Next up, Sun launches new ad campaign, "The Network is the Network" complete with diagrams of computer terminals and arrows and happy faces. Under their definition of "peer" which could mean anything from running shoes to supercomputers (this is from the article, people) their radical new plan connects those computers together in a mostly decentralized way to share information. This changes everything!
I mean even FreeBSD is easier to install then Debian.
Funny you should say that. I recently set up an ancient Pentium 133 as a stereo component and I figured since I was going to re-install as part of the process that I would use the opportunity to try out FreeBSD. Now this is a machine with a new hard-drive and has no shortage of problems with the BIOS and for whatever reason I can't get F10 diagnostics to work-- so feel free to ignore this...
But the FreeBSD install requires some deep magic apparently, because several tries later I was still unable to get the boot loader to work (maybe I should boot from floppy?)... Debian, however, was a cinch to install. LILO handles the inept BIOS without complaint, so rebooting actually succeeded, and adding packages with apt-get is a breeze.
Just to say I could, I did install FreeBSD on another system. I'm really into the ports system. However, it seems a little odd alongside a complete package system that downloads precompiled binaries. And don't even remind me that I spent a day trying to recompile my kernel for sound. Now that I finally figured it out, I suppose I should submit some documentation patches since the instructions were only mildly helpful.
And just to make sure we're not biased in favor of Debian: compared to Red Hat and Yellow Dog, Debian is hard to install, but easy to work with. The RH and YDL installs were practically automatic on the machines I've done-- including setting up the X server, which on Debian can be a tad tricky. So far I haven't found the perfect distro, but each of them seems to have a major selling point or two that make the variety worthwhile.
As long as we're thinking big... What I think would be really nice is a program that managed packages as source code in/usr/src and installed them individually into/usr/local/$package_name, linking them into/usr/bin and/usr/sbin as needed/desired. The *BSD ports system is most of the way there.
It would be even cooler if the package manager worked using CVS whenever possible so doing updates wouldn't necessarily require recompiling all those files that didn't change from one version to the next (after all, I hear that's part of the beauty of make and company).
It would also be cool if such a package manager existed (and yes, I'm going to write one if someone doesn't point to an existing example) and if a shared pool of patches and package tarballs/CVS info existed to facilitate downloading source code and getting Makefiles (and other files too I suppose) up to speed for a particular distro/architecture, so that installing from source was mostly painless. And that when packages were updated by their actual maintainers, that lists were available with the new version number, an MD5 hash, and the URL for the new tarball or CVS, and the reason for the update (so that users of this ports system could tell whether it was an essential security update or a feature(s) add).
As for the automatic dependency handling in apt, sure it's nice, but dependencies, schmependencies... when you're going from source, if your make process failed, you have unsatisfied dependencies. Time to check the make log and the README.:)
If you want an accurate idea of how much stuff sells for on eBay you need to look at completed auctions. As of the time I post this, your laptop linked is at $122.50. I suspect, given it's specs, it's likely to go to $300 and probably even $400 before the end of the auction. Even ancient laptops are extremely pricy-- a similar ThinkPad 380 just sold for $310 but only had a p/166 with a 2gb HD.
Personally, I'd look at the internet appliances if that's what I wanted (and I do and I am *grin*)-- especially the Audrey, with the touch screen and the flashed ROM stuff going for it. I'm not interested in waiting for my laptop to boot just to listen to mp3s on it's little speakers.
Well, and that's just the problem. I'd guess over half of new users want to run Linux precisely because it has some powerful tool available that convinced them to give it a try.
I knew I wanted a Unix-like system so I could run the IRC client that I was used to, learn Apache/CGI/Perl development, and read mail in Pine-- all things I'd gotten accustomed to doing on Unix shell accounts during college. (Plus Mac OS crashed on me constantly, what a pain!) Now running IRC and Pine were no big deal, neither is a default install of Apache and Perl. But before you know it you've got services all over the place and maybe some serious security issues. Besides, you know how it worked for you as a user, but you never had to think about the administration side.
And so yes, I agree that it's important for Linux newbies to get the basics down. After all, just learning about process management, disk space, and things like that means (if you ask me) learning the Unix way of doing things. Now maybe GNOME and KDE have made this less problematic, but how many Linux users just want to surf the web and do email? If they bought a new computer anywhere in the USA it came with an OS that is fairly reliable for those two things. So why would anyone switch at that point?
I guess, to me, the upshot is: yes, setting up a non-default Apache server or a Samba share probably is biting off more than you can chew if you don't know du from df. But we also need to recognize that there may be ways to work with Linux that don't require understanding the deep magic to get something going-- even if it is something we traditionally think of as complex. That was part of the point of the article "geez, look at all the pointy-clicky configuration stuff that's out there now."
Huh. I go there, Internet Explorer tells me that scripting is usually safe, but I say "no turn off scripting" just to be cautious/capricious. I get a white screen. Personally I think someone milked this KPMG crew for quite a chunk of change but doesn't know the first thing about making a usable website. *sigh*
Does anyone else see any irony behind the fact that the GPL doesn't require derived works to credit the original authors,
BS! You can't go willy-nilly removing copyright statements from GPL code-- and if I'm not mistaken that would constitute crediting the original authors. The only way you can possibly take someone's name off a package is to modify the package so completely that no portion of the remaining code is from that author. Even so, the ethical thing to do is leave their name in the relevant places. After all, the derived work in some way depends on the original work for its form and function and it doesn't hurt you to leave it attributed.
If the DFSG demanded that everything in a package be modifiable then it would not accept either the GPL, the LGPL, or the GNU FDL (and in fact, on the debian-legal list there has been a lot of traffic lately regarding the "freeness" of the GNU FDL). The GPL itself is clearly copyrighted and does not allow modifications. The GNU FDL allows for "invariant" sections, which is how things like the Manifesto get into the emacs documentation, but are technically verboten to remove-- and that's exactly *why* the FDL itself is being looked at as well.
Frankly, I don't see a reason why the LDP cannot go into non-free-- or be split into free and non-free components. LDP is readily available on the web, and almost anyone with set of ISOs is going to have the non-free stuff too, so it's available if needed. Users won't have to suffer unless they want to.
Microsoft have invested heavily in creating anti-copying technology that would make it safe for record labels and other content owners to sell their products online.
Bwahahaha! Is this the same protection they used for Windows XP which took, what, two days to crack?
I am not a lawyer either, but I'm pretty sure you can't go around copying stuff you don't own and then giving it away no matter how altruistic your motives are. The best defense in this case would be that there is no market, per se, for these trailers since they are free of charge to begin with.
At least they don't use the word "virii" when they mean "viruses". In Latin, you typically don't find the plural of words ending in "-us" turned into "-ii", but rather "-i"-- so if you still don't want to just speak English, the correct Latin word would be "viri". But then we have to be careful to indicate that we mean "viri" as in the plural form of the word for poisons and noxious stuff and not "viri" the plural form of the word for man.:)
34 companies isn't exactly a monopoly. Compare to a Microsoft-owned one company scenario.
What's dangerous, however, is that this 34 company oligopoly is the one that is likely to be the main influence in the SSS-CA and any regulation that results if that bill ever passes. They will have no qualms crushing your freedom to support their revenue models... "Liberty Alliance". Some joke.
emusic.com looks pretty f*cking cool. Who cares that they don't have Britney Spears or that the files are only 128kbps (that's good enough for casual listening)? Their minimum contract of three months at $15/month is the best bargain I've seen since I saw a whole operating system on the internet available as a free download.
$45 barely buys three CDs new, and maybe 6 CDs used. But looking throught the emusic.com catalog I could easily find 100 CDs that I would download in the first month. In fact, I expect I will have to write a Perl script that takes Artist/Album names and fetches the whole album to my mp3 server's file library pretty much as needed.
And frankly, I don't care if this helps the RIAA, because this hopefully will show them that offering unencumbered mp3s (or any other open format) encourages people to pay for the product when it is offered at a reasonable price. What none of these content providers have figured out yet is that they are only going to make $XYZ a year anyway-- and that they can either help all their consumers obtain the works legitimately or there will be a built-in incentive for someone to attempt to crack the system.
emusic.com could charge me $25-$50 a month or have some huge signup fee for unlimited access (and offer different service levels so that there were still some low-priced options) and it would still be worth it in my opinion. But if they followed the subscription model described in this article, there is no way I would sign up-- even if it were the *only* way to obtain music. I'd rather be stuck sifting through the used bins of all the stuff that's already out on unrestricted media and having to rip it myself.
A week? That's optimistic. I wouldn't be surprised if it's cracked during the beta cycle and a hacked client (or whatever is necessary to the crack) is release *same day* as the official launch of the service.
You are probably right that there will always be an app or two that run only on Windows. But right now there are plenty of cool apps that run only on Unix and Linux-- okay, maybe not plenty, but enough that run natively on Unix, whereas the Windows version is just not the same.
But what this does that's so awesome is demonstrate the power of Free Software-- and by so doing demonstrates to the average user what else they might gain by changing to a lifestyle where Free Software is emphasized over proprietary. Sure the average Windows user isn't going to be compiling from source in the next year. Definitely the average user of any system isn't going to be making subtle tweaks to their source code ever.
But by having a large library of cross-platform applications software, maybe users will migrate to Free Software (being exposed to it and having a vague understanding of why user freedom is a good thing). I know the availability of things like emacs and Perl on Windows makes my life better either way. Since I must use Windows at work, that these tools are available for the machines at work means I have the opportunity to do my job in a way that's comfortable for me and show off some Free Software can accomplish at the same time. Once there are no killer Windows apps left to replace, there will be no reason to have Windows at all-- and wouldn't that be nice?:)
That doesn't sound like it will handle my Ogg files.;)
Thanks for the suggestion. I'm fine running mpg321 or ogg123 on the CLI, but not sure how those stack up CPU-wise, so I will check out Xaudio. It would be nice to do two things at once on that machine every now and again.
No. The reason MS is the 800 pound gorilla is that any time they see someone else's innovation they work it into their operating system, buy the company outright, or code a workalike and release it for free. And I think what you see on SourceForge is a very public process, beginning with the idea phase, early development and planning, up to the few mature beta projects and then no insignificant number of release quality packages. I bet if you had similar full access to all the developer machines at Microsoft you'd find a similar number of ideas that never gained traction, a lot of early code that got chucked after the skeleton was written, enough parallel development of certain things to make you wonder who's in charge and if they are awake, and plenty of beta stuff that never makes it to release, and a handful of release quality stuff.
Note on the word release: *any* release is really just a beta until users get a chance to test it, only with Microsoft they can't call their first release beta because then no one would pay for it, the updates in the form of service packs are usually no charge-- free software can call stuff beta forever if they want, giving a package a release number is good for the public process of knowing when to feature freeze and do more rigorous testing for a while, but not much else.
Personally I think Sterling is flat out wrong. He does give some okay insights into why geeks might be a little wary of the government (after all the spooks are scary to just about anyone, and the spooks are actually interested in the geeks), but as to his assertion that there is no innovation going on... I think that's just plain nonsense.
Of course, when computing finally reached the level where you could put a really powerful machine on someone's lap there was a rush to fill in all the blanks, software-wise. But there's still plenty to be done, and people are working on it. Natural language parsing, non-mouse/non-keyboard interfaces, wearables, portables, AI, agent software, there may be lots of great clients out there (like Excel or Powerpoint), but just wait until there is an Excel server and two people can really easily start to work on the same spreadsheet simultaneously... I could go on and on with the things that are likely happening or have happened that we aren't really aware of yet...
His quick dismissal of Moore's law is based on small computers, look at the things big computers are doing that before were essentially impossible! Even if the underlying software or programming paradigms are the same, the real world application of the tools is drastically improving.
Finally, just because the last fifty years have seen this rate of growth in computing that's astronomical, we cannot and should not expect it to always feel like that: it's not even safe. When you get on the freeway you accelerate from zero to 60 rather quickly, but once you're on the road, just because you've stopped accelerating doesn't mean you're slowing down. Tech is perhaps done accelerating, but even at cruising speed we're still going places.
I've just recently had RH7.2 on a P133 laptop, but am in the process of switching it to FreeBSD4.4-- an OS with a little demon for its mascot is more my style. I previously ran RH6.2 on this machine and found that the RH7.2 release is significantly slower than the RH6.2 system running a compiled-from-source KDE2.2.x (I only upgraded because I was hoping to take advantage of X4-- but never ran X4 because the video chip on the laptop is unsupported as yet)...
Anyway, long story short: XMMS (or any mp3/ogg player) on a P/133 is just about going to max out a Linux system. Even minor events (like running a ps or ls from a terminal) can cause an interruption to the playback. On RH7.2 in KDE I'd get a lot of cuts at all times (simply from updating the XMMS window itself I think), whereas before as long as I didn't do anything but play the mp3 I'd be fine.
The question is not "How can i downgrade to a non-M$ system and lose functionality?"
I prefer not to think of Linux as less functional. We zealots like to refer to it as differently functional.;)
As for Office formats, unless someone is paying me to use the documents they're sending me, sending them in a file format my friends/family know I can't view is no different than me translating everything into Latin before I send it. How do you think they would react to that?
My comments on Flash are moronic? I'll give you moronic! The only people being punished by not being able to see Flash are kids! KIDS! I've never run into more sites that were utterly dependent on Flash than those sites geared towards children. I suppose I should run Windows-- for the children? Of course, the Flash has never actually added value to the web pages, it wrecks navigation, and it might be neat, but if I wanted to watch TV, I'd watch TV. HTTP stands for Hyper-Text Transfer Protocol, not "cheesy, useless animation protocol". Hypertext != moving graphics. Hypertext == text with links in it.:) *whoa, where this soapbox come from?*
Games... gaming is not relevant in a discussion about a Windows-less office. Linux has mine sweeper and solitaire and all those other good Windows games that come on a standard Windows desktop. Unless there is something inherent in Linux that prevents gaming software from being written for Linux, I think game manufacturers will continue to target dominant platforms just as they always have. Most games are available for more than one OS/platform/architecture already (Windows, Mac, Playstation, etc). I assume Linux is not going to get great games until it's more apparent that there is revenue to be generated by porting. This seems unlikely since the people mostly likely to insist on those games are the same people most likely to have a Windows partition handy and to not have a problem with that. Since it's practically impossible to buy a new computer without Windows pre-installed, the odds that someone who wants to use the computer for video games and them not having Windows is essentially zero.
But your last statement is the most telling, and is actually no less true now than it was about 20 years ago the first time I heard it said... "you don't decide on your OS or platform first, you decide what applications you want to run, then you make sure you have a machine and OS that can run them". Obviously the answer to this question for most offices right now includes Microsoft Office, and the best place to run MS Office is a version of MS Windows.
*mail: set up an auto-forward of all mail sent to your Exchange account. send all mail from a non-Exchange account. eventually wean people off the exchange address.
*office documents: demand that people send you the data as XML or HTML or RTF or TXT or any of the other zillions of formats available. if they will not: pout.
*web browsing: the only place Linux falls down is on terribly designed web pages and Flash. those sites are not worth viewing anyway. consider yourself rescued from bad web pages.:)
*games: you kids today and your fancy 3D rendering. in my day, games had 256 colors if they were lucky... some games were drawn in mono as vector graphics, other games relied solely on that faculty known as the imagination-- and presented the entire experience using only the written word. In fact, the more I play the new games, the more I like the older games. Once in a while I even use these analog games that rely on having people in the room with me. It's kind of fun!
As you can see, either you have an employment-related restriction and you can either try and find a new job (I know the desire to not use Windows at work makes me think about this once in a while, but on the whole it's unrealistic-- and they're paying for MS stuff, not me) or you have a comfort level that you seem to think you need. If you fall into this latter camp, all I can say is that comfort is relative, you have to weigh the comfort of interoperability/etc with the comfort of Free Software.
Just like it's hard to find people who are willing to man tech support call centers as an entry-level job in the IT industry? Besides, for some level of what's needed, admining a Linux office can't be any tougher than admining a Windows office.
So guess who would be the first to undergo a quick training-- yes, the old Windows admins. I think it would be a pretty poor business policy to just can the folks who've been doing the sysadmin job up to this point just because they have a slightly different experience. Linux isn't advanced magic at the desktop/office LAN level.
But frankly, I don't want sysadmins around anyway whose attitude is "you're a moron if you need your password reset"... sysadmins don't get paid to be wizards, they get paid to make sure the systems stay maintained and the users stay productive (from a technical perspective).
Re:Totally insulting price!!
on
This is IT?
·
· Score: 2
For $3000 I want to sit the f*ck down while I'm traveling. At 17mph it's going to be quite the hike to go anywhere even in a relatively dense city. Screw it. For less than $1000 I can get a scooter which will go a few extra mph, allow for passengers and runs around 70 miles on a tank of gas (which is less than a gallon in many cases). Sure, the scooter can tip over, but it's not even close to the motorcycle level in terms of difficulty. I think the primary danger in motorcycle riding is high speeds and unsafe driving practices (drinking, riding on wet roads, speed, etc)-- many of which are mitigated by the relatively underpowered nature of something like a Spree or a Vespa.
Hmmm. Did you continue reading to where he explained why Microsoft was so successful? It had nothing to do with financial mismanagement at mid-scale server companies.
Slashdot Mirror
Next up, Sun launches new ad campaign, "The Network is the Network" complete with diagrams of computer terminals and arrows and happy faces. Under their definition of "peer" which could mean anything from running shoes to supercomputers (this is from the article, people) their radical new plan connects those computers together in a mostly decentralized way to share information. This changes everything!
I mean even FreeBSD is easier to install then Debian.
Funny you should say that. I recently set up an ancient Pentium 133 as a stereo component and I figured since I was going to re-install as part of the process that I would use the opportunity to try out FreeBSD. Now this is a machine with a new hard-drive and has no shortage of problems with the BIOS and for whatever reason I can't get F10 diagnostics to work-- so feel free to ignore this...
But the FreeBSD install requires some deep magic apparently, because several tries later I was still unable to get the boot loader to work (maybe I should boot from floppy?)... Debian, however, was a cinch to install. LILO handles the inept BIOS without complaint, so rebooting actually succeeded, and adding packages with apt-get is a breeze.
Just to say I could, I did install FreeBSD on another system. I'm really into the ports system. However, it seems a little odd alongside a complete package system that downloads precompiled binaries. And don't even remind me that I spent a day trying to recompile my kernel for sound. Now that I finally figured it out, I suppose I should submit some documentation patches since the instructions were only mildly helpful.
And just to make sure we're not biased in favor of Debian: compared to Red Hat and Yellow Dog, Debian is hard to install, but easy to work with. The RH and YDL installs were practically automatic on the machines I've done-- including setting up the X server, which on Debian can be a tad tricky. So far I haven't found the perfect distro, but each of them seems to have a major selling point or two that make the variety worthwhile.
As long as we're thinking big... What I think would be really nice is a program that managed packages as source code in /usr/src and installed them individually into /usr/local/$package_name, linking them into /usr/bin and /usr/sbin as needed/desired. The *BSD ports system is most of the way there.
:)
It would be even cooler if the package manager worked using CVS whenever possible so doing updates wouldn't necessarily require recompiling all those files that didn't change from one version to the next (after all, I hear that's part of the beauty of make and company).
It would also be cool if such a package manager existed (and yes, I'm going to write one if someone doesn't point to an existing example) and if a shared pool of patches and package tarballs/CVS info existed to facilitate downloading source code and getting Makefiles (and other files too I suppose) up to speed for a particular distro/architecture, so that installing from source was mostly painless. And that when packages were updated by their actual maintainers, that lists were available with the new version number, an MD5 hash, and the URL for the new tarball or CVS, and the reason for the update (so that users of this ports system could tell whether it was an essential security update or a feature(s) add).
As for the automatic dependency handling in apt, sure it's nice, but dependencies, schmependencies... when you're going from source, if your make process failed, you have unsatisfied dependencies. Time to check the make log and the README.
If you want an accurate idea of how much stuff sells for on eBay you need to look at completed auctions. As of the time I post this, your laptop linked is at $122.50. I suspect, given it's specs, it's likely to go to $300 and probably even $400 before the end of the auction. Even ancient laptops are extremely pricy-- a similar ThinkPad 380 just sold for $310 but only had a p/166 with a 2gb HD.
Personally, I'd look at the internet appliances if that's what I wanted (and I do and I am *grin*)-- especially the Audrey, with the touch screen and the flashed ROM stuff going for it. I'm not interested in waiting for my laptop to boot just to listen to mp3s on it's little speakers.
Well, and that's just the problem. I'd guess over half of new users want to run Linux precisely because it has some powerful tool available that convinced them to give it a try.
I knew I wanted a Unix-like system so I could run the IRC client that I was used to, learn Apache/CGI/Perl development, and read mail in Pine-- all things I'd gotten accustomed to doing on Unix shell accounts during college. (Plus Mac OS crashed on me constantly, what a pain!) Now running IRC and Pine were no big deal, neither is a default install of Apache and Perl. But before you know it you've got services all over the place and maybe some serious security issues. Besides, you know how it worked for you as a user, but you never had to think about the administration side.
And so yes, I agree that it's important for Linux newbies to get the basics down. After all, just learning about process management, disk space, and things like that means (if you ask me) learning the Unix way of doing things. Now maybe GNOME and KDE have made this less problematic, but how many Linux users just want to surf the web and do email? If they bought a new computer anywhere in the USA it came with an OS that is fairly reliable for those two things. So why would anyone switch at that point?
I guess, to me, the upshot is: yes, setting up a non-default Apache server or a Samba share probably is biting off more than you can chew if you don't know du from df. But we also need to recognize that there may be ways to work with Linux that don't require understanding the deep magic to get something going-- even if it is something we traditionally think of as complex. That was part of the point of the article "geez, look at all the pointy-clicky configuration stuff that's out there now."
Huh. I go there, Internet Explorer tells me that scripting is usually safe, but I say "no turn off scripting" just to be cautious/capricious. I get a white screen. Personally I think someone milked this KPMG crew for quite a chunk of change but doesn't know the first thing about making a usable website. *sigh*
Does anyone else see any irony behind the fact that the GPL doesn't require derived works to credit the original authors,
BS! You can't go willy-nilly removing copyright statements from GPL code-- and if I'm not mistaken that would constitute crediting the original authors. The only way you can possibly take someone's name off a package is to modify the package so completely that no portion of the remaining code is from that author. Even so, the ethical thing to do is leave their name in the relevant places. After all, the derived work in some way depends on the original work for its form and function and it doesn't hurt you to leave it attributed.
If the DFSG demanded that everything in a package be modifiable then it would not accept either the GPL, the LGPL, or the GNU FDL (and in fact, on the debian-legal list there has been a lot of traffic lately regarding the "freeness" of the GNU FDL). The GPL itself is clearly copyrighted and does not allow modifications. The GNU FDL allows for "invariant" sections, which is how things like the Manifesto get into the emacs documentation, but are technically verboten to remove-- and that's exactly *why* the FDL itself is being looked at as well.
Frankly, I don't see a reason why the LDP cannot go into non-free-- or be split into free and non-free components. LDP is readily available on the web, and almost anyone with set of ISOs is going to have the non-free stuff too, so it's available if needed. Users won't have to suffer unless they want to.
Microsoft have invested heavily in creating anti-copying technology that would make it safe for record labels and other content owners to sell their products online.
Bwahahaha! Is this the same protection they used for Windows XP which took, what, two days to crack?
I am not a lawyer either, but I'm pretty sure you can't go around copying stuff you don't own and then giving it away no matter how altruistic your motives are. The best defense in this case would be that there is no market, per se, for these trailers since they are free of charge to begin with.
I stand corrected! Thanks. :)
If Clippy is dead, then explain why half the people on my floor here at work have a little dog or cat or whatever running around on their monitors.
At least they don't use the word "virii" when they mean "viruses". In Latin, you typically don't find the plural of words ending in "-us" turned into "-ii", but rather "-i"-- so if you still don't want to just speak English, the correct Latin word would be "viri". But then we have to be careful to indicate that we mean "viri" as in the plural form of the word for poisons and noxious stuff and not "viri" the plural form of the word for man. :)
34 companies isn't exactly a monopoly. Compare to a Microsoft-owned one company scenario.
What's dangerous, however, is that this 34 company oligopoly is the one that is likely to be the main influence in the SSS-CA and any regulation that results if that bill ever passes. They will have no qualms crushing your freedom to support their revenue models... "Liberty Alliance". Some joke.
Thanks for the link.
emusic.com looks pretty f*cking cool. Who cares that they don't have Britney Spears or that the files are only 128kbps (that's good enough for casual listening)? Their minimum contract of three months at $15/month is the best bargain I've seen since I saw a whole operating system on the internet available as a free download.
$45 barely buys three CDs new, and maybe 6 CDs used. But looking throught the emusic.com catalog I could easily find 100 CDs that I would download in the first month. In fact, I expect I will have to write a Perl script that takes Artist/Album names and fetches the whole album to my mp3 server's file library pretty much as needed.
And frankly, I don't care if this helps the RIAA, because this hopefully will show them that offering unencumbered mp3s (or any other open format) encourages people to pay for the product when it is offered at a reasonable price. What none of these content providers have figured out yet is that they are only going to make $XYZ a year anyway-- and that they can either help all their consumers obtain the works legitimately or there will be a built-in incentive for someone to attempt to crack the system.
emusic.com could charge me $25-$50 a month or have some huge signup fee for unlimited access (and offer different service levels so that there were still some low-priced options) and it would still be worth it in my opinion. But if they followed the subscription model described in this article, there is no way I would sign up-- even if it were the *only* way to obtain music. I'd rather be stuck sifting through the used bins of all the stuff that's already out on unrestricted media and having to rip it myself.
A week? That's optimistic. I wouldn't be surprised if it's cracked during the beta cycle and a hacked client (or whatever is necessary to the crack) is release *same day* as the official launch of the service.
You are probably right that there will always be an app or two that run only on Windows. But right now there are plenty of cool apps that run only on Unix and Linux-- okay, maybe not plenty, but enough that run natively on Unix, whereas the Windows version is just not the same.
:)
But what this does that's so awesome is demonstrate the power of Free Software-- and by so doing demonstrates to the average user what else they might gain by changing to a lifestyle where Free Software is emphasized over proprietary. Sure the average Windows user isn't going to be compiling from source in the next year. Definitely the average user of any system isn't going to be making subtle tweaks to their source code ever.
But by having a large library of cross-platform applications software, maybe users will migrate to Free Software (being exposed to it and having a vague understanding of why user freedom is a good thing). I know the availability of things like emacs and Perl on Windows makes my life better either way. Since I must use Windows at work, that these tools are available for the machines at work means I have the opportunity to do my job in a way that's comfortable for me and show off some Free Software can accomplish at the same time. Once there are no killer Windows apps left to replace, there will be no reason to have Windows at all-- and wouldn't that be nice?
That doesn't sound like it will handle my Ogg files. ;)
Thanks for the suggestion. I'm fine running mpg321 or ogg123 on the CLI, but not sure how those stack up CPU-wise, so I will check out Xaudio. It would be nice to do two things at once on that machine every now and again.
No. The reason MS is the 800 pound gorilla is that any time they see someone else's innovation they work it into their operating system, buy the company outright, or code a workalike and release it for free. And I think what you see on SourceForge is a very public process, beginning with the idea phase, early development and planning, up to the few mature beta projects and then no insignificant number of release quality packages. I bet if you had similar full access to all the developer machines at Microsoft you'd find a similar number of ideas that never gained traction, a lot of early code that got chucked after the skeleton was written, enough parallel development of certain things to make you wonder who's in charge and if they are awake, and plenty of beta stuff that never makes it to release, and a handful of release quality stuff.
Note on the word release: *any* release is really just a beta until users get a chance to test it, only with Microsoft they can't call their first release beta because then no one would pay for it, the updates in the form of service packs are usually no charge-- free software can call stuff beta forever if they want, giving a package a release number is good for the public process of knowing when to feature freeze and do more rigorous testing for a while, but not much else.
Personally I think Sterling is flat out wrong. He does give some okay insights into why geeks might be a little wary of the government (after all the spooks are scary to just about anyone, and the spooks are actually interested in the geeks), but as to his assertion that there is no innovation going on... I think that's just plain nonsense.
Of course, when computing finally reached the level where you could put a really powerful machine on someone's lap there was a rush to fill in all the blanks, software-wise. But there's still plenty to be done, and people are working on it. Natural language parsing, non-mouse/non-keyboard interfaces, wearables, portables, AI, agent software, there may be lots of great clients out there (like Excel or Powerpoint), but just wait until there is an Excel server and two people can really easily start to work on the same spreadsheet simultaneously... I could go on and on with the things that are likely happening or have happened that we aren't really aware of yet...
His quick dismissal of Moore's law is based on small computers, look at the things big computers are doing that before were essentially impossible! Even if the underlying software or programming paradigms are the same, the real world application of the tools is drastically improving.
Finally, just because the last fifty years have seen this rate of growth in computing that's astronomical, we cannot and should not expect it to always feel like that: it's not even safe. When you get on the freeway you accelerate from zero to 60 rather quickly, but once you're on the road, just because you've stopped accelerating doesn't mean you're slowing down. Tech is perhaps done accelerating, but even at cruising speed we're still going places.
I've just recently had RH7.2 on a P133 laptop, but am in the process of switching it to FreeBSD4.4-- an OS with a little demon for its mascot is more my style. I previously ran RH6.2 on this machine and found that the RH7.2 release is significantly slower than the RH6.2 system running a compiled-from-source KDE2.2.x (I only upgraded because I was hoping to take advantage of X4-- but never ran X4 because the video chip on the laptop is unsupported as yet)...
Anyway, long story short: XMMS (or any mp3/ogg player) on a P/133 is just about going to max out a Linux system. Even minor events (like running a ps or ls from a terminal) can cause an interruption to the playback. On RH7.2 in KDE I'd get a lot of cuts at all times (simply from updating the XMMS window itself I think), whereas before as long as I didn't do anything but play the mp3 I'd be fine.
The question is not "How can i downgrade to a non-M$ system and lose functionality?"
;)
:) *whoa, where this soapbox come from?*
I prefer not to think of Linux as less functional. We zealots like to refer to it as differently functional.
As for Office formats, unless someone is paying me to use the documents they're sending me, sending them in a file format my friends/family know I can't view is no different than me translating everything into Latin before I send it. How do you think they would react to that?
My comments on Flash are moronic? I'll give you moronic! The only people being punished by not being able to see Flash are kids! KIDS! I've never run into more sites that were utterly dependent on Flash than those sites geared towards children. I suppose I should run Windows-- for the children? Of course, the Flash has never actually added value to the web pages, it wrecks navigation, and it might be neat, but if I wanted to watch TV, I'd watch TV. HTTP stands for Hyper-Text Transfer Protocol, not "cheesy, useless animation protocol". Hypertext != moving graphics. Hypertext == text with links in it.
Games... gaming is not relevant in a discussion about a Windows-less office. Linux has mine sweeper and solitaire and all those other good Windows games that come on a standard Windows desktop. Unless there is something inherent in Linux that prevents gaming software from being written for Linux, I think game manufacturers will continue to target dominant platforms just as they always have. Most games are available for more than one OS/platform/architecture already (Windows, Mac, Playstation, etc). I assume Linux is not going to get great games until it's more apparent that there is revenue to be generated by porting. This seems unlikely since the people mostly likely to insist on those games are the same people most likely to have a Windows partition handy and to not have a problem with that. Since it's practically impossible to buy a new computer without Windows pre-installed, the odds that someone who wants to use the computer for video games and them not having Windows is essentially zero.
But your last statement is the most telling, and is actually no less true now than it was about 20 years ago the first time I heard it said... "you don't decide on your OS or platform first, you decide what applications you want to run, then you make sure you have a machine and OS that can run them". Obviously the answer to this question for most offices right now includes Microsoft Office, and the best place to run MS Office is a version of MS Windows.
*mail: set up an auto-forward of all mail sent to your Exchange account. send all mail from a non-Exchange account. eventually wean people off the exchange address.
:)
*office documents: demand that people send you the data as XML or HTML or RTF or TXT or any of the other zillions of formats available. if they will not: pout.
*web browsing: the only place Linux falls down is on terribly designed web pages and Flash. those sites are not worth viewing anyway. consider yourself rescued from bad web pages.
*games: you kids today and your fancy 3D rendering. in my day, games had 256 colors if they were lucky... some games were drawn in mono as vector graphics, other games relied solely on that faculty known as the imagination-- and presented the entire experience using only the written word. In fact, the more I play the new games, the more I like the older games. Once in a while I even use these analog games that rely on having people in the room with me. It's kind of fun!
As you can see, either you have an employment-related restriction and you can either try and find a new job (I know the desire to not use Windows at work makes me think about this once in a while, but on the whole it's unrealistic-- and they're paying for MS stuff, not me) or you have a comfort level that you seem to think you need. If you fall into this latter camp, all I can say is that comfort is relative, you have to weigh the comfort of interoperability/etc with the comfort of Free Software.
Just like it's hard to find people who are willing to man tech support call centers as an entry-level job in the IT industry? Besides, for some level of what's needed, admining a Linux office can't be any tougher than admining a Windows office.
So guess who would be the first to undergo a quick training-- yes, the old Windows admins. I think it would be a pretty poor business policy to just can the folks who've been doing the sysadmin job up to this point just because they have a slightly different experience. Linux isn't advanced magic at the desktop/office LAN level.
But frankly, I don't want sysadmins around anyway whose attitude is "you're a moron if you need your password reset"... sysadmins don't get paid to be wizards, they get paid to make sure the systems stay maintained and the users stay productive (from a technical perspective).
For $3000 I want to sit the f*ck down while I'm traveling. At 17mph it's going to be quite the hike to go anywhere even in a relatively dense city. Screw it. For less than $1000 I can get a scooter which will go a few extra mph, allow for passengers and runs around 70 miles on a tank of gas (which is less than a gallon in many cases). Sure, the scooter can tip over, but it's not even close to the motorcycle level in terms of difficulty. I think the primary danger in motorcycle riding is high speeds and unsafe driving practices (drinking, riding on wet roads, speed, etc)-- many of which are mitigated by the relatively underpowered nature of something like a Spree or a Vespa.
Hmmm. Did you continue reading to where he explained why Microsoft was so successful? It had nothing to do with financial mismanagement at mid-scale server companies.