Users misspell domain names, causing actions they do not want...
so, what technical solutions could we use for this problem? We could of course leave it as it is, a la survival of the fittest. Or we could try educating users not to mispell (fat chance). But are there other options?
Credit card companies and banks have been dealing with somewhat similar probems, their solution usually involves checksum digits included in each number
Could we apply a simmilar system to domain names?
i.e advertise a www address as httX:/Y/zzz.com where httX tells the browser that the next character is a checksum, and Y is the one-digit checksum for "zzz.com"?
users of older browsers would still be able to visit such a domain by leaving out the checksum
Or, make a new top level domain,.check, where the second-level-domain has to be a checksum for the rest? http://zzz.Y.check
this would require no changes in current software, but would require companies to use something else but.com in their domains.
looks good, but the keyboard layout is problematic
It only allows the y and z keys for jump and fire, and those keys are very uncomfortable to press with on hand on a german keyboard. changing the keyboard layout under windows has no effect on the game.
As mentioned in another post, "selling" was a bad choice of words, what I meant is of course "making available".
As to the withdrawing: You can not withdraw something with the current copyright. Once you have distributed it, the people who received it may keep it forever, or give it to someone else. Once the copies have been distributed, there will always be some on the market.
All you can prevent (and for a limited time only) is the production of additional copies, until the copyright expires.
(Not to mention people may still quote your work, make parodies of it, and a number of other things covered by fair use)
So you can only make it harder and more expensive for someone to get a copy, not impossible.
You are right that this would not be possible with the system I proposed. This is intentional.
you create something for the public, and in exchange the public grants you a time-limited monopoly on the production of copies of said work. If you do not not want to provide the public with said work, why should the public be interrested in maintaining your monopoly on it?
If you create something, then you are the owner of it, you automatically have a moral right to the object. you can keep it, sell it, copy it, destroy it, whatever. few ethic schools of thought will deny you this right.
Bat what you do not have by default, is the exclusive right over every copy of your creation.
This right is granted to you by the public for a limited time only, it is essentially a contract with the the public, as a reward for creating something for the public. This contract is a relatively recent invention.
And this contract is it which entitles the public to certain demands.
Or to put it another way: Noone needs Jesica Simpson's latest work (whatever that is), not even she herself. But she may of course keep it indefinitelly, since it is hers. What she however may not do, is to keep others from having a copy of her work forever. This is the way it is with current copyright
All we are arguing here are the exact terms of "not forever". I say that the current terms are too much a disadvantage for the public to justify continuing this contract between the public and the creators.
The question is, would people create less with a more limited copyright? Or would they create more, since they could build more on the works of others?
Seems like this was one of the guys who made the music for the starcraft broodwar intro "morituri te salutant". One of my favorite songs, so anything he makes just has to be good:)
you can get it from blizzard btw: bwintro.zip More songs from blizzard here
ah yes, seems like we would need some sort of exceptions for this. Maybe only apply it to works available to the public at one time. So in-house solutions would not be affected either.
Still, these non-published works should still be submitted to the copyright office for archival, otherwise when the copyright expires, there won't be anyone around to release it.
Copyright grabts you an exclusive first-sale right. If you stopp selling something, then obviously you do not want this right anymore.
Copyright should include an obligation to make the work available to the public, in exchange for the exclusive sale right granted to the copyright holder by the public.
I see no reason why "hording" should be allowed. If the copyright owner is not interested in selling the work anymore, be it book, game, software, or whatever, then copyright should expire.
"hording" a work would still be possible by selling it only for exorbitant prices intended to scare of everyone, so this is not a perfect solution. But it would require more effort, and orphaned works would be impossible
But what if you do not have any control over either firewall? Say, A is you at work, or at a public access point that only offers nat, and B is the person you want to help who has no idea how to configure his router.
If A sends a SYN, it will be blocked by B's firewall, but thats why they have to agree bevorehand about the sequence number and the time to send the ACK.
The following ACK from B should not get blocked, since B'S firewall does not block outgoing packets (Assuming standard nat, and no additional blocks), and A's firewall will see it as the response to the previous SYN from A.
These sollutions all seem to work only if only one side is behind a nat, or if a dedicated third party server is available that both sides can use.
From my rudimantary understanding of tcp/ip, I am wondering if the following would work too:
A and B are behind a NAT or a firewall that blocks all incomming connections.
Asuming A and B have some (inefficient) way to communicate, like email:
- A and B agree on a TCP sequence number and a time per mail.
- Both A and B send a SYN with that number at the defined time. Both SYNs are blocked by the firewall on the other end.
- Both A and B send a ACK with the correspondending sequence number, as if they had received each others SYN. The ACK should not get blocked by the firewall, since it is a response to a previous SYN.
- Both A and B send a SYN/ACK as usual.
Now A and B should have an open TCP connection between them (or two, but since they have the same sequence no. they might be indistinguishable)
As long as they keep it open, it should allow them to communicate despite the firewalls.
You are wrong, but not in this statement directly. Yes, you will be prevented from sending emails with your gmail address from your SMTP server. What you are wrong about is doing this in the first place; it breaks the SMTP standard.
are you sure about that? from my skimming over rfc 821 and 2821, neither forbids or even discourages it.
Well, this is google, so they just might give me a key.
reply-to has the big disadvantage that I would then have to monitor two adresses for replies, Because some people will inevitably mess it up and reply to the wrong one. And then I could just as well switch completely to the other address.
There are how many million domain names registered?
As long as not ALL of them implement this, spamers will always have domain names they can use. And if this goes at the same speed of ipv6...
The bounce traffic wil end up with these domains instead of the protected ones, wich I guess is good for domain owners, but it wil NOT reduce the volume of spam, or force spammers to identify themselves in any way.
For domain owners it might be a good thing, but for the average user, it has no advantages.
Correct me if I am wrong, but if I understand this correctly, and if filtering with this becomes widely adopted, then it will also prevent me from sending mails with my gmail-address from my smtp server. So I would have to use their web-interface, or hope they wil eventually make a smtp-server useable for a fee
Not that this is not their right and all, and I could just stop using it if I don' like it, free service, yada yada.. Still, this gives a little too much control to my email-domain-provider about which smtp-server I can use, than I am comfortable with.
In a perfect world, the amount of legaleze I need to agree on would have some relation wih the product. a 10 page contract, that requies in depth study and advanced knowledge of the law to fully understand, might be appropriate for a million dollar deal, but not for a 20$ piece of software from walmart.
of course I have no good idea how to enforce this.
a possibility would be to define certain standart conditions with descritive labels, that companies could use, like NO-WARRANTY, MUST-NOT-COPY, ONLY-NONCOMMERCIAL. then limit the additional legaleze to only a few paragraphs for "minor" purchases by law.
(this draft of a "solution" has numerous problems. still, what are your thougths?)
It's very easy to resize a PDF. Just select "fit to page".
sure, if the size you want to refit it to happens to have the same proportions. Otherwise you will get borders.
and about the commercial apps: They have to rely on OCR-like features
No, you can ger the text directly (in most cases), unless the publisher has deliberately obfuscated it.
I am not talking about plain text. that's why I said that they share all the weaknesses for layout and markup recognition that common OCR tools have
Lots of tools, even built in to the printer driver, to do this, 2-up, 4-up, etc. (PDF were designed fro publishing, you need to do this a lot.)
Sure, if I want to simply print 2to4 or something like this, I can do this with any sort of document, even with gifs. But what if I want to print it in a slightly smaller font size? or remove the 6 cm footer that is for some reason at the bottom of each page? or print it without the huge photos that the publisher choose to include on every second page, even thought they serve no purpose?
And I didn't even mention the numerous problems that a pdf with complex tables gives to blind people so far, despite the tools suplied by adobe.
It is just my opinion that a proper format specification (and lots of other documents)should contain meta-data. Otherwise they are unsuitable for anything besides printing of viewing it exactly as they were created. And the orginal document most likely did contain this meta-data, so why throw it away?
Why would you want to do this anyway? You'll just end up with anoher document.
We are talking about protocol specs here. Can you not think of any valid reasons to export information from such a spec? Do you not think it would be usefull to be able to quote any part of the spec, be it plain text, formula, or table, or to combine parts of the spec with some other sources of information? Is it not a valid request to reformat such a spec to my liking, so that I can read it easier, print it on less paper, make a presentation out of it? Or maybe i want to print out or view all the lines that contain $foo, without the rest. Or I might want to create a summary that only contains the parts of the document that are important for a certain task.
Heck, in some cases it is difficult to even change the paper format of a pdf without affecting the layout.
PDF is only usefull if you never expect anybody to do anything else besides printing or viewing it on screen. In most cases where PDF is used nowadays, this asumption is false (or the publisher just does not care). And it is certainly false in the case of protocoll specifications.
and about the commercial apps: They have to rely on OCR-like features for this task, and share all the weaknesses for layout and markup recognition that common OCR tools have. A PDF file just does not contain any contextual meta-information, the best they can do is guess.
Try to export a table from a pdf, and you might change your mind heck, try to export text in columns at all, or mathematic formulas, or anything besides plain text or graphics. Then try to do that automatically for a lot of pages.
There are workarounds, but they are ugly and mostly rely on OCR.
Slashdot Mirror
Users misspell domain names, causing actions they do not want...
.check, where the second-level-domain has to be a checksum for the rest?
.com in their domains.
so, what technical solutions could we use for this problem?
We could of course leave it as it is, a la survival of the fittest. Or we could try educating users not to mispell (fat chance). But are there other options?
Credit card companies and banks have been dealing with somewhat similar probems, their solution usually involves checksum digits included in each number
Could we apply a simmilar system to domain names?
i.e advertise a www address as
httX:/Y/zzz.com
where httX tells the browser that the next character is a checksum, and Y is the one-digit checksum for "zzz.com"?
users of older browsers would still be able to visit such a domain by leaving out the checksum
Or, make a new top level domain,
http://zzz.Y.check
this would require no changes in current software, but would require companies to use something else but
Any other ideas? What do you think?
Sounds great! But is it also 100% Debian compatible?
meant to say x and z.
y and z are swapped on a german keyboard
But nevermind, found my old gamepad
nice game
looks good, but the keyboard layout is problematic
It only allows the y and z keys for jump and fire, and those keys are very uncomfortable to press with on hand on a german keyboard.
changing the keyboard layout under windows has no effect on the game.
has anyone found out a way to change that?
Oh no! It is lovebot!
Run for your lives!
(stolen from Bob the Angry Flower)
As mentioned in another post, "selling" was a bad choice of words, what I meant is of course "making available".
As to the withdrawing: You can not withdraw something with the current copyright. Once you have distributed it, the people who received it may keep it forever, or give it to someone else. Once the copies have been distributed, there will always be some on the market.
All you can prevent (and for a limited time only) is the production of additional copies, until the copyright expires.
(Not to mention people may still quote your work, make parodies of it, and a number of other things covered by fair use)
So you can only make it harder and more expensive for someone to get a copy, not impossible.
You are right that this would not be possible with the system I proposed. This is intentional.
you create something for the public, and in exchange the public grants you a time-limited monopoly on the production of copies of said work. If you do not not want to provide the public with said work, why should the public be interrested in maintaining your monopoly on it?
The name is just from my flaky memory.
Usenet seems to confirm it, but I think i got it from somewhere else
IIRC the song starts with "Ave Caesar, morituri te salutant".
But maybe it does not have an official name at all.
If you create something, then you are the owner of it, you automatically have a moral right to the object.
you can keep it, sell it, copy it, destroy it, whatever. few ethic schools of thought will deny you this right.
Bat what you do not have by default, is the exclusive right over every copy of your creation.
This right is granted to you by the public for a limited time only, it is essentially a contract with the the public, as a reward for creating something for the public.
This contract is a relatively recent invention.
And this contract is it which entitles the public to certain demands.
Or to put it another way:
Noone needs Jesica Simpson's latest work (whatever that is), not even she herself.
But she may of course keep it indefinitelly, since it is hers. What she however may not do, is to keep others from having a copy of her work forever. This is the way it is with current copyright
All we are arguing here are the exact terms of "not forever".
I say that the current terms are too much a disadvantage for the public to justify continuing this contract between the public and the creators.
The question is, would people create less with a more limited copyright? Or would they create more, since they could build more on the works of others?
Seems like this was one of the guys who made the music for the starcraft broodwar intro "morituri te salutant". :)
One of my favorite songs, so anything he makes just has to be good
you can get it from blizzard btw: bwintro.zip
More songs from blizzard here
ah yes, seems like we would need some sort of exceptions for this.
Maybe only apply it to works available to the public at one time.
So in-house solutions would not be affected either.
Still, these non-published works should still be submitted to the copyright office for archival, otherwise when the copyright expires, there won't be anyone around to release it.
agreed. "Sale" was a bad choice of words, "make available" works better.
Ok, how about this approach?
Copyright grabts you an exclusive first-sale right. If you stopp selling something, then obviously you do not want this right anymore.
Copyright should include an obligation to make the work available to the public, in exchange for the exclusive sale right granted to the copyright holder by the public.
I see no reason why "hording" should be allowed. If the copyright owner is not interested in selling the work anymore, be it book, game, software, or whatever, then copyright should expire.
"hording" a work would still be possible by selling it only for exorbitant prices intended to scare of everyone, so this is not a perfect solution. But it would require more effort, and orphaned works would be impossible
What do you think?
But what if you do not have any control over either firewall?
Say, A is you at work, or at a public access point that only offers nat, and B is the person you want to help who has no idea how to configure his router.
If A sends a SYN, it will be blocked by B's firewall, but thats why they have to agree bevorehand about the sequence number and the time to send the ACK.
The following ACK from B should not get blocked, since B'S firewall does not block outgoing packets (Assuming standard nat, and no additional blocks), and A's firewall will see it as the response to the previous SYN from A.
At least that is how i figure it.
These sollutions all seem to work only if only one side is behind a nat, or if a dedicated third party server is available that both sides can use.
From my rudimantary understanding of tcp/ip, I am wondering if the following would work too:
A and B are behind a NAT or a firewall that blocks all incomming connections.
Asuming A and B have some (inefficient) way to communicate, like email:
- A and B agree on a TCP sequence number and a time per mail.
- Both A and B send a SYN with that number at the defined time. Both SYNs are blocked by the firewall on the other end.
- Both A and B send a ACK with the correspondending sequence number, as if they had received each others SYN. The ACK should not get blocked by the firewall, since it is a response to a previous SYN.
- Both A and B send a SYN/ACK as usual.
Now A and B should have an open TCP connection between them (or two, but since they have the same sequence no. they might be indistinguishable)
As long as they keep it open, it should allow them to communicate despite the firewalls.
this just means that 50% of p2p traffic is spam, making 33% of overall trafic both spam AND p2p. slashdot uses the remaining 1%.
sounds about right.
You are wrong, but not in this statement directly. Yes, you will be prevented from sending emails with your gmail address from your SMTP server. What you are wrong about is doing this in the first place; it breaks the SMTP standard.
are you sure about that? from my skimming over rfc 821 and 2821, neither forbids or even discourages it.
explicit steps, like writing down my address from memory, or from a printout of my mail?
that's hardly unlikey.
Well, this is google, so they just might give me a key.
reply-to has the big disadvantage that I would then have to monitor two adresses for replies, Because some people will inevitably mess it up and reply to the wrong one.
And then I could just as well switch completely to the other address.
There are how many million domain names registered?
As long as not ALL of them implement this, spamers will always have domain names they can use. And if this goes at the same speed of ipv6...
The bounce traffic wil end up with these domains instead of the protected ones, wich I guess is good for domain owners, but it wil NOT reduce the volume of spam, or force spammers to identify themselves in any way.
For domain owners it might be a good thing, but for the average user, it has no advantages.
Correct me if I am wrong, but if I understand this correctly, and if filtering with this becomes widely adopted, then it will also prevent me from sending mails with my gmail-address from my smtp server.
So I would have to use their web-interface, or hope they wil eventually make a smtp-server useable for a fee
Not that this is not their right and all, and I could just stop using it if I don' like it, free service, yada yada..
Still, this gives a little too much control to my email-domain-provider about which smtp-server I can use, than I am comfortable with.
In a perfect world, the amount of legaleze I need to agree on would have some relation wih the product. a 10 page contract, that requies in depth study and advanced knowledge of the law to fully understand, might be appropriate for a million dollar deal, but not for a 20$ piece of software from walmart.
of course I have no good idea how to enforce this.
a possibility would be to define certain standart conditions with descritive labels, that companies could use, like NO-WARRANTY, MUST-NOT-COPY, ONLY-NONCOMMERCIAL. then limit the additional legaleze to only a few paragraphs for "minor" purchases by law.
(this draft of a "solution" has numerous problems. still, what are your thougths?)
sure, if the size you want to refit it to happens to have the same proportions. Otherwise you will get borders.
I am not talking about plain text. that's why I said that they share all the weaknesses for layout and markup recognition that common OCR tools have
Sure, if I want to simply print 2to4 or something like this, I can do this with any sort of document, even with gifs. But what if I want to print it in a slightly smaller font size? or remove the 6 cm footer that is for some reason at the bottom of each page? or print it without the huge photos that the publisher choose to include on every second page, even thought they serve no purpose?
And I didn't even mention the numerous problems that a pdf with complex tables gives to blind people so far, despite the tools suplied by adobe.
It is just my opinion that a proper format specification (and lots of other documents)should contain meta-data. Otherwise they are unsuitable for anything besides printing of viewing it exactly as they were created. And the orginal document most likely did contain this meta-data, so why throw it away?
Why would you want to do this anyway? You'll just end up with anoher document.
We are talking about protocol specs here.
Can you not think of any valid reasons to export information from such a spec?
Do you not think it would be usefull to be able to quote any part of the spec, be it plain text, formula, or table, or to combine parts of the spec with some other sources of information?
Is it not a valid request to reformat such a spec to my liking, so that I can read it easier, print it on less paper, make a presentation out of it?
Or maybe i want to print out or view all the lines that contain $foo, without the rest.
Or I might want to create a summary that only contains the parts of the document that are important for a certain task.
Heck, in some cases it is difficult to even change the paper format of a pdf without affecting the layout.
PDF is only usefull if you never expect anybody to do anything else besides printing or viewing it on screen.
In most cases where PDF is used nowadays, this asumption is false (or the publisher just does not care).
And it is certainly false in the case of protocoll specifications.
and about the commercial apps: They have to rely on OCR-like features for this task, and share all the weaknesses for layout and markup recognition that common OCR tools have.
A PDF file just does not contain any contextual meta-information, the best they can do is guess.
Try to export a table from a pdf, and you might change your mind
heck, try to export text in columns at all, or mathematic formulas, or anything besides plain text or graphics.
Then try to do that automatically for a lot of pages.
There are workarounds, but they are ugly and mostly rely on OCR.