Find some "gold" that is free supply, and figure a way to deny, then sell access.
That seems to be their philosophy. Make sure to exploit any possible claim, not make sure that internal R&D funding is sustainable by fair claims and licensing of its results.
I don't think patenting was originally conceived to serve their kind of "prospecting."
MTBF means too little. I want graphs, e.g., ...
on
Intel using FreeBSD
·
· Score: 1
Instead of a single MTBF number that doesn't say much to me, I'd like to see a graph of PCSTF (percent surviving trouble free) vs time. Then you could see the difference between bad QA shipping a solid design (+) and planned obsolescence (o), e.g. (for 36 months):
Or a scatter gram of first failures vs time, color coded with the kind of failure (disk, memory, mobo stuff, etc.) And for those systems capable of reporting soft failures, those would be interesting too.
Also a scatter gram of soft failures vs time backwards from same-item hard failure.
I wish Consumer Reports would do their stuff with cumulative graphs and scattergrams. They have enough data to show some interesting things.
I wonder how the MPG ratings would plot vs time after tuneup.
BTW, anyone know of a system with ECC memory that detects and reports soft error statistics? How do I know my ECC memory hasn't just been covering up a stuck single-bit error since the beginning?
The point is that name registration is not a "normal business." It is a critical e-world public utility function, and needs (apparently) to have acceptable practices and procedures defined for it, to prevent and/or punish abuses, and to ensure reliable and fair transactions.
If things are as bad as comments here on slashdot indicate, perhaps a petition for a proper review sent to the top of the regulatory pyramid and/or congressperson(s) on relevant committes would result in a better implementation.
OTTOMH, I offer the following ideas to protect registration clients without threatening the cashflow of legitimate businesses:
Separate the the first-come-first-serve part of claiming names from the necessarily centralized processing of administrative details. In take-a-number style, use certifiably trustable general purpose servers that archive and return UTC timestamped and digitally signed copies of whatever text you send them (via secure form). In this case the text would say something like, "I, Some Nameclaimer, rest of id, hereby claim the domain name Nameclaimer.com." The returned and archived certificate would establish your claim priority, which you would have 3 business days to exercise.
Eliminate the power of an individual registering business to damage your claim. You have three days, so you attach your timestamped claim certificate with your other registration details and send it to your first choice registration business (most likely within three minutes, but allow for contingencies). The earliest time stamp would have priority in any contest. If you didn't get immediate satisfaction (i.e., digitally signed and UTC time stamped evidence of having completed your claim) from one registering business, you would have that three day window to go to another, and none could accidentally or on purpose lose your claim for you.
Make transfer of title/ownership atomic and at the secure option of the owner. No limbo time allowed! This really should go without saying, but apparently it didn't.
Process expirations with due notification, something like health or car insurance.
I suppose if there was a worry about a time stamping server pulling something funny (like being a cybersquat front), you could send an encrypted version (openable by yourself and perhaps some impartial legal entity) for timestamping first to several time stampers, then the clear text version. Assuming that would create legal evidence. IANAL.
Why not exploit new business opportunities instead of fixing the problem?;-/
One wonders if the status of a given name is even maintained in a single transactional database. Or maybe they have defined name claim and release transactions, but not transfer? Can there be race conditions between competing registering businesses?
Also, how can one be sure the very act of checking a name doesn't pass it to a speculator? It's apparently not encrypted, so who is in a position to snoop all those form submissions? Maybe one should be careful not to "check" unless ready to commit immediately. Hm. Are registering companies allowed to sell their server logs? What if they just extract the names being checked?
The challenge is to embrace our success without losing the qualities that got us here. If I'm not meeting that challenge, I want to hear from you just how, with details, when that happens.
Fair enough. Please meditate on the differences between Olympic, Pro, and amateur sports, and what is different/same in the hearts and experiences of players, fans, and promoters.
Maybe a new business opportunity. IRL there is a risk to letting people into your home. Some services advertise that their people are "bonded and insured," meaning (more or less, IANAL) there's money deposited somewhere which can be tapped in case of trouble.
If there were laws to support bonding of visiting software (I mean laws with consequences that can (really, really) NOT be absorbed by the unscrupulous as cost of doing business), then users could choose to lower their risks in a way backed with predictable legal recourse.
Big commercial operations could afford to provide this kind of assurance (assuming they aren't dependent on deception), but there ought to be a way for a small contributor to give assurances too. Open source is great, but I am not sure I have time to inspect all the code myself, especially if you include OS and libraries (;-), so it would be nice to have versions signed by trusted reviewers. Anybody have a list of trusted reviewers? Should they be bonded ?? Paid?
Get in front of a digital video camera and do your crash course. Convert and post the result to a heavy duty server as quicktime or realaudio, etc., then announce it on slashdot and see what happens.
An Apple G4 with lots of memory and Final Cut Pro, plus a Canon XL with Firewire link thereto should be possible for less than USD8k. That and some talent should do it. Apply to Wellcome for funds?
The way I read what you quoted is that US state judges are bound to observe federal laws and treaties, no matter what their state laws may say.
They should not only ask, they should show and ask
on
Another Software Spy
·
· Score: 1
One way would be to ask permission to preload a text/plain e-mail body with specified info into your mail program, and then to ask you to read and send it. That would be ok with me. Then I would know what I was sending. A further improvement would be an authenticated PGP key for the destination, along with a digitally signed declaration of info use policy at the top of the preloaded message.
If there were a final public challenge mechanism available, we wouldn't get such obvious stuff patented. All they would have to do is set up a web site where patent applications in final form were published for comment, like slashdot articles.
A parallel distributed search for prior art via public review is going to be a lot more effective than one or two examiners going through their search, IMHO.
Seems like a better way (distributed parallel search vs. pat. offc. staff) to show prior art and/or obviousness. And since patents are supposed to be for the benefit of the public,...?
Um, has the process of generating patentable inventions by AI or genetic algorithms been patented yet?
Slashdot Mirror
That seems to be their philosophy. Make sure to exploit any possible claim, not make sure that internal R&D funding is sustainable by fair claims and licensing of its results.
I don't think patenting was originally conceived to serve their kind of "prospecting."
100.o
.90....o..o
.80.+........o
.70....+..+..+..x..+..+..+..+..+..+..+..+.
.60................o
.50..................o
.40....................o
.30.....................o
.20......................o
.10.......................o
..0........................o
....0..3..6..9.12.15.18.21.24.27.30.33.36
Or a scatter gram of first failures vs time, color coded with the kind of failure (disk, memory, mobo stuff, etc.) And for those systems capable of reporting soft failures, those would be interesting too.
Also a scatter gram of soft failures vs time backwards from same-item hard failure.
I wish Consumer Reports would do their stuff with cumulative graphs and scattergrams. They have enough data to show some interesting things.
I wonder how the MPG ratings would plot vs time after tuneup.
BTW, anyone know of a system with ECC memory that detects and reports soft error statistics? How do I know my ECC memory hasn't just been covering up a stuck single-bit error since the beginning?
If things are as bad as comments here on slashdot indicate, perhaps a petition for a proper review sent to the top of the regulatory pyramid and/or congressperson(s) on relevant committes would result in a better implementation.
OTTOMH, I offer the following ideas to protect registration clients without threatening the cashflow of legitimate businesses:
I suppose if there was a worry about a time stamping server pulling something funny (like being a cybersquat front), you could send an encrypted version (openable by yourself and perhaps some impartial legal entity) for timestamping first to several time stampers, then the clear text version. Assuming that would create legal evidence. IANAL.
One wonders if the status of a given name is even maintained in a single transactional database. Or maybe they have defined name claim and release transactions, but not transfer? Can there be race conditions between competing registering businesses?
Also, how can one be sure the very act of checking a name doesn't pass it to a speculator? It's apparently not encrypted, so who is in a position to snoop all those form submissions? Maybe one should be careful not to "check" unless ready to commit immediately. Hm. Are registering companies allowed to sell their server logs? What if they just extract the names being checked?
Good luck, and beware of the invisible hand ;-)
If there were laws to support bonding of visiting software (I mean laws with consequences that can (really, really) NOT be absorbed by the unscrupulous as cost of doing business), then users could choose to lower their risks in a way backed with predictable legal recourse.
Big commercial operations could afford to provide this kind of assurance (assuming they aren't dependent on deception), but there ought to be a way for a small contributor to give assurances too. Open source is great, but I am not sure I have time to inspect all the code myself, especially if you include OS and libraries (;-), so it would be nice to have versions signed by trusted reviewers. Anybody have a list of trusted reviewers? Should they be bonded ?? Paid?
An Apple G4 with lots of memory and Final Cut Pro, plus a Canon XL with Firewire link thereto should be possible for less than USD8k. That and some talent should do it. Apply to Wellcome for funds?
The way I read what you quoted is that US state judges are bound to observe federal laws and treaties, no matter what their state laws may say.
One way would be to ask permission to preload a text/plain e-mail body with specified info into your mail program, and then to ask you to read and send it. That would be ok with me. Then I would know what I was sending. A further improvement would be an authenticated PGP key for the destination, along with a digitally signed declaration of info use policy at the top of the preloaded message.
A parallel distributed search for prior art via public review is going to be a lot more effective than one or two examiners going through their search, IMHO.
Um, has the process of generating patentable inventions by AI or genetic algorithms been patented yet?
;-/
That was my post. Wonder why it came out AC.