So the CRTC has been convinced that VOIP should be regulated.
By Rogers, the dominant Cable company (and thus data supplier).
Since Rogers sells the data pipe, what is the additional "surcharge" needed to support VOIP? Yes, the sale/rental of the converter.
Now, Bell has been told (by the CRTC) that they can't just sell you the VOIP adaptor.
Because... that would hurt Rogers business.
Its a big win by Rogers. Makes them a "Bell" in that they are protected now too.
What IS the value of VOIP (pricing). Why is is NOT a few pennies a month, to support some QOS infrastructure? I pay Rogers for 60GB of traffic a month (used to be unlimited) -- why can't I use that for voice communications? 3 hours of talk a day, compressed, would be 30MB of traffic. Just data.
Is there another reason why Rogers is concerned about WHAT the data is? Oh, yeah, I remember... they are a phone company too (cell phones). So, VOIP would be a great way to charge a LOT more for that 30MB per day.
I stand by my comment: requiring 3 - 5% or better tolerance is not acceptable. We had to hand-match Apple ][ drives back in '77 to ensure that the two drives could exchange data. First and last time I've ever had to do this.
As to "100%" reliable... it is possible to "accidentally" record a data pattern that duplicates the synchronization header. Just about the only system that this was possible on was the Apple ][.
(Side note: 3270 format "soft sector" came first. Followed by "hard sector". Apple didn't actually innovate -- except withe the strange "group encoding" chosen to allow the 1Mhz 6502 to decode the bit stream in software.)
However, saving the expense of (1) a proper floppy controller, (2) a proper video generator, (3) delegating everything to the 6502, did make the machine remarkably affordable.
Just not remarkably dependable.
But Apple does get cred for making the computer popular for home and school use.
Ratboy
PS. U an looking for an Apple ][ DOS 3.3 boot disk. Email me if you can make me a copy.
The apple used minimal circuitry for everything. Including floppy. Bit decode was done in software. Which meant that the drives had to be almost EXACTLY the same rotational speed, or they couldn't exchange disks.
As to soft sector being more reliable -- sorry, that's also not true. Hard sector is actually more reliable. Simple put, if a sector is damaged, it is possible to skip over it, and read sectors after it on the same track. Not possible (with ANY reliability) using a soft-sector format. As to cost savings? The hole detector is in the drive anyway (for either 1 hole, or n holes).
pic, grap (you will have to add), gnuplot, postscript.
All more complicated than OO. pic is a drawing language for troff. grap will do simple graphs (but you will have to download and install it), gnuplot is a workhorse for doing plots, but is many times more difficult than OO. And, you can always program in postscript directly! [and there are many more ways to make simple charts, including paint style programs).
Or, use OO.org
With OO.org, try:
Select some numbers that you want in a bar chart (highlight), then click (and hold) the icon on the left that looks like a pie-chart. A push-out menu appears, select the pie-chart within that. Position the chart (don't worry, you can move it later), and follow the wizard for the rest.
Best of all, the chart is live with your data -- change the data, the graph will change.
I'm curious as to why this didn't meet your needs...
Yes, your unix/linux box can be configured to automatically send dump information to a server. This is a useful feature -- but needs to be explicitly enabled.
If dump information from the Windows box can be sent to a central server that is controllable (eg. not an outside agency), then I am all for this feature -- plus I want to be able to disable this feature.
Same as being able to forward logging information (again, under installation control on those Unix boxes).
So, having this feature available is a "good thing" (tm) but I would like to see it configurable to use an internal server. Glad to see Microsoft is offering this as a problem resolution mechanism!
As a desktop for those without engineering degrees -- I would go to a locked-down Gnome configuration, running OpenOffice, on a Linux or BSD core.
Not because its cheap, but... (drum role)
because its locked down. I can preset what the user does, and manage it remotely.
For a data center, I would use Linux/Solaris (although NOT for high flexibility).
Indeed, the only time I would consider Windows is to run "ad hoc" software -- not mission critical, not revenue generating, etc.: games, spyware, chatware, P2Pware, and others.
rogers (the cable company here in toronto), just announced a 60GB data cap per month (should probably be a story to submit).
that works out to 100 cds, uncompressed. at a price of, say $5 average (we just want to cut out the actual brick and mortar store), this means that i am obviously downloading at least $500 worth of music per month.
i am not running any servers (no telnet, sshd, mail service, nntp, chat service or web pages) because that is against the agreement i have with rogers.
i can't be downloading movies, because, well, that would be illegal.
i only get a few megabytes of email every month. and my web browsing? if i get 10mb of data in a month, that's sure a lot of reading (mostly slashdot, thank you).
so, i must be using almost all of the 60gb of music, right?
so, my internet connection should be $540 per month.
So, what is that solution anyway (and, I agree that limiting that way is basically suckage)?
There are other ways of finding services - including broadcast, NIS maps, &etc. (and, no, I don't know how MS services are published; I am a "Unix head". I presume that MS has followed standard solutions here).
Anyway... if you *really* want to do it your way -- look for the machine that worked the last time first; then look at the machines in the arp cache, then broaden the search. I do beleive that MS machines natter at each other all the time, so the arp cache should have a good reflection of your local machines.
RIAA doesn't own anything. They simply represent the copyright holders.
And, It *does* matter where the bits come from... If you can show that a sequence of (say) 10,000 bits comes from a random number generator, it's yours.
If that *same* sequence of bits is sampled, copyright comes into force.
So, the "ownership" of the copyright comes from the source of the bits.
The RBL is a sledgehammer; brought down on an ISP.
If an ISP tolerates SPAMMERS for longer than a set time (I think serveral milliseconds is reasonable, but I could be a bit over the edge), the ISP looses email connectivity. Period.
No network effect for you! And ALL of your customers.
If it is important, and you know me, send it to another account -- you know, the ones that don't check -- or phone, or use regular mail.
If you run a "list", reconsider. It may not work well. Publish the information on the web instead (use pull technologies, not push). In the post-spam world, push is reserved for people who really want it. Who actually invest in it (setting up their own servers, and buying blackberries/cell-phones etc).
If SPAMMERS take over machines on a Cable ISP -- block the whole damn thing! Yes, Gran and Gramps may get upset, and that IS the point. (oh, you say, RBLs already DO THIS! Damn straight).
180,000 IP block, or 10x that; it is a sledgehammer. Use it. Hell, our local DSL provider (Bell Symaptico) COMPLETELY blocks port 25 out-going AND in-coming. And that's the way it is. [they avoid the sledgehammer, by making it impossible to be hit. Good for now, and when we win over the SPAMMERS, they can be more reasonable].
This is a war, kids, and its not finished yet.
If you are discomforted -- blaim the SPAMMERS. Fucking bottom-feeders.
You want the receiver to distinguish between an "authorized" broadcast and an "unauthorized" broadcast.
So the broadcast must be tagged in such a way as to allow this identification to be made (since potential transmitters cannot be controlled -- as you observed).
Encrypt the broadcast, or transmit a hash (or equivalent). Be sure to consider "replay attacks". An old message can be recorded and played back to the receiver at a later time.
And that's really all that you can do. Be sure to choose a cipher or hash technique that will outlive the useful life of the equipment as well. Once the signal has been sent, you must assume that an outside party has it; and can study it at their convenience. Simply because potential receivers cannot be controlled, either.
* Apple makes hardware that looks good, but (generally) sucks. Goes right back to the Apple ][. If they can finally build reasonable hardware -- good on them. The Apple ][ suxored (software floppies?!?), the Mac suxored (only 128KB ?!?) the Mac II suxored (unshielded cables, unshielded power supplies. Mac OS suxored (compared with Unix, compared with Smalltalk). Apple gave us exploding laptops! (major suxor!). Am I to believe that they can produce quality now?
* Apple software is "servicable", but not brilliant. It is closed, and generally refuses to interoperate (remember, this is the company that gave us the 'cr' delimiter, and resource forks)
* Apple needs the fanboy/grrl base. Without that they would have been so dead...
* Apple may be using an "open" core, but they won't open up the GUI for cross compatibility. I want to run Apple applications on my Sparc Solaris 9 box (at least the GUI), or from my Redhat box, or my AIX boxes, or my HPUX (well, you get the idea -- I have a few more different systems in the lab as well). But I can't. I could use the Apple as my desktop machine, but it won't play with my preferred keyboard and mouse (Lexmark Model M). Come on, open up the GUI client, guys!
to call up whatever browser you want for the help system.
And, the option should allow running the help subsystem as whatever "user" (priviledge level) you want. Or, run at the priviledges of the current user (but no higher).
So:
Is there a simple configuration (or registry) entry that does this? What privilege level is the help browser run at in MS Word?
Same thing with "embedded media". Does MS Word (or the OS) search out a compliant filter? And what privilege level is used for the filter? Is it the same as (say) Redhat 9, where I use "plugger" to parse a compliancy table, and can run separate programs of my choosing? Or is it "integrated" into the OS Core, and I have no such control?
Just wondering (I really don't know -- being a Solaris/Linux guy).
"Redirects to a page should be treated as having far less PageRank value than the page itself. That will fix the problem."
Won't work. In a nutshell (expounded upon below) -- Google doesn't know what the "real page" is.
Specifically, I generate "301" on my own sites:
http://blurt.org/page_of_files
will generate a 301 redirect to:
http://blurt.org/page_of_files/
which then gets an index of files. Ok?
Now, sometimes I relocate data (popular big files):
http://blurt.org/p_o_f/bigfile.html
would get a 302 redirect to
http://big_ass_isp.com/myplace/bigfile.html
I don't want to loose my "pagerank" over the content location! And I am doing it to myself here...
Google will look at my site, and see the redirect. It spiders the page, and it is under MY url (where it should be).
But, if Google spiders the OTHER site (say, though an automatically generated links page), it may eliminate the proper URL.
But, REDUCING "pagerank" because I am providing for more bandwidth? (which is what you are suggesting). Why on earth would anyone want that?
Now, what SHOULD happen is that the page should have a tag on it that says "Please index me if you accessed me directly" or "Please index me if you accessed me via REDIRECT from there".
That would do it (I think -- not much of a web master, I'm afraid).
As it is, since the big_ass_isp site is under my control, I place a robots.txt file there to prevent Google from spidering big_ass_isp view of my pages.
But, I can't control other people from generating a REDIRECT to me (but I *can* additionally tag the page).
I don't think that there is much that Google can do about this. (Well, they could honour a custom tag - and propose of the W3 folk, which is what I think will happen).
So, its the fault of "old-school" Unix hacks (now managers).
Ok -- I am one of those (grey hair, beard, used to be a Staff Engineer at SUN).
Now, Unix *has* been open. Open implementations, open specs. There was a strange kerfuffle with AT&T, along with some restrictions on Minix (that gave us Linux).
But -- we thought that EVEN if software wasn't redistributable, it should come with source. After all, its kind of useless without it. We thought that the OS itself is a commoditity. Unix is Unix is... Unix. Different flavour, same great taste. May not be the best, but certainly better than the rest.
Even DEC VMS came with source, for $DEITY sake! (on microfiche, but it *was* delivered). For $DEITY sake, it just needed gentle ASKING to get the source for SunOS!
It wasn't until these new-fangled micros came out that the source was COMPLETELY closed. CP/M-80 came as a binary only! The horror. So did MS-DOS, and Windows. Couldn't even get the source on microfiche for reference.
Now, as it turns out, Microsoft is a contemporary of SUN (I think Microsoft predates SUN by a bit). Still, the philosophy is different. SUN builds computers. They happen to need an OS. Microsoft builds OSs (but not for SUN computers).
Now, lets go one level deeper. The SPARC architecture is open. Windows is closed. MIPs is open; Intel is closed.
Yes, I have made money in the closed world. Shameful, but people seem to like it (check out all the games available on Windows). I just take offense to being pointed to as the "culprit" here.
We had user groups devoted to sharing source before you "younger folks" were born. Remember DECUS? Remember SHARE? All "open source" or OSS as you would see it today. Including OS, compilers, and application code.
Ratboy (not because I am young -- because I get obsessed with detail)
I consider myself an expert assembly level programmer. I recently took on a job optimizing some C code. The code was the back end bi-level compression code in a JBIG library. Bit twiddling, and updates to a compression table.
No big deal, I though. Unrolled the inner loop, and extracted patterns (as in "this case cannot possibly happen on the next n bits being 0, so bulk update and skip forward"). I wrote a program to generate the cases of interest, and tested.
Very good results -- client is happy. I converted to assembler, hand scheduling instructions. On a Pentium II (400Mhz) that I used for my speed test machine, I handily beat GCC (3.something). By a factor of 2, which was expected. Compiling the code using Microsoft C from the.NET distribution... they beat me. By a healthy margin. Almost 20% in the inner loop!
Analyzing the result and checking Intel documentation... It shouldn't have been that way. However, Microsoft may have better connections with Intel, and therefore better documentation on the processor.
On code that is vector code -- yes, I will agree with your point. But I think that advantage will fall by the wayside as compiler vendors implement C99 features, and auto-vectorization.
But don't sell the current crop of C compilers short - in particular, MSVC 7 is VERY VERY good at generating x86 code.
The problem is that email is only useful with the "network effect". Of *course* I can implement this (spam free email) -- and be an island of one. Indeed, even the smallest step towards this is not tolerated.
(which would be -- grab my public key from the MIT key server, and encrypt mail you want to send to me. Personally, I would be FASCINATED to see what spam would warrant encryption and the CPU cycles that it entails!).
And, I won't "sell out" the technology. Which means that I, and ONLY I, control my mail server.
So you can have your spam-free world, but (1) you will be talking to less people and/or (2) you will have no control.
Solve all three of these problems and you will have a winner.
The same problem will afflict VOIP. But, people seem willing to give up control of the technology (witness the purchase of VOIP technology from middle vendors, who cannot possibly be adding value to the IP based connection you ALREADY have!).
Slashdot Mirror
So the CRTC has been convinced that VOIP should be regulated.
By Rogers, the dominant Cable company (and thus data supplier).
Since Rogers sells the data pipe, what is the additional "surcharge" needed to support VOIP? Yes, the sale/rental of the converter.
Now, Bell has been told (by the CRTC) that they can't just sell you the VOIP adaptor.
Because... that would hurt Rogers business.
Its a big win by Rogers. Makes them a "Bell" in that they are protected now too.
What IS the value of VOIP (pricing). Why is is NOT a few pennies a month, to support some QOS infrastructure? I pay Rogers for 60GB of traffic a month (used to be unlimited) -- why can't I use that for voice communications? 3 hours of talk a day, compressed, would be 30MB of traffic. Just data.
Is there another reason why Rogers is concerned about WHAT the data is? Oh, yeah, I remember... they are a phone company too (cell phones). So, VOIP would be a great way to charge a LOT more for that 30MB per day.
Ratboy
blue meanie:
I stand by my comment: requiring 3 - 5% or better tolerance is not acceptable. We had to hand-match Apple ][ drives back in '77 to ensure that the two drives could exchange data. First and last time I've ever had to do this.
As to "100%" reliable... it is possible to "accidentally" record a data pattern that duplicates the synchronization header. Just about the only system that this was possible on was the Apple ][.
(Side note: 3270 format "soft sector" came first. Followed by "hard sector". Apple didn't actually innovate -- except withe the strange "group encoding" chosen to allow the 1Mhz 6502 to decode the bit stream in software.)
However, saving the expense of (1) a proper floppy controller, (2) a proper video generator, (3) delegating everything to the 6502, did make the machine remarkably affordable.
Just not remarkably dependable.
But Apple does get cred for making the computer popular for home and school use.
Ratboy
PS. U an looking for an Apple ][ DOS 3.3 boot disk. Email me if you can make me a copy.
Now that's a hoot.
The apple used minimal circuitry for everything. Including floppy. Bit decode was done in software. Which meant that the drives had to be almost EXACTLY the same rotational speed, or they couldn't exchange disks.
As to soft sector being more reliable -- sorry, that's also not true. Hard sector is actually more reliable. Simple put, if a sector is damaged, it is possible to skip over it, and read sectors after it on the same track. Not possible (with ANY reliability) using a soft-sector format. As to cost savings? The hole detector is in the drive anyway (for either 1 hole, or n holes).
And, yes, my Apple ][ still works.
Ratboy
OpenOffice.org spreadsheet was not suitable?
You need bar graphs --
pic, grap (you will have to add), gnuplot, postscript.
All more complicated than OO. pic is a drawing language for troff. grap will do simple graphs (but you will have to download and install it), gnuplot is a workhorse for doing plots, but is many times more difficult than OO. And, you can always program in postscript directly! [and there are many more ways to make simple charts, including paint style programs).
Or, use OO.org
With OO.org, try:
Select some numbers that you want in a bar chart (highlight), then click (and hold) the icon on the left that looks like a pie-chart. A push-out menu appears, select the pie-chart within that. Position the chart (don't worry, you can move it later), and follow the wizard for the rest.
Best of all, the chart is live with your data -- change the data, the graph will change.
I'm curious as to why this didn't meet your needs...
Ratboy
It's an insight. It's funny. It poignant.
Ratboy.
And boy did THAT blow up in American Greetings face. The NEXT parody was directly aimed for the genitals...
I am sure that American Greetings is going to think long and hard about the results the next time...
Ratboy.
Well...
Yes, your unix/linux box can be configured to automatically send dump information to a server. This is a useful feature -- but needs to be explicitly enabled.
If dump information from the Windows box can be sent to a central server that is controllable (eg. not an outside agency), then I am all for this feature -- plus I want to be able to disable this feature.
Same as being able to forward logging information (again, under installation control on those Unix boxes).
So, having this feature available is a "good thing" (tm) but I would like to see it configurable to use an internal server. Glad to see Microsoft is offering this as a problem resolution mechanism!
Ratboy.
As a desktop for those without engineering degrees -- I would go to a locked-down Gnome configuration, running OpenOffice, on a Linux or BSD core.
Not because its cheap, but... (drum role)
because its locked down. I can preset what the user does, and manage it remotely.
For a data center, I would use Linux/Solaris (although NOT for high flexibility).
Indeed, the only time I would consider Windows is to run "ad hoc" software -- not mission critical, not revenue generating, etc.: games, spyware, chatware, P2Pware, and others.
But YMMV
Ratboy.
Ontario, Canada...
7 million seats
Big enough?
Ratboy
VMS source was distributed on microfiche.
It seemed to be written in MACRO-11 and BLISS/32.
Of course, my old memory may be failing me...
Ratboy.
Um... CA *did* hire Mr. Torvalds. CA funds (a part of) OSDL, which Torvalds work for.
So, they *do* want attention paid to their issues.
So there you go.
Ratboy.
Let's see.
rogers (the cable company here in toronto), just announced a 60GB data cap per month (should probably be a story to submit).
that works out to 100 cds, uncompressed. at a price of, say $5 average (we just want to cut out the actual brick and mortar store), this means that i am obviously downloading at least $500 worth of music per month.
i am not running any servers (no telnet, sshd, mail service, nntp, chat service or web pages) because that is against the agreement i have with rogers.
i can't be downloading movies, because, well, that would be illegal.
i only get a few megabytes of email every month. and my web browsing? if i get 10mb of data in a month, that's sure a lot of reading (mostly slashdot, thank you).
so, i must be using almost all of the 60gb of music, right?
so, my internet connection should be $540 per month.
glad that's settled.
ratboy
So, what is that solution anyway (and, I agree that limiting that way is basically suckage)?
There are other ways of finding services - including broadcast, NIS maps, &etc. (and, no, I don't know how MS services are published; I am a "Unix head". I presume that MS has followed standard solutions here).
Anyway... if you *really* want to do it your way -- look for the machine that worked the last time first; then look at the machines in the arp cache, then broaden the search. I do beleive that MS machines natter at each other all the time, so the arp cache should have a good reflection of your local machines.
YMMV
Ratboy.
RIAA doesn't own anything. They simply represent the copyright holders.
And, It *does* matter where the bits come from... If you can show that a sequence of (say) 10,000 bits comes from a random number generator, it's yours.
If that *same* sequence of bits is sampled, copyright comes into force.
So, the "ownership" of the copyright comes from the source of the bits.
Ratboy
If the drive needs service, it's probably too late.
The data has to be written to the drive encrypted, look into AES-256 on Linux (possibly with tweak).
Some hardware assist may help to get you going: my company makes a data assurance solution (plug).
In any case, with an encrypted solution: The data keys are revoked, and Bob's your Uncle - the data is gone. You need to secure your key repository.
If the drive needs service, writes may fail, so any "pre-service" solution won't be secure.
So, bring out the sledge, and whack that thing!
Ratboy
That is the way... And it is good.
The RBL is a sledgehammer; brought down on an ISP.
If an ISP tolerates SPAMMERS for longer than a set time (I think serveral milliseconds is reasonable, but I could be a bit over the edge), the ISP looses email connectivity. Period.
No network effect for you! And ALL of your customers.
If it is important, and you know me, send it to another account -- you know, the ones that don't check -- or phone, or use regular mail.
If you run a "list", reconsider. It may not work well. Publish the information on the web instead (use pull technologies, not push). In the post-spam world, push is reserved for people who really want it. Who actually invest in it (setting up their own servers, and buying blackberries/cell-phones etc).
If SPAMMERS take over machines on a Cable ISP -- block the whole damn thing! Yes, Gran and Gramps may get upset, and that IS the point. (oh, you say, RBLs already DO THIS! Damn straight).
180,000 IP block, or 10x that; it is a sledgehammer. Use it. Hell, our local DSL provider (Bell Symaptico) COMPLETELY blocks port 25 out-going AND in-coming. And that's the way it is. [they avoid the sledgehammer, by making it impossible to be hit. Good for now, and when we win over the SPAMMERS, they can be more reasonable].
This is a war, kids, and its not finished yet.
If you are discomforted -- blaim the SPAMMERS. Fucking bottom-feeders.
Ratboy
You want the receiver to distinguish between an "authorized" broadcast and an "unauthorized" broadcast.
So the broadcast must be tagged in such a way as to allow this identification to be made (since potential transmitters cannot be controlled -- as you observed).
Encrypt the broadcast, or transmit a hash (or equivalent). Be sure to consider "replay attacks". An old message can be recorded and played back to the receiver at a later time.
And that's really all that you can do. Be sure to choose a cipher or hash technique that will outlive the useful life of the equipment as well. Once the signal has been sent, you must assume that an outside party has it; and can study it at their convenience. Simply because potential receivers cannot be controlled, either.
Advice: Hire a cryptographer.
Ratboy
* Apple makes hardware that looks good, but (generally) sucks. Goes right back to the Apple ][. If they can finally build reasonable hardware -- good on them. The Apple ][ suxored (software floppies?!?), the Mac suxored (only 128KB ?!?) the Mac II suxored (unshielded cables, unshielded power supplies. Mac OS suxored (compared with Unix, compared with Smalltalk). Apple gave us exploding laptops! (major suxor!). Am I to believe that they can produce quality now?
* Apple software is "servicable", but not brilliant. It is closed, and generally refuses to interoperate (remember, this is the company that gave us the 'cr' delimiter, and resource forks)
* Apple needs the fanboy/grrl base. Without that they would have been so dead...
* Apple may be using an "open" core, but they won't open up the GUI for cross compatibility. I want to run Apple applications on my Sparc Solaris 9 box (at least the GUI), or from my Redhat box, or my AIX boxes, or my HPUX (well, you get the idea -- I have a few more different systems in the lab as well). But I can't. I could use the Apple as my desktop machine, but it won't play with my preferred keyboard and mouse (Lexmark Model M). Come on, open up the GUI client, guys!
Ratboy
There should be an option --
to call up whatever browser you want for the help system.
And, the option should allow running the help subsystem as whatever "user" (priviledge level) you want. Or, run at the priviledges of the current user (but no higher).
So:
Is there a simple configuration (or registry) entry that does this? What privilege level is the help browser run at in MS Word?
Same thing with "embedded media". Does MS Word (or the OS) search out a compliant filter? And what privilege level is used for the filter? Is it the same as (say) Redhat 9, where I use "plugger" to parse a compliancy table, and can run separate programs of my choosing? Or is it "integrated" into the OS Core, and I have no such control?
Just wondering (I really don't know -- being a Solaris/Linux guy).
Ratboy
"Redirects to a page should be treated as having far less PageRank value than the page itself. That will fix the problem."
Won't work. In a nutshell (expounded upon below) -- Google doesn't know what the "real page" is.
Specifically, I generate "301" on my own sites:
http://blurt.org/page_of_files
will generate a 301 redirect to:
http://blurt.org/page_of_files/
which then gets an index of files. Ok?
Now, sometimes I relocate data (popular big files):
http://blurt.org/p_o_f/bigfile.html
would get a 302 redirect to
http://big_ass_isp.com/myplace/bigfile.html
I don't want to loose my "pagerank" over the content location! And I am doing it to myself here...
Google will look at my site, and see the redirect. It spiders the page, and it is under MY url (where it should be).
But, if Google spiders the OTHER site (say, though an automatically generated links page), it may eliminate the proper URL.
But, REDUCING "pagerank" because I am providing for more bandwidth? (which is what you are suggesting). Why on earth would anyone want that?
Now, what SHOULD happen is that the page should have a tag on it that says "Please index me if you accessed me directly" or "Please index me if you accessed me via REDIRECT from there".
That would do it (I think -- not much of a web master, I'm afraid).
As it is, since the big_ass_isp site is under my control, I place a robots.txt file there to prevent Google from spidering big_ass_isp view of my pages.
But, I can't control other people from generating a REDIRECT to me (but I *can* additionally tag the page).
I don't think that there is much that Google can do about this. (Well, they could honour a custom tag - and propose of the W3 folk, which is what I think will happen).
Ratboy
So, its the fault of "old-school" Unix hacks (now managers).
Ok -- I am one of those (grey hair, beard, used to be a Staff Engineer at SUN).
Now, Unix *has* been open. Open implementations, open specs. There was a strange kerfuffle with AT&T, along with some restrictions on Minix (that gave us Linux).
But -- we thought that EVEN if software wasn't redistributable, it should come with source. After all, its kind of useless without it. We thought that the OS itself is a commoditity. Unix is Unix is... Unix. Different flavour, same great taste. May not be the best, but certainly better than the rest.
Even DEC VMS came with source, for $DEITY sake! (on microfiche, but it *was* delivered). For $DEITY sake, it just needed gentle ASKING to get the source for SunOS!
It wasn't until these new-fangled micros came out that the source was COMPLETELY closed. CP/M-80 came as a binary only! The horror. So did MS-DOS, and Windows. Couldn't even get the source on microfiche for reference.
Now, as it turns out, Microsoft is a contemporary of SUN (I think Microsoft predates SUN by a bit). Still, the philosophy is different. SUN builds computers. They happen to need an OS. Microsoft builds OSs (but not for SUN computers).
Now, lets go one level deeper. The SPARC architecture is open. Windows is closed. MIPs is open; Intel is closed.
Yes, I have made money in the closed world. Shameful, but people seem to like it (check out all the games available on Windows). I just take offense to being pointed to as the "culprit" here.
We had user groups devoted to sharing source before you "younger folks" were born. Remember DECUS? Remember SHARE? All "open source" or OSS as you would see it today. Including OS, compilers, and application code.
Ratboy
(not because I am young -- because I get obsessed with detail)
When the cable industry started up in Canada, the cable companies slurped up US signals, and distributed them to Canadian cable subscribers.
No, the US providers where not paid.
Furthermore, US ads where replaced with local ads.
Niiice.
And now that industry has the HUTZPAH to complain that I download TV shows!?!
And, its offtopic - and I don't anything about Aussie IP law, anyway.
Ratboy
Since the job was T&M, I didn't have the freedom to explore Intels compiler. Sorry -- and it would have been very interesting.
Ratboy.
Ameline:
.NET distribution... they beat me. By a healthy margin. Almost 20% in the inner loop!
Um... not true.
I consider myself an expert assembly level programmer. I recently took on a job optimizing some C code. The code was the back end bi-level compression code in a JBIG library. Bit twiddling, and updates to a compression table.
No big deal, I though. Unrolled the inner loop, and extracted patterns (as in "this case cannot possibly happen on the next n bits being 0, so bulk update and skip forward"). I wrote a program to generate the cases of interest, and tested.
Very good results -- client is happy. I converted to assembler, hand scheduling instructions. On a Pentium II (400Mhz) that I used for my speed test machine, I handily beat GCC (3.something). By a factor of 2, which was expected. Compiling the code using Microsoft C from the
Analyzing the result and checking Intel documentation... It shouldn't have been that way. However, Microsoft may have better connections with Intel, and therefore better documentation on the processor.
On code that is vector code -- yes, I will agree with your point. But I think that advantage will fall by the wayside as compiler vendors implement C99 features, and auto-vectorization.
But don't sell the current crop of C compilers short - in particular, MSVC 7 is VERY VERY good at generating x86 code.
Ratboy (hanging up the assembler badge, sadly).
And that's not the problem.
The problem is that email is only useful with the "network effect". Of *course* I can implement this (spam free email) -- and be an island of one. Indeed, even the smallest step towards this is not tolerated.
(which would be -- grab my public key from the MIT key server, and encrypt mail you want to send to me. Personally, I would be FASCINATED to see what spam would warrant encryption and the CPU cycles that it entails!).
And, I won't "sell out" the technology. Which means that I, and ONLY I, control my mail server.
So you can have your spam-free world, but (1) you will be talking to less people and/or (2) you will have no control.
Solve all three of these problems and you will have a winner.
The same problem will afflict VOIP. But, people seem willing to give up control of the technology (witness the purchase of VOIP technology from middle vendors, who cannot possibly be adding value to the IP based connection you ALREADY have!).
And in that there is a solution.
Just some shit for you to ponder...
Ratboy.