CDs are stereo only, and generally cost more than DVDs, which are 5.1 surround *and* have a video track.
I would like to see classical concerts released in DVD format. These I would buy.
Are CDs obsolete? Not yet. The question should be "why are sales dropping?" Here in Zambonia, we pay a levy on unrecorded CDs. I wonder what the *combined* sales figures (recorded and unrecorded) are. Add to that the flash levy, and then compare. After all, flash devices didn't exist when CDs were released.
I support the media levy, which, by the way, the Kanukistan version of the RIAA lobbied FOR (for years), and now is lobbying AGAINST. What a crock of shite.
"Of course, Stallman had to hijack our definition of "free" to correspond to his beliefs, ignoring the fact that "freeware" existed before him."
Let's examine this. The term "freeware" was coined by Andrew Fluegelman in 1982. Richard Stallman started GNU in 1983. GNU Manifesto is dated 1985. So, yes, you are correct "Freeware" existed before RMS.
But... "Freeware" was a trademarked term.
In 1979, RMS was already mulling over "freedom in software issues". A quote from that time: "the prospect of charging money for software was a crime against humanity."
So the answer to the question: Did RMS hijack the term "freeware"? No, "freeware" was a trademarked term. Yes, Fluegelman used the term "free" in relation to software one year before GNU, and 3 years before the GNU Manifesto. No, RMS was thinking about free and software freedom in 1979, 3 years before Fluegelman coined "freeware".
I build MythTV systems out for people. Here is a rough breakdown:
HTPC case, IR remote integrated, display (VFD) integrated, $300. Note that the VFD/IR is over $100 by itself -- and the case has to be "living room acceptable".
Power supply - main characteristic - QUIET. $40.
Mainboard. Note that we are using this for (potentially) running multiple HD streams, and transcoding HD, so Athlon X2, and 2GB (dual channel) memory. 4 PCI slots, 1 PCIe. $100 for the mainboard/processor, $60 for memory.
Now, what does it sum up to? $910. Which is close to a thousand (oh, the quiet passive CPU cooling adds $30).
What do you get? A bog-quiet box, with all sorts of A/V possibilities, that looks very nice. Run MythTV on it, which doesn't take ALL of its power (most of the time -- when transcoding you can't have enough!), so put VMware on there, and some virtual servers as well (I run XP and Solaris 10 on mine).
I run a small business at home. Programming services.
I use Rogers as my ISP. My other choice is Bell Sympatico.
But Bell isn't sympathetic at all. Bell blocks port 25 outbound AND INBOUND. No mail for me, unless I use a.sympatico.ca domain. Maybe they do other port blocking -- I don't know. Upgrading to a "business account" doesn't solve the issue either. I would have to purchase an account with a mail redirector.
Rogers only blocks port 25 outbound. Started a couple of years ago, with no warning. I told Rogers that I run "light" servers, and they seem to have no trouble with that... Once in a while, I contact Rogers, TRYING to upgrade my service to "business class" -- with a static IP, and the right to run servers. I don't mind my data cap. But, no luck, they just won't sell me the service!
I consider you lucky in that you CAN get business class service; Rogers could cut me off at any time with no recourse (except buying a racked system somewhere).
The way Microsoft implemented "asm" was wrong. Either the compiler contains a machine simulator in order to optimize the "asm" along with the C code, or any sequence containing "asm" needed most optimizations disabled.
The GCC approach (declaring used registers -- runtime characteristics of the asm code) allows the "asm" to be included in optimized sequences.
Now, you may not care about standards, but, come on, C99 is 8 years old! I certainly expected a compiler vendor with as many resources as Microsoft to adopt it. ESPECIALLY when gcc already supports (most) of it. Basically, I can't use C99 because *if* the code needs to be compiled with CL, it doesn't work. So, I am stuck with C89 (the 18 year old standard).
I write code; expressions of algorithms cast into a programming language. I *don't* write to a particular platform, or operating system. I find it disturbing that the largest consumer driver of computers can simply disregard standards which would make it easy for me to employ their platform.
I am sure that there will be a lot of "snicker" replies -- how can the magistrate be so stupid...
But this is an interesting idea. RAM holds information, specifically the IP addresses in this case.
"Sorry, we don't have the IP address available; they are never recorded". To which the reply is: "They ARE recorded. In RAM. So copy RAM".
Why this is a useful result: It means that it *could* become illegal to build a computer that has "unreadable" memory, because *that* memory may be where information needed by a court is being kept, and it needs copying.
Which means that "secure writeable storage" for DRM becomes illegal (at least on computers).
But, back to the topic, the magistrate is dead on. Of course, the RAM could simply be dumped onto a hard disks, lather, rinse, repeat. I don't think INTERPRETATION of the document was discussed!
Technically, email is just a protocol(s) between computers.
Your computer finds my computer (possibly via other computers) in order to deliver a message on behalf of you for me.
Now, I don't charge money for your use of the protocol. Indeed, I rely on network effects, and encourage use of the protocol by giving out my "email address".
Why should one of the intermediate computers charge you for using my protocol? Sounds silly to me. I *want* your email.
Spam, of course, I don't want. But it happens because of (1) people without sruples, and (2) that aformentioned network effect.
I take steps to combat spam -- specifically, all outgoing email is relayed through an upstream ISP, and I don't allow relaying through my system.
Which should be sufficient (from my perspective). Sufficient because all ISPs could require the same limits, and if those limits where enforced there could be no high-volume spam. Specifically, any mass mailing HAS to go THROUGH an ISP (as mine have to).
The only way it breaks down if there are ISPs without scruples (and there are).
But these can produce spam anyway (even with a system like "goodmail" in place). Which makes "goodmail" rather useless.
I don't want to impose ANY MORE LIMITS on my domain, and I certainly don't want to charge for email -- My choice is to provide the service exactly as I do now; and my users must accept "no relaying" and "must use upstream ISP relay outbound" restrictions. Just no further.
SUN is a publically traded company. They, by law "firmly commit" to the shareholders. After that, the rest is:
- open source Solaris. Been in the works since Solaris 8. Includes ZFS, dtrace, and other good stuff.
- buy OpenOffice.org, invest in it, and open source it.
- nfs, nis, etc.
Exactely what DO you want SUN to do? SUN even opened the processor arch for the SPARC. It was even the IEEE STANDARD processor -- and what happened? The COMMUNITY let Intel "win" with a proprietary arch. The COMMUNITY let Microsoft "win" with a proprietary network file sharing.
NIS (NIS+) support is crucial for automount. We are talking about heterogenous networks.
And, if needed for automount, it *should* be supported for login.
Now, other options may be supported as well for login, but NIS should be there.
Why does NIS suck? It is simply a service that provides key/value lists from replicated servers. And, in Linux, the options for nsswitch naming services are: NIS, NIS+, or HESIOD.
If Microsoft releases the information, it would require incorporation into MANY operating systems (Linux, Solaris, HP/UX, AIX, and more).
These operating systems *interoperate* with a number of standards.
What would be MUCH MORE useful would be if Microsoft actually implemented those standards. As a start:
- use NIS (NIS+) for signon, hosts, services, etc.
- support NFS
- support SUN automounts
- support LPD (CUPS)
This would allow Microsoft based workstations to "just work" when put into networks, along side of those "other" operating systems. The parenthetical standard would be "nice to have", but not required.
Start with the basic infrastructure. Honestly, there should be no need for SAMBA!
After that, Microsoft should support a POSIX standard shell and POSIX utilities (to allow scripting compatible with the other operating systems). This would allow consistent administration.
And then Microsoft should support C89 and C99. After all C99 is an 8 year old standard! Microsoft STILL doesn't support "long long", and other C99 constructs.
Microsoft should also support POSIX APIs, and X. If an application uses Windows APIs natively, there should be a mapping layer to translate the application to X.
But this is only a fanatasy. I woke up from that dream years ago.
1 - If the communication or thread switching overhead exceeds the thread computation, it is not worth threading.
2 - It is (unfortunately) easy to build in "lock stepping" into otherwise independent threads. These systems scale from 1..n cores; after n cores no further scaling is seen.
3 - It *is* difficult to build correct parallel systems. Especially with points 1 and 2 in mind (and, yes, I *have* built parallel high-speed device drivers that are lock-free to avoid switching).
4 - *Proving* that a multi-thread program is correct is quite difficult, especially when using lock-free constructs.
Actually, the Microsoft patent pretty much describes groff/troff being able to run external commands. They do use the word "automatic" in the patent... but they seem to mean "automatic when a link is clicked".
But that isn't stated. Which means the combination of a markup language with external program selection has been done. The only NEW thing is the addition of a lookup table. I would argue that the shell implements the lookup table aspect (and, in certain systems, is implemented that way -- eg. VMS DCL). So, I wouldn't worry about that one.
The Eolas patent seems to want to leverage off of browser extensions. They claim that any programmatic use of a browser extension or applet is covered. But this one is dangerous. The IDEA of wrapping an API is in widespread use (eg. Linux NDIS network wrapper). This would let Eolas attack (for instance) development environments that wrap APIs around components, especially if the dev env is scriptable. This is a superset of the first one, but could be applied in a much more focussed (and difficult to defend) fashion. (After all, we normally don't think of a shell "!" extension in vi, for example, as a wrapper for shell in the editor -- even though it is, in a way).
But my suspicion is that neither of these will be overturned, and some will claim that Linux now infringes on "n+1" patents. I don't think that either company will act on the patent (Eolas may, and if that happens, some may just pay out).
Your assertion that P2P is (just) a hack to avoid central authority and move bandwidth is partly right.
Yers, I use "P2P" (bittorrent) as a hack. But the DIRECT problem I use bittorrent to address is the disparity between my download speed (5ish mbits/second) vs. my upload speed (256ish kbits/second).
I prefer to be in control of my own network resources -- and not rely on "central authority". So, yes, that is the "end reason" for using bittorrent.
But there would be no reason to use a "P2P" solution if my upload and download rates where the same, or if my upload rate where larger. As it is presently, it takes 20 uploads to match a download (for me), making the case for using P2P.
Indeed, it would be better if Dell charged *more* for Ubuntu than Windows.
Ubuntu is a better choice for novice users. Better security (through isolation of components). All basic software needs covered (browsing, word processing, photo editing, email), with (arguably) best-of-breed software (firefox vs. ie).
Stable, good looking and performant system. Easy to install and remote maintainance is possible. Better standards compliance. Less resource intensive, allowing a longer life on your hardware purchase.
This should be worth something. If Dell charged a $50 PREMIUM for the Ubuntu OS, it may convince some users that it is worth pirating! (and, by extension, using).
I'm sorry, I don't get your point. Microsoft is the "arrogant" one here.
For instance, does XP support NFS out of the box? How about NIS? How do I get XP to reflect home directories and common NIS passwd signon?
Microsoft is arrogant-- their solution is "better". Ignoring interoperability with any other common OS (SUN OS, Solaris, AIX, HP/UX).
What about POSIX support "out of the box" (a standards compliant shell environment would be nice)? X Window support?
As a programmer, how about a C99 compliant compiler?
Yes, I use Linux and Solaris and I even know WHY. I also develop for Windows. I don't mind it, because it's a massive market. But, really, it isn't "Linux" that's arrogant, it's Windows.
Let's think about this: No, I probably CAN'T make a channel over the internet that you can't detect. But I CAN make it very difficult for you to know that the channel exists. (and, yes, I have been involved with projects that needed to protect against thermal measurement to help recover key bits).
Example - the timing of requests to DNS servers (especially DNS servers under the vendors control). After all, morse code was an effective communication method.
If all I wanted to do was to pass on your keystrokes (a very low bandwidth thing), I could certainly do it in a wya that wouldn't be picked up by most systems. Modifying a few bits in normal packet headers (not so good), modifying the html requests (a bit better). timing (a bit better), retreiving "updates" or some static resources, coupled with one of the other methods from my vendor site (looks innocuous, after all what's an extra space or ten?).
You really can't know unless you (or someone you trust) has reviewed the code. Even there, the code vendor COULD be in cahoots with the compiler vendor.
Not that I believe that Opera suffers from this...
Around 5 years ago I was working in the lab for a (major) hardware vendor. I had just received a request to investigate full resolution digital video for theaters. The idea was that movies were to be distributed via hard disk instead of on film. The hard disks could be returned and new feature movies put on them, whereas the film is simply destroyed.
I went to my systems administrator, and said "I need a lot of disk space...". He asked "How much?". "A lot... um...". "Well, 'lot' could mean a gig, 10 big, or even a terrabyte.". My reply: "A terrabyte would be a good start... mumble". "WHAT!". "Yeash, how about a terrabyte now, and 10 terrabytes next week?".
The sysadmin fainted.
And this was only 5 years ago. Unfortunately, I/O speed has not kept up -- imagine trying to statically mirror two of these drives for a backup. (The video application used a lot of smaller drives, and naturally had a lot more heads, so it could maintain I/O rates. Even then the data had to be laid out carefully to maximize output speed).
Still, I am looking forward to my first 3TB RAID5 home storage unit.
Mice, graphics, networking JUST WORKS. You may not benefit from the "whiz-bang 3D super gaming feature". On the other hand, if this IS a concern to you, it puts you in very rarified company. Indeed, most people just use light word processing, maybe a digital camera, webcam, the web, and email. They want to play music and DVDs.
AND THAT'S IT. Are you wondering why Microsoft and others are planning "software as web service"? I will let you in on a really dark secret... Windows 98 would be "good enough" for most people, coupled with Openoffice.org (or Office 97) and Firefox, if it is behind a NAT router.
The Linux kernel has more drivers than Microsoft Vista kernel.
Linux distributions come with more software than Microsoft Vista distribution.
Openoffice.org is a very capable replacement for Microsoft Office. Yes, some things are easier in Microsoft Office. And I would expect that -- after all, Microsoft Office costs a whole lot more.
Yes, Firefox "does something". So does the Linux kernel. Specifically, the Linux kernel manages hardware resources. Mark me as failing your intelligence test.
I generate a PostScript file (possibly PDF) if I want "exact printing". I can test the PostScript against Adobe and Ghostscript (especially with gs' "safety" mode).
I generate an ASCII file, with illustrations in a neutral format (X bitmaps, XFig, JPEG, or PostScript).
I give a marked up document in (La)Tex, or TROFF.
(La)Tex is generally presumed "defect free", even though it does constitute a "monoculture" (Knuths bug pay-out policy contributes).
TROFF has two common implementations - AT&T (Solaris), and groff. Both considered very stable.
Ghostscript is very stable; I cannot speak to the Adobe implementation. There are other PostScript implementations available as well (Harlequin?)
These days, I still use (La)Tex for structured work, ODF for "ad-hoc" work (letters and throw-away memos, where the visual result is more important than sharing).
I view document preparation a bit like programming, in that the language should be considered as separate from the compiler implementation, and there should be tools to allow the language to be extended (elements by function or purpose, not format). Most wysiwyg systems fail at this (including Openoffice.org) which limits their usefulness to me.
Thanks for your comment. Yes, the opening statement is a "flame-ish" punch to get attention. My signature does mention "insightful troll", as a reflection of writing style.
And, I do say that PowerShell can be the best thing since sliced bread where it's appropriate. You are correct in that when all methods are available, PowerShell can be used as a fantastic integration tool.
It still fails in "ad-hoc" usage, except where "ad-hac" means common desktop command line use.
For my purposes (reading others log files, mostly), the Unix paradigm works better. And PowerShell also works here, but only as an implementation of the Unix paradigm. The object nature doesn't help (yet -- maybe log file implementors will introduce an object classification that everyone can agree with).
Slashdot Mirror
CDs are stereo only, and generally cost more than DVDs, which are 5.1 surround *and* have a video track.
I would like to see classical concerts released in DVD format. These I would buy.
Are CDs obsolete? Not yet. The question should be "why are sales dropping?" Here in Zambonia, we pay a levy on unrecorded CDs. I wonder what the *combined* sales figures (recorded and unrecorded) are. Add to that the flash levy, and then compare. After all, flash devices didn't exist when CDs were released.
I support the media levy, which, by the way, the Kanukistan version of the RIAA lobbied FOR (for years), and now is lobbying AGAINST. What a crock of shite.
"Of course, Stallman had to hijack our definition of "free" to correspond to his beliefs, ignoring the fact that "freeware" existed before him."
Let's examine this. The term "freeware" was coined by Andrew Fluegelman in 1982. Richard Stallman started GNU in 1983. GNU Manifesto is dated 1985. So, yes, you are correct "Freeware" existed before RMS.
But... "Freeware" was a trademarked term.
In 1979, RMS was already mulling over "freedom in software issues". A quote from that time: "the prospect of charging money for software was a crime against humanity."
So the answer to the question: Did RMS hijack the term "freeware"? No, "freeware" was a trademarked term. Yes, Fluegelman used the term "free" in relation to software one year before GNU, and 3 years before the GNU Manifesto. No, RMS was thinking about free and software freedom in 1979, 3 years before Fluegelman coined "freeware".
So, twist it any damn way you want.
I build MythTV systems out for people. Here is a rough breakdown:
HTPC case, IR remote integrated, display (VFD) integrated, $300. Note that the VFD/IR is over $100 by itself -- and the case has to be "living room acceptable".
Power supply - main characteristic - QUIET. $40.
Mainboard. Note that we are using this for (potentially) running multiple HD streams, and transcoding HD, so Athlon X2, and 2GB (dual channel) memory. 4 PCI slots, 1 PCIe. $100 for the mainboard/processor, $60 for memory.
SD (standard def) TV. Dual tuner - $140 (with hardware MPEG-2 encode).
QAM/ATSC HD tuner - $80.
Suitable video card (full HD output) - $150
Montego sound card (optical digital in/out) - $40
Now, what does it sum up to? $910. Which is close to a thousand (oh, the quiet passive CPU cooling adds $30).
What do you get? A bog-quiet box, with all sorts of A/V possibilities, that looks very nice. Run MythTV on it, which doesn't take ALL of its power (most of the time -- when transcoding you can't have enough!), so put VMware on there, and some virtual servers as well (I run XP and Solaris 10 on mine).
You are lucky!
.sympatico.ca domain. Maybe they do other port blocking -- I don't know. Upgrading to a "business account" doesn't solve the issue either. I would have to purchase an account with a mail redirector.
I run a small business at home. Programming services.
I use Rogers as my ISP. My other choice is Bell Sympatico.
But Bell isn't sympathetic at all. Bell blocks port 25 outbound AND INBOUND. No mail for me, unless I use a
Rogers only blocks port 25 outbound. Started a couple of years ago, with no warning. I told Rogers that I run "light" servers, and they seem to have no trouble with that... Once in a while, I contact Rogers, TRYING to upgrade my service to "business class" -- with a static IP, and the right to run servers. I don't mind my data cap. But, no luck, they just won't sell me the service!
I consider you lucky in that you CAN get business class service; Rogers could cut me off at any time with no recourse (except buying a racked system somewhere).
Off topic
The way Microsoft implemented "asm" was wrong. Either the compiler contains a machine simulator in order to optimize the "asm" along with the C code, or any sequence containing "asm" needed most optimizations disabled.
The GCC approach (declaring used registers -- runtime characteristics of the asm code) allows the "asm" to be included in optimized sequences.
Now, you may not care about standards, but, come on, C99 is 8 years old! I certainly expected a compiler vendor with as many resources as Microsoft to adopt it. ESPECIALLY when gcc already supports (most) of it. Basically, I can't use C99 because *if* the code needs to be compiled with CL, it doesn't work. So, I am stuck with C89 (the 18 year old standard).
I write code; expressions of algorithms cast into a programming language. I *don't* write to a particular platform, or operating system. I find it disturbing that the largest consumer driver of computers can simply disregard standards which would make it easy for me to employ their platform.
I am sure that there will be a lot of "snicker" replies -- how can the magistrate be so stupid...
But this is an interesting idea. RAM holds information, specifically the IP addresses in this case.
"Sorry, we don't have the IP address available; they are never recorded". To which the reply is: "They ARE recorded. In RAM. So copy RAM".
Why this is a useful result: It means that it *could* become illegal to build a computer that has "unreadable" memory, because *that* memory may be where information needed by a court is being kept, and it needs copying.
Which means that "secure writeable storage" for DRM becomes illegal (at least on computers).
But, back to the topic, the magistrate is dead on. Of course, the RAM could simply be dumped onto a hard disks, lather, rinse, repeat. I don't think INTERPRETATION of the document was discussed!
Technically, email is just a protocol(s) between computers.
Your computer finds my computer (possibly via other computers) in order to deliver a message on behalf of you for me.
Now, I don't charge money for your use of the protocol. Indeed, I rely on network effects, and encourage use of the protocol by giving out my "email address".
Why should one of the intermediate computers charge you for using my protocol? Sounds silly to me. I *want* your email.
Spam, of course, I don't want. But it happens because of (1) people without sruples, and (2) that aformentioned network effect.
I take steps to combat spam -- specifically, all outgoing email is relayed through an upstream ISP, and I don't allow relaying through my system.
Which should be sufficient (from my perspective). Sufficient because all ISPs could require the same limits, and if those limits where enforced there could be no high-volume spam. Specifically, any mass mailing HAS to go THROUGH an ISP (as mine have to).
The only way it breaks down if there are ISPs without scruples (and there are).
But these can produce spam anyway (even with a system like "goodmail" in place). Which makes "goodmail" rather useless.
I don't want to impose ANY MORE LIMITS on my domain, and I certainly don't want to charge for email -- My choice is to provide the service exactly as I do now; and my users must accept "no relaying" and "must use upstream ISP relay outbound" restrictions. Just no further.
"don't firmly commit".
SUN is a publically traded company. They, by law "firmly commit" to the shareholders. After that, the rest is:
- open source Solaris. Been in the works since Solaris 8. Includes ZFS, dtrace, and other good stuff.
- buy OpenOffice.org, invest in it, and open source it.
- nfs, nis, etc.
Exactely what DO you want SUN to do? SUN even opened the processor arch for the SPARC. It was even the IEEE STANDARD processor -- and what happened? The COMMUNITY let Intel "win" with a proprietary arch. The COMMUNITY let Microsoft "win" with a proprietary network file sharing.
NIS (NIS+) support is crucial for automount. We are talking about heterogenous networks.
And, if needed for automount, it *should* be supported for login.
Now, other options may be supported as well for login, but NIS should be there.
Why does NIS suck? It is simply a service that provides key/value lists from replicated servers. And, in Linux, the options for nsswitch naming services are: NIS, NIS+, or HESIOD.
It's not that simple.
If Microsoft releases the information, it would require incorporation into MANY operating systems (Linux, Solaris, HP/UX, AIX, and more).
These operating systems *interoperate* with a number of standards.
What would be MUCH MORE useful would be if Microsoft actually implemented those standards. As a start:
- use NIS (NIS+) for signon, hosts, services, etc.
- support NFS
- support SUN automounts
- support LPD (CUPS)
This would allow Microsoft based workstations to "just work" when put into networks, along side of those "other" operating systems. The parenthetical standard would be "nice to have", but not required.
Start with the basic infrastructure. Honestly, there should be no need for SAMBA!
After that, Microsoft should support a POSIX standard shell and POSIX utilities (to allow scripting compatible with the other operating systems). This would allow consistent administration.
And then Microsoft should support C89 and C99. After all C99 is an 8 year old standard! Microsoft STILL doesn't support "long long", and other C99 constructs.
Microsoft should also support POSIX APIs, and X. If an application uses Windows APIs natively, there should be a mapping layer to translate the application to X.
But this is only a fanatasy. I woke up from that dream years ago.
How can you tell an extroverted Linux user?
He looks at YOUR shoes while talking to you.
A couple of points:
1 - If the communication or thread switching overhead exceeds the thread computation, it is not worth threading.
2 - It is (unfortunately) easy to build in "lock stepping" into otherwise independent threads. These systems scale from 1..n cores; after n cores no further scaling is seen.
3 - It *is* difficult to build correct parallel systems. Especially with points 1 and 2 in mind (and, yes, I *have* built parallel high-speed device drivers that are lock-free to avoid switching).
4 - *Proving* that a multi-thread program is correct is quite difficult, especially when using lock-free constructs.
Actually, the Microsoft patent pretty much describes groff/troff being able to run external commands. They do use the word "automatic" in the patent... but they seem to mean "automatic when a link is clicked".
But that isn't stated. Which means the combination of a markup language with external program selection has been done. The only NEW thing is the addition of a lookup table. I would argue that the shell implements the lookup table aspect (and, in certain systems, is implemented that way -- eg. VMS DCL). So, I wouldn't worry about that one.
The Eolas patent seems to want to leverage off of browser extensions. They claim that any programmatic use of a browser extension or applet is covered. But this one is dangerous. The IDEA of wrapping an API is in widespread use (eg. Linux NDIS network wrapper). This would let Eolas attack (for instance) development environments that wrap APIs around components, especially if the dev env is scriptable. This is a superset of the first one, but could be applied in a much more focussed (and difficult to defend) fashion. (After all, we normally don't think of a shell "!" extension in vi, for example, as a wrapper for shell in the editor -- even though it is, in a way).
But my suspicion is that neither of these will be overturned, and some will claim that Linux now infringes on "n+1" patents. I don't think that either company will act on the patent (Eolas may, and if that happens, some may just pay out).
Go figure.
Your assertion that P2P is (just) a hack to avoid central authority and move bandwidth is partly right.
Yers, I use "P2P" (bittorrent) as a hack. But the DIRECT problem I use bittorrent to address is the disparity between my download speed (5ish mbits/second) vs. my upload speed (256ish kbits/second).
I prefer to be in control of my own network resources -- and not rely on "central authority". So, yes, that is the "end reason" for using bittorrent.
But there would be no reason to use a "P2P" solution if my upload and download rates where the same, or if my upload rate where larger. As it is presently, it takes 20 uploads to match a download (for me), making the case for using P2P.
It's not the price difference.
Indeed, it would be better if Dell charged *more* for Ubuntu than Windows.
Ubuntu is a better choice for novice users. Better security (through isolation of components). All basic software needs covered (browsing, word processing, photo editing, email), with (arguably) best-of-breed software (firefox vs. ie).
Stable, good looking and performant system. Easy to install and remote maintainance is possible. Better standards compliance. Less resource intensive, allowing a longer life on your hardware purchase.
This should be worth something. If Dell charged a $50 PREMIUM for the Ubuntu OS, it may convince some users that it is worth pirating! (and, by extension, using).
I'm sorry, I don't get your point. Microsoft is the "arrogant" one here.
For instance, does XP support NFS out of the box? How about NIS? How do I get XP to reflect home directories and common NIS passwd signon?
Microsoft is arrogant-- their solution is "better". Ignoring interoperability with any other common OS (SUN OS, Solaris, AIX, HP/UX).
What about POSIX support "out of the box" (a standards compliant shell environment would be nice)? X Window support?
As a programmer, how about a C99 compliant compiler?
Yes, I use Linux and Solaris and I even know WHY. I also develop for Windows. I don't mind it, because it's a massive market. But, really, it isn't "Linux" that's arrogant, it's Windows.
aQuantive has ~900 million in assets, and a current shareholder equity of ~600 million. Net revenue is the first quarter of this year was ~14 million.
Microsoft is aquiring them for 6 billion? Sounds about as dumb-ass as Google with DoubleClick (what, 31x premium?).
aQuantive stock is now 64.75 *up 27.88*. I missed it.
As a Microsoft shareholder, I don't like this. What a waste of money! I mean, paying 3 billion would be high, but 6 billion?
Let's think about this: No, I probably CAN'T make a channel over the internet that you can't detect. But I CAN make it very difficult for you to know that the channel exists. (and, yes, I have been involved with projects that needed to protect against thermal measurement to help recover key bits).
Example - the timing of requests to DNS servers (especially DNS servers under the vendors control). After all, morse code was an effective communication method.
If all I wanted to do was to pass on your keystrokes (a very low bandwidth thing), I could certainly do it in a wya that wouldn't be picked up by most systems. Modifying a few bits in normal packet headers (not so good), modifying the html requests (a bit better). timing (a bit better), retreiving "updates" or some static resources, coupled with one of the other methods from my vendor site (looks innocuous, after all what's an extra space or ten?).
You really can't know unless you (or someone you trust) has reviewed the code. Even there, the code vendor COULD be in cahoots with the compiler vendor.
Not that I believe that Opera suffers from this...
Around 5 years ago I was working in the lab for a (major) hardware vendor. I had just received a request to investigate full resolution digital video for theaters. The idea was that movies were to be distributed via hard disk instead of on film. The hard disks could be returned and new feature movies put on them, whereas the film is simply destroyed.
I went to my systems administrator, and said "I need a lot of disk space...". He asked "How much?". "A lot... um...". "Well, 'lot' could mean a gig, 10 big, or even a terrabyte.". My reply: "A terrabyte would be a good start... mumble". "WHAT!". "Yeash, how about a terrabyte now, and 10 terrabytes next week?".
The sysadmin fainted.
And this was only 5 years ago. Unfortunately, I/O speed has not kept up -- imagine trying to statically mirror two of these drives for a backup. (The video application used a lot of smaller drives, and naturally had a lot more heads, so it could maintain I/O rates. Even then the data had to be laid out carefully to maximize output speed).
Still, I am looking forward to my first 3TB RAID5 home storage unit.
A major hinderance?
Mice, graphics, networking JUST WORKS. You may not benefit from the "whiz-bang 3D super gaming feature". On the other hand, if this IS a concern to you, it puts you in very rarified company. Indeed, most people just use light word processing, maybe a digital camera, webcam, the web, and email. They want to play music and DVDs.
AND THAT'S IT. Are you wondering why Microsoft and others are planning "software as web service"? I will let you in on a really dark secret... Windows 98 would be "good enough" for most people, coupled with Openoffice.org (or Office 97) and Firefox, if it is behind a NAT router.
The Linux kernel has more drivers than Microsoft Vista kernel.
Linux distributions come with more software than Microsoft Vista distribution.
Openoffice.org is a very capable replacement for Microsoft Office. Yes, some things are easier in Microsoft Office. And I would expect that -- after all, Microsoft Office costs a whole lot more.
Yes, Firefox "does something". So does the Linux kernel. Specifically, the Linux kernel manages hardware resources. Mark me as failing your intelligence test.
Sorry you are wrong, the percentage calculation should be:
409 / 259,723 * 100 ~ 0.16%
By definition, 1/100 is 1%, 1/100 = 0.01 * 100 = 1
1996? Wow, have *I* been misled. Mid 1996 is the vintage of my Dual Pentium Pro 200Mhz, and I *really* thought that it was state-of-the-art.
Colour me disappointed...
If I wanted to share documents in 1999, or now...
I generate a PostScript file (possibly PDF) if I want "exact printing". I can test the PostScript against Adobe and Ghostscript (especially with gs' "safety" mode).
I generate an ASCII file, with illustrations in a neutral format (X bitmaps, XFig, JPEG, or PostScript).
I give a marked up document in (La)Tex, or TROFF.
(La)Tex is generally presumed "defect free", even though it does constitute a "monoculture" (Knuths bug pay-out policy contributes).
TROFF has two common implementations - AT&T (Solaris), and groff. Both considered very stable.
Ghostscript is very stable; I cannot speak to the Adobe implementation. There are other PostScript implementations available as well (Harlequin?)
These days, I still use (La)Tex for structured work, ODF for "ad-hoc" work (letters and throw-away memos, where the visual result is more important than sharing).
I view document preparation a bit like programming, in that the language should be considered as separate from the compiler implementation, and there should be tools to allow the language to be extended (elements by function or purpose, not format). Most wysiwyg systems fail at this (including Openoffice.org) which limits their usefulness to me.
Thanks for your comment. Yes, the opening statement is a "flame-ish" punch to get attention. My signature does mention "insightful troll", as a reflection of writing style.
And, I do say that PowerShell can be the best thing since sliced bread where it's appropriate. You are correct in that when all methods are available, PowerShell can be used as a fantastic integration tool.
It still fails in "ad-hoc" usage, except where "ad-hac" means common desktop command line use.
For my purposes (reading others log files, mostly), the Unix paradigm works better. And PowerShell also works here, but only as an implementation of the Unix paradigm. The object nature doesn't help (yet -- maybe log file implementors will introduce an object classification that everyone can agree with).
Have a GREAT weekend!