things, and implements cache correctly.
hmm... that kind of reminds me of the record changer, if any of you guys are old enought to remember what they do.
Basically, around the late 1970's record changers became unpopuluar (people were afraid they damaged the records and were sonically inferior to regular one disc players), yet I still own one of them and use it daily. The fact that you can not buy record changers anymore stop me from listening to one record after another on the changer, no.
The same thing with mp3s. All current hardware can play Mp3s, and it will be able to do so in the futuire (umm.. xmms-0.91 or later should continue working fine on my LinuxPPC R5 system (even 2 years from now), as long as I don't break it's depencies).
I can see that possibly hardware might have disablable Mp3 players, but on computers, their will always be hacks to get around other programs that might try to block the output from Mp3s.
So how do record from SDMI file to Mp3. Simple. You just copy the analog output of SDMI to your digital recorder. You feed that sound into Mp3 encoder... Yes their is a slight loss of sound quality, but if you use monster gold plated cable, it should be barely audiable if audiable at all.
What about analog watermarking?
From what I have read analog watermarking is total bullshit. Nobody is going to buy recording that have the sound quality audibly degraded just to keep the RIAA happy. People will rather fall back to Tapes and CDs to rip before using that DVD-Audio and SDIA.
What if DVD-Audio is a success with all it's anti-copy stuff?:
You got to be kidding. The first Audio-DVDs are two track with a higher sampling rate / frequency response. Maybe that will improve your dog's hearing of the music or that die hard audiofile that has a sound proofed room with dead slience, but for most people, no. Some DVD-Audio formats, actually degrade the sound, by adding analog (audiable) encyrption stuff.
Mulitchannel DVD-Audio discs. Hmm.. for one they are not comming out tommrow. Another problem with multichannel DVD's is they will not help out recordings recorded in stereo (skip the cheesey surrond effects, please!) Another problem, is most people's stereos are still only two channels, and going to 5.1 channels is expensive. Multichannel DVDs are looking, at least for now to be another RIAA pipe dream.
Actually Mac OS 8.5.x and better (and I think Windows 98), don't have to have jaggy icons. Mac OS 8.5.x supports alpha-chanel transperency and 32-bit icons, but they are a tadbit difficult to edit, since the tools suck to edit them (I use clip2icons and PhotoDeluxe [I cann't afford the real thing, and no I am not booting Linux/GIMP to edit a 32-bit icon!]).
They look nice if you know what you are doing with anti-aliased corners, and stuff.
The reason why NeXTstep does not have jaggys on the icons is that they use tiffs. Tiffs support 32-bit color natively and support alpha-transperency (I think).
Actually, somebody in like 1986 or early-1987 ported that demostation program to the Mac Plus.
It was a pretty cool demo. Some how it used a resolution hack (or at least made it look that way) so you got a high resolution (more then 72-dpi) using the standard Mac Plus video card. It's lines were smoother then smooth, but it seemed to take like a 1 1/2 minutes to initilize the screen at that resolution.
I have seen some pretty cool demos of Super3d on a MacPlus (I still have it), which looked cool, and you could make your own 3d images, but the resolution was far inferior (think 72 dpi) then that amiga clone demo.
If somebody could explain how that demo could create more then 72 dpi on a B&W mac plus, I would be a very happy person.
I know I shouldn't fall to this level, but I though that was pretty darn funny. hehe
That was certainly a great joke.
(/me returns to laughing on the floor)
Several reasons:
1) GNUstep is needs some work, in several areas. When it gets closer to 1.0.
2) GNUstep requires glibc 2.1, it won't work with libc5/glibc1.99
3) NeXTStep look is not liked by some people. But others love it.
4) The standard GUIs people use look like Mac or Windows, NeXTStep feels quite different.
Two things I can't stand with the screenshot:
1) They had to call the trash can garbage. That really is kind of confusing, since your files are typically trashed not garbaged (sounds to much like garbaled, and plus you don't use english that way). If you really must, at least call it something that makes sense like 'Trash' or 'Recycle Bin', etc. Afraid of lawsuit, huh?
2) What's up with this white mouse cursor. Most normal systems use a black curser (X11 and Macs), the only reason to use a white curser is to avoid a lawsuit. At any rate, studies show that having a black cursor == less eye strain, easier to find (most work is done on white paper/white screen).
It's not that easy just to open the win32 API. It might be to the benifit of Linux x86, but it wouldn't go much farther.
Even with Win32 APIs out there, it would be a major challenge to develop a OS from scratch. Since the Win32 APIs don't make any programs up, you would have to write a whole new desktop enviroment from sratch.
For one, it would be usless in Mac OS X, since Mac OS X is big endian PowerPC OS [mainly], while the win32 API is are mainly oriented to little endian x86. (Yes there was a Windows NT 3.5.1 port to CHRP PowerPC [running in little endian mode] a few years back, but it failed in general, because x86 binary programs could not run on the PowerPC.
Would it kill posix?:
OF Course NOT. Posix is a set of APIs for *nix-like systems, designed for scalblity, power, and stablity. Win32 APIs are designed to bring Windows a stable set of 32-bit APIs. Most *nix-like OSs rely heavly on posix APIs, so they will be in use for year and years and years.
At any rate, the main benfit of releasing win32 APIs, Windows would be more stable, faster (since everybody knew about the APIs). Also it would greatly help out projects like WINE.
Cows and farm animals can not be easily made resistant to the e.coli virus. Cows can't be changed (yet) by humans to get rid of the e.coli virus. No vaccine exists yet.
This is different with Windows. Microsoft could have and can engineer Windows easily to be resistant against the Back Orifice attacks out there. Linux / Unix is resitant to these kind of types of attacks. Did it require advance sicence and gene engineering? No. It just required some simple engineering ideas.
I would be mad if my security system in the office could be broken into via. some back door / secret way in (and the manifacture convently forgot to tell me).
If that back door was announced on TV (rendering my security system useless), I would feel more secure, since I now know about this exploit.
Since I now know that my security system is useless (and the world also knows that), I feel compleded to fix my security system up.
CDC is doing the right thing with Back Orifice 2000, releasing the source code. Now I can know for sure how they are breaking in, and I can fix it (using some little hack onto Windows, since I can not directly modify the source code). (Microsoft fixing it soon... haha)
I disagree on what you call stupid features. Yes, that Windows sliding menu stuff might be a waste of CPU/memory resources, it looks pretty (at least to some people). Of course, if Microsoft was smart like the KDE 2.0 team, for example they would have made the menu sliding feature optional, and completely disablable if you hate it, using 0 system resources. And yes the also give you 7 other styles to choose from if you kind of sorta like it. But Microsoft forces you to use this feature, creating bloat, and slowing you down.
Is anybody making you use a GUI on your system? No you don't have to, to enjoy the the power of Linux (you must use a GUI to use Windows). (since Windows == GUI + big patch to DOS).
The nice thing about Linux is it extermely competive, you don't like blah blah feature in your program, you can either contact the author and suggest he makes it optional, if he refuses, you can always code fork (under most licenses like the GPL or BSD). Why else would have things like E been invented if some commerical company had made it so we all had to use TWM, no questions asked.
Easter Eggs should not add bloat if added to a program correctly. Most good easter eggs are pretty well hidden in the program, and don't waste memory and CPU when you are not using them.
I got to say my two favorite ones are from the MacOS:
System 7.5.x - The Ignuana Flag, by Sal Slagon the AppleScript Guy.
Mac OS 8.5/6 - The interactive presentation of authors of the Mac OS.
I don't think they add too much bloat, although they may be part of the reason why the Mac OS has gotten so big so quickly.
Yes, making a OS moduluar has it problems, such as you create depencies. But that's life. You have to use depency every day in life (you need money before you can buy that car), the other way just doesn't work. The same thing is true with Windows, a program might require you have Windows 98 installed and some werid graphics lib before you can use it. The Mac isn't much different either, you often need shared libaries for programs to run (Word 6 on the Mac required like 21 shared libaries to just run! Forently, Word 98 doesn't require any, because of a tool that automantains the shared libaries on your system.)
Forently, packaging systems makes this easy to deal with, for example installing the qt package will tell you kindly, you must have X11, and this and that installed.
But not everybody needs/wants qt on their system. So build it as part of your X Server? That would add bloat.
So this is a good idea to make stuff modulized.
That is a major issue.
Commerical Developing for Linux is right now a major risk, in the fact that something far better to your commerical program becomes avalible for Linux for free (under the GPL). Think about the risk. You would quickly lose user base, even though your product has special features not in the GPL version of the program.
This has been a problem with the MacOS and speech software. Companies are afraid to develop speech software for the Mac, because they are afraid Apple will make Speech software part of the Mac OS for free.
I don't understand what you are mumbling about.
E and Window Maker are both Window Mangers, and one has to run at a time.
If you are suggesting intergrating the code bases into one source code, that would not make sense, since Window Maker aims to be something totally different then Enlightenment.
If you are talking about GNOME compatiblity, both E and Window Maker work great with GNOME. Window Maker also works pretty good with KDE, although E could use some intergration work.