I ended up going with nVidia, since their 3D acceleration seems to work better under Linux than ATi's. My CPU is powerful enough to decode MPEGs on it's own. And as for an accelerated framebuffer: I don't have any graphical apps that run on the framebuffer, and anyways the whole point of X is to display multiple terminal emulators at the same time.:)
Having said that, I am very interested in which piece of currently available graphics hardware you suggest, that does all the things you mentioned in Linux. Is it price competitive vs ATi/nVidia?
I could say something about versions of MacOS before X, and their complete lack of a little concept I like to call *protected memory* or that cool new *preemptive multitasking.* In cases where the OS doesn't take steps to protect itself, the system is only as stable as the most buggy application that one runs on it. This is true for DOS, MacOS and PalmOS.
Anyways, regardless of the "sad, unstable, pathetic joke" of an OS that came on it many pieces of hardware can be made to run Linux, including i386s and up, PowerPCs, and ARMs (including XScales).
FreeBSD doesn't seem to be doing too badly, and it's competing right in Linux's own backyard: x86. The problem isn't some kind of Linux conspiracy, it's that people want a usable system without having to write half the userland themselves while hacking on the kernel. This also explains the obsession with Unix-workalikes: They already have lots of usable programs. If you're going to start from scratch you have to accept that there will be a huge time investment before anyone else cares. Maybe Linux seems like it got popular overnight (only took 14 years right?), but keep in mind that RMS had been hacking on the GNU tools since 10 years before Linux's inception. So, yes I'm all for new ideas and I like trying out new OSes, but don't try and blame the amorphous cloud that is Linux for somehow keeping people from trying new OSes.
The same way that geeks got Linux booting on the Mac G3s designed for MacOS 8. Also, you may be interested to know that Palm switched from those horrible 16MHz Motorollas quite a while ago and now most (all?) PalmOS handhelds just have ARM (or ARM compatible) CPUs like everyone else.
Someone has already contribted a beta build of gpe-mini-browser which is based on GTK-webcore, which IIRC is related to KHTML in some way. mini-browser has some issues right now, such as lack off HTTPS support, but renders well and is impressively fast on low power hardware. It runs great on my Zaurus C1000 (with an Xscale @ 416MHz). It would not suprise me to see a port of konqueror show up at some point. Firefox is also likely to be an easy port since it will compile and run on a Zaurus (both have ARM-archictecture CPUs), though IMHO Firefox is too heavy for a system like the 770.
Hmmm, maybe there are some factors not being considered. Perhaps users won't complain unless pass a certain "annoyance threshhold." Maybe 3 bugs won't annoy a user enough to complain but 4 will. Different users will of course have different annoyance threshholds. In addition to this perhaps one user (A) complaining to another user (B) about a bug may cause B to also complain about the bug even if normally it wouldn't have bothered B enough to complain in the first place. Just some thoughts...
"This is why at the end of the day Linux will have an equivilant for EVERYTHING that exists on OSX or XP or whatever but it's always going to lack the refined standardized feel that most people like in a finished product."
Having used GNU/Linux since 1998 I've seen it make vast strides in this area. In 1998 it wasn't true that there was an equivalent to everything offered on Windows (and MacOS). When I started using it, the choice of web browsers was Netscape 4.05, lynx (or links or w3m) or Mosaic. I guess you could call Netscape an "equivalent" to it's Windows and Mac counterpart, but it was barely usable! Open more than 3 windows and it crashes. Try to submit something with a text entry box and it submits random garbage. There wasn't really anything you could call an equivalent file manager. WYSIWYG word processors were mostly painful to use, buggy and underfeatured. Autodetection of hardware was a joke. There wasn't even a very usable AIM client. Linux was simply not a usable desktop OS at the time.
Today we take a lot of things for granted. Firefox for Linux gives almost exactly the same user experience as Firefox on Windows. Despite (very minor) shortcomings Open Office (or even Abiword) is a very usable word processor. Autodetection of hardware is getting to a point where most of the time you can plug something in and it just works. In the area of instant messaging clients Gaim is now at the point where many Windows users are using it instead of AIM and MSN. Some say that Open Source always seems to be catching up commercial software but never surpassing it. From my point of view the gap between Open Source and commercial software is closing, not rapidly, but steadily.
It's not a problem of interested developers. The Zaurus scene has plenty of those (check out oesf.org/forums). My C1000 which has been out for about a year and isn't even available in the US except through importers, has several different Linux distributions and OpenBSD available. People have ported X servers, got a full debian distro to run on it and generally hacked it up in some very interesting ways. But it will never have SDIO because of all the red tape around SD in general.
The only way to write SD (or SDIO) drivers is by pay some exorbitant fee to the SD consortium gods or by reverse engineering it. The fine people behind SD have let it be known in no uncertain terms that if you release open source SD drivers you will get slapped with the lawsuit of a lifetime in return. So SD drivers have to come from the manufacturer of the device or some well funded commercial operation (SD specs ain't cheap!). And if the manufacturer doesn't have SDIO at launch, what incentive do they have to add it after launch? No potential return on investment means no investment in the first place.
For a better description of the situation check here.
Yeah, I heard SDIO drivers will be released simultaneously with a legal $20 DVD player for Linux and Duke Nukem Forever...
Seriously though, SD drivers, let alone SDIO, just aren't realistic to expect if they don't *ship with the device.* The Sharp Zaurus SL-5500 has been out for over 3 years now and you still have to run the 2.4.18 kernel it shipped with because no updated or open source SD drivers are available. IIRC, all the Sharp Zaurus SL-C750 and up have hardware support for SDIO but can't use it because there are no drivers. There is a reason that a lot of smaller companies still ship products that only accept MMC cards.
for apple.com/trailers: Use the firefox "Media Player Connectivity" plugin for Firefox. It takes an inline, plugin requiring piece of media like a quicktime, WMV, avi or mpeg, and turns replaces the space on that page with a nice button to click on that launches the media player you choose to handle that format in a separate window. Or you can set it up to just let you right click and download the movie. Click here.
As for this little piece of stupidity called.viv, I don't think there is a current easy solution...
Very good analogy. It can be argued both ways though: In much the same way that we can predict that next year's pop bands will suck just as much as this years, we can safely assume that Windows Vista will suck just as much as the current version of Windows, albeit in new and interesting ways.
yeah, but storage media with moving parts just doesn't cut it for mobile applications. They suffer higher seek times, higher power consumption and tend to be more fragile. And need I mention spin-up time? Ask the Palm Life Drive owners how much they enjoy waiting for their hard drives to spin up every time they want to look up a phone number or write themselves a short memo.
To top it off, flash media can survive a lot more writes than people seem to think. By distributing writes among evenly among all the sectors on the device the life of the device can be greatly increased.
purely anecdotal: I accidentally put my usb flash drive through the wash. I let it dry a couple days then plugged it in and it came up fine. No filesystem corruption at all. I'd like to see someone do that to a microdrive and tell me the results.:)
This is strictly anecdotal in nature...but on more than one occasion I've run into people that have upgraded or replaced Windows machines that had begun to seem slow simply from the incredible amount of spyware, adware, malware and random crap they had running in their system tray. This wasn't always especially outdated or slow hardware either. And it isn't always spyware...Raise your hand if you've seen the stupifying number of things running in the system tray on a brand new Sony, HP or . It's mind boggling. But I digress I wouldn't be very surprised to find out that, the reason many Windows users upgrade or replace their computer is simply because of how "slow" it's gotten from all the random crap that accumulates over time in addition to all the stuff that is running in the background straight from the factory. Is part of the reason Mac users don't upgrade as often simply due to less accumulation of cruft?
On a slightly different note: I don't think that the reason there is no (or very little) Mac spyware/adware/malware is really only due to installed base. This might be reaching a little far but different operating systems (especially "minority" operating systems) seem to have different feels to their user and developer community. Many people who develop software for Windows seem to be out to make a fast buck, and therefore adware and spyware abounds. It is simply a logical extension of shareware. Even the freeware developed for Windows, usually isn't free for commercial use. Most Windows users wouldn't know what do with source code so it is rarely requested and even more rarely provided. Much of the free (as in speech) and open source software on Windows seems to originate on Linux. The BeOS community was kind of a mixed bag. It was made up of Linux, Mac and Windows users and as such was a mixed bag. A kind of comradarie did seem to prevail though. Users and developers seemed compelled to support their fellow users. There was freeware, some with source available, and the shareware usually either nagged you once in a while or was so reasonably priced compared to it's feature set that it would be considered worth the price. The Linux user community on the other hand seems downright hostile towards anyone who won't release their program under a free license. As such developers either deal with the fact that they can't really charge for software or they put donation links on their web pages. Either way, it seems like very few bother trying to sell closed source software to "Linux zealots.":)
I can't comment directly on the Mac community because I've never owned a Mac (the Mac Mini might changed that...), but it seems like Mac users in general don't mind the concept of commercial software and are willing to pay a reasonable price for good software. This seems to encourage entrepreneurial programmers to actually develop feature rich programs, knowing that people would be willing to pay for them. This is in contrast to many Windows users who want things that are free (as in beer) but don't care about freedom (as in speech). Now that MacOS is based on *BSD, lots of free (speech) and open source software is being ported as well.
So the point of this long rambling rant: 1) I think Windows users upgrade (hardware) more often because of spyware/adware/cruft 2) I think Windows users are a target of spyware/adware not only because Windows is the largest installed base, but because of the mentality that they don't want to pay for things, yet they also don't care about truely free (as in speech) software 3) I should create some little program (GPL of course:)) to prevent myself from trying to make coherent posts to Slashdot at 4 in the morning...
-Mr. Lizard
(I hit preview and realized how much I had actually typed. I don't feel like proofreading it, so hopefully it makes some kind of sense. Must sleep...)
The transfer rates are very impressive for something with such a low totational speed though. I imagine that lots of geeks will be buying these things and modding them with traditional hard drive motors to unleash their real potential. Imagine getting one these spinning at even 5,400 RPMs. Think of the possibilities!
To know who was it that said, to know recursion you must first know recursion, you must first know who was it that said, to know recursion you must first know recursion.
Maybe they think that the coffee shops trying to blanket the city in wifi will use this as a backbone to provide Internet access for their wifi hotspots. Less expensive and less of a hassle than getting a T1 line run. Also it has provisions for increasing bandwidth without needing to upgrade equipment.
To continue the analogy: Sure it would be nice if the engine acted to reinforce the frame, but if the company can't even get the engine to work right in the first place then what good is it anyways? Also, is it really better for the consumer to be locked into a certain engine because putting in a different one would break most of the car? Wouldn't it be neat if it was designed in such a way that you put could have a 4 cylinder, 6 cylinder, or 8 cylinder engine without having to worry that the rest of the car is going to fall apart around your ears because you changed something? That's called modularity.
...And the winner is IBM! Seriously, with all the nextgens using PPCs they're in a great position since no matter who wins in the console wars they're selling processors. Conspiracy theories may now commence.
I ended up going with nVidia, since their 3D acceleration seems to work better under Linux than ATi's. My CPU is powerful enough to decode MPEGs on it's own. And as for an accelerated framebuffer: I don't have any graphical apps that run on the framebuffer, and anyways the whole point of X is to display multiple terminal emulators at the same time. :)
Having said that, I am very interested in which piece of currently available graphics hardware you suggest, that does all the things you mentioned in Linux. Is it price competitive vs ATi/nVidia?
-Mr. Lizard
I could say something about versions of MacOS before X, and their complete lack of a little concept I like to call *protected memory* or that cool new *preemptive multitasking.* In cases where the OS doesn't take steps to protect itself, the system is only as stable as the most buggy application that one runs on it. This is true for DOS, MacOS and PalmOS.
Anyways, regardless of the "sad, unstable, pathetic joke" of an OS that came on it many pieces of hardware can be made to run Linux, including i386s and up, PowerPCs, and ARMs (including XScales).
FreeBSD doesn't seem to be doing too badly, and it's competing right in Linux's own backyard: x86. The problem isn't some kind of Linux conspiracy, it's that people want a usable system without having to write half the userland themselves while hacking on the kernel. This also explains the obsession with Unix-workalikes: They already have lots of usable programs. If you're going to start from scratch you have to accept that there will be a huge time investment before anyone else cares. Maybe Linux seems like it got popular overnight (only took 14 years right?), but keep in mind that RMS had been hacking on the GNU tools since 10 years before Linux's inception. So, yes I'm all for new ideas and I like trying out new OSes, but don't try and blame the amorphous cloud that is Linux for somehow keeping people from trying new OSes.
-Mr. Lizard
The same way that geeks got Linux booting on the Mac G3s designed for MacOS 8. Also, you may be interested to know that Palm switched from those horrible 16MHz Motorollas quite a while ago and now most (all?) PalmOS handhelds just have ARM (or ARM compatible) CPUs like everyone else.
Someone has already contribted a beta build of gpe-mini-browser which is based on GTK-webcore, which IIRC is related to KHTML in some way. mini-browser has some issues right now, such as lack off HTTPS support, but renders well and is impressively fast on low power hardware. It runs great on my Zaurus C1000 (with an Xscale @ 416MHz). It would not suprise me to see a port of konqueror show up at some point. Firefox is also likely to be an easy port since it will compile and run on a Zaurus (both have ARM-archictecture CPUs), though IMHO Firefox is too heavy for a system like the 770.
Hmmm, maybe there are some factors not being considered. Perhaps users won't complain unless pass a certain "annoyance threshhold." Maybe 3 bugs won't annoy a user enough to complain but 4 will. Different users will of course have different annoyance threshholds. In addition to this perhaps one user (A) complaining to another user (B) about a bug may cause B to also complain about the bug even if normally it wouldn't have bothered B enough to complain in the first place. Just some thoughts...
"This is why at the end of the day Linux will have an equivilant for EVERYTHING that exists on OSX or XP or whatever but it's always going to lack the refined standardized feel that most people like in a finished product."
Having used GNU/Linux since 1998 I've seen it make vast strides in this area. In 1998 it wasn't true that there was an equivalent to everything offered on Windows (and MacOS). When I started using it, the choice of web browsers was Netscape 4.05, lynx (or links or w3m) or Mosaic. I guess you could call Netscape an "equivalent" to it's Windows and Mac counterpart, but it was barely usable! Open more than 3 windows and it crashes. Try to submit something with a text entry box and it submits random garbage. There wasn't really anything you could call an equivalent file manager. WYSIWYG word processors were mostly painful to use, buggy and underfeatured. Autodetection of hardware was a joke. There wasn't even a very usable AIM client. Linux was simply not a usable desktop OS at the time.
Today we take a lot of things for granted. Firefox for Linux gives almost exactly the same user experience as Firefox on Windows. Despite (very minor) shortcomings Open Office (or even Abiword) is a very usable word processor. Autodetection of hardware is getting to a point where most of the time you can plug something in and it just works. In the area of instant messaging clients Gaim is now at the point where many Windows users are using it instead of AIM and MSN. Some say that Open Source always seems to be catching up commercial software but never surpassing it. From my point of view the gap between Open Source and commercial software is closing, not rapidly, but steadily.
It's not a problem of interested developers. The Zaurus scene has plenty of those (check out oesf.org/forums). My C1000 which has been out for about a year and isn't even available in the US except through importers, has several different Linux distributions and OpenBSD available. People have ported X servers, got a full debian distro to run on it and generally hacked it up in some very interesting ways. But it will never have SDIO because of all the red tape around SD in general.
The only way to write SD (or SDIO) drivers is by pay some exorbitant fee to the SD consortium gods or by reverse engineering it. The fine people behind SD have let it be known in no uncertain terms that if you release open source SD drivers you will get slapped with the lawsuit of a lifetime in return. So SD drivers have to come from the manufacturer of the device or some well funded commercial operation (SD specs ain't cheap!). And if the manufacturer doesn't have SDIO at launch, what incentive do they have to add it after launch? No potential return on investment means no investment in the first place.
For a better description of the situation check here.
Yeah, I heard SDIO drivers will be released simultaneously with a legal $20 DVD player for Linux and Duke Nukem Forever...
Seriously though, SD drivers, let alone SDIO, just aren't realistic to expect if they don't *ship with the device.* The Sharp Zaurus SL-5500 has been out for over 3 years now and you still have to run the 2.4.18 kernel it shipped with because no updated or open source SD drivers are available. IIRC, all the Sharp Zaurus SL-C750 and up have hardware support for SDIO but can't use it because there are no drivers. There is a reason that a lot of smaller companies still ship products that only accept MMC cards.
-Mr. Lizard
By the time these are out good ol' global warming will have kicked into full swing and you won't need to worry about that whole "cold" thing.
for apple.com/trailers: Use the firefox "Media Player Connectivity" plugin for Firefox. It takes an inline, plugin requiring piece of media like a quicktime, WMV, avi or mpeg, and turns replaces the space on that page with a nice button to click on that launches the media player you choose to handle that format in a separate window. Or you can set it up to just let you right click and download the movie. Click here.
.viv, I don't think there is a current easy solution...
As for this little piece of stupidity called
Like, for example, a site or two getting slashdotted?
...Leaves 2 hard drives dead and one CD-ROM drive wounded.
I like how I'm the only one in this thread who got modded down, regardless of the fact that 4 other people said almost the exact same thing...
Very good analogy. It can be argued both ways though: In much the same way that we can predict that next year's pop bands will suck just as much as this years, we can safely assume that Windows Vista will suck just as much as the current version of Windows, albeit in new and interesting ways.
-Mr. Lizard
yeah, but storage media with moving parts just doesn't cut it for mobile applications. They suffer higher seek times, higher power consumption and tend to be more fragile. And need I mention spin-up time? Ask the Palm Life Drive owners how much they enjoy waiting for their hard drives to spin up every time they want to look up a phone number or write themselves a short memo.
:)
To top it off, flash media can survive a lot more writes than people seem to think. By distributing writes among evenly among all the sectors on the device the life of the device can be greatly increased.
purely anecdotal: I accidentally put my usb flash drive through the wash. I let it dry a couple days then plugged it in and it came up fine. No filesystem corruption at all. I'd like to see someone do that to a microdrive and tell me the results.
This is strictly anecdotal in nature...but on more than one occasion I've run into people that have upgraded or replaced Windows machines that had begun to seem slow simply from the incredible amount of spyware, adware, malware and random crap they had running in their system tray. This wasn't always especially outdated or slow hardware either. And it isn't always spyware...Raise your hand if you've seen the stupifying number of things running in the system tray on a brand new Sony, HP or . It's mind boggling. But I digress I wouldn't be very surprised to find out that, the reason many Windows users upgrade or replace their computer is simply because of how "slow" it's gotten from all the random crap that accumulates over time in addition to all the stuff that is running in the background straight from the factory. Is part of the reason Mac users don't upgrade as often simply due to less accumulation of cruft?
:)
:)) to prevent myself from trying to make coherent posts to Slashdot at 4 in the morning...
On a slightly different note: I don't think that the reason there is no (or very little) Mac spyware/adware/malware is really only due to installed base. This might be reaching a little far but different operating systems (especially "minority" operating systems) seem to have different feels to their user and developer community. Many people who develop software for Windows seem to be out to make a fast buck, and therefore adware and spyware abounds. It is simply a logical extension of shareware. Even the freeware developed for Windows, usually isn't free for commercial use. Most Windows users wouldn't know what do with source code so it is rarely requested and even more rarely provided. Much of the free (as in speech) and open source software on Windows seems to originate on Linux. The BeOS community was kind of a mixed bag. It was made up of Linux, Mac and Windows users and as such was a mixed bag. A kind of comradarie did seem to prevail though. Users and developers seemed compelled to support their fellow users. There was freeware, some with source available, and the shareware usually either nagged you once in a while or was so reasonably priced compared to it's feature set that it would be considered worth the price. The Linux user community on the other hand seems downright hostile towards anyone who won't release their program under a free license. As such developers either deal with the fact that they can't really charge for software or they put donation links on their web pages. Either way, it seems like very few bother trying to sell closed source software to "Linux zealots."
I can't comment directly on the Mac community because I've never owned a Mac (the Mac Mini might changed that...), but it seems like Mac users in general don't mind the concept of commercial software and are willing to pay a reasonable price for good software. This seems to encourage entrepreneurial programmers to actually develop feature rich programs, knowing that people would be willing to pay for them. This is in contrast to many Windows users who want things that are free (as in beer) but don't care about freedom (as in speech). Now that MacOS is based on *BSD, lots of free (speech) and open source software is being ported as well.
So the point of this long rambling rant:
1) I think Windows users upgrade (hardware) more often because of spyware/adware/cruft
2) I think Windows users are a target of spyware/adware not only because Windows is the largest installed base, but because of the mentality that they don't want to pay for things, yet they also don't care about truely free (as in speech) software
3) I should create some little program (GPL of course
-Mr. Lizard
(I hit preview and realized how much I had actually typed. I don't feel like proofreading it, so hopefully it makes some kind of sense. Must sleep...)
The transfer rates are very impressive for something with such a low totational speed though. I imagine that lots of geeks will be buying these things and modding them with traditional hard drive motors to unleash their real potential. Imagine getting one these spinning at even 5,400 RPMs. Think of the possibilities!
To know who was it that said, to know recursion you must first know recursion, you must first know who was it that said, to know recursion you must first know recursion.
In terms of weather the Seattle roll-out should make a decent simulation of how it would perform in Britain.
But, would you be impressed enough to cover the $500 bill he racks up the first time he connects to the service?
Maybe they think that the coffee shops trying to blanket the city in wifi will use this as a backbone to provide Internet access for their wifi hotspots. Less expensive and less of a hassle than getting a T1 line run. Also it has provisions for increasing bandwidth without needing to upgrade equipment.
To continue the analogy: Sure it would be nice if the engine acted to reinforce the frame, but if the company can't even get the engine to work right in the first place then what good is it anyways? Also, is it really better for the consumer to be locked into a certain engine because putting in a different one would break most of the car? Wouldn't it be neat if it was designed in such a way that you put could have a 4 cylinder, 6 cylinder, or 8 cylinder engine without having to worry that the rest of the car is going to fall apart around your ears because you changed something? That's called modularity.
...And the winner is IBM! Seriously, with all the nextgens using PPCs they're in a great position since no matter who wins in the console wars they're selling processors. Conspiracy theories may now commence.
In fact it's aimed at such a mature audience that we'll all have to wait 5 - 10 years from now until we're old enough for it...