spamers will be authed which is pointless because its easy to go register 100 new domains with keys (once you figure out how to do it) but it will be a real pain for startups and small businesses
But spammers won't be able to buy reliable accreditation or maintain a good reputation, no matter how many domains they buy. Maybe buying 100's of domains might help them this month, but eventually it won't help them much at all.
Suppose your mailserver gets a piece of email that's fully authenticated. Before accepting the message for delivery, it can ask a reputation service, (or domain name block list, using today's limited version of such), whether this email is from a reputable place, or whether it's from such a spammy or untrustable place that we should reject it, or whether it's from a "new" domain without any real info and thus should greylist it, temporarily rejecting the message for a few hours.
Your mailserver will then use the answer to help decide how to handle the email.
Presumably, a market will spring up to help these new domains and businesses that don't want to put up with many-hours of delays on their mail getting delivered, namely; accreditation agencies.
An accreditation agency can put its own reputation behind new domains. Companies can go to this accreditation agency, sign some forms, pay a fee, pay a bond, and become accredited.
Accreditation companies that do a good job will get a good following, with people paying attention to their claims of trustworthyness and non-spaminess. Of course, no one will listen to accreditation services that say good things of domains that turn out to be spammers, so they'll presumably go out of business.
In any event, this "buy 100's of domains" worry isn't something I'm concerned about being a problem in the long term.
I'm under the impression that this should have been done by doing a dns lookup on $REV_IP_QUAD.$USER._at.$DOMAIN.
You can set that sort of record up with an "exists:" mechanism. (Though $USER has to be turned into xtext, to handle special characters, and of course you have to handle both ipv4 and ipv6.)
With 30 candidates for just one of the offices, and over 50 things to vote on in total--well, you simply can't do that with a simple two-checkbox ballot
Ok, then make it a 30-checkboxes ballot. Check one and that's it. Why should having to vote on 50 things make the interpretation of who you voted for more confusing?
That wouldn't be a 30-checkboxes ballot, but more like a 300-checkboxes ballot. (I'm assumming an average of 5 candidates for 20 offices, and 10 candidates for 10 offices, and yes-versus-no for 50 referrendums. This is admittedly extreme, but you have to design to accomodate the worst cases.)
With 300 checkboxes, you simply won't get clear, unambiguous results 100% of the time.
If voters were given one big sheet of paper to write checkmarks on, you'll get some ballots back with random marks on them. I know I'd probably hold the uncapped pen in one hand while checking that I hadn't missed any offices/referrendums I knew of or that were on my cheat-sheet I had brought with me, and I'd be likely to make a random mark somewhere on the ballot. Even if I didn't mess up and write an unnoticed stray mark somewhere, someone else probably will do so.
I live in Georgia, where we've moved to the no-paper-trail electronic voting machines, so it's now not possible for a voter to submit such an ambiguous ballot to the voting machinery. (Of course, it's now possible for the voting machinery to undetectably log a completely different ballot than the one submitted, so that one improvement in the process is more than outweighed by the fact that Georgia election results can no longer be trusted, but that's a different story.)
But before we switched over, most counties used a punchcard like system, similar to Florida's. After I punched my entries, I'd always pull out my punchcard from the older punch-machine, and run through the entire list again, making sure that the punched holes matched up with the numbers on the printed ballot, and that they were fully punched out. (It wasn't until after the Florida debacle that I found out that not everyone does this.) Everyone is now aware of the different sorts of errors this system can introduce.
So the fact of the matter is that neither make-a-mark-in-boxes, or punch-holes-in-boxes is going to be 100% foolproof. So saying "just make checkboxes" isn't a real solution.
I think that about the only way you can mostly get rid of these technical problems is to have a paper ballot (possibly automatically printed with electronic voting machinery), that is validated when the voter submits it to a central box.
But if you don't have such a system, you will get ambiguous ballots, and that means that you will need a rule to deal with them. If you throw them out, then in a sense you're disenfanchising voters. If you go by an "intent of the voter", you wind up with a subjective mess.
At least if we ever get voter-verified, machine-validated paper trails added into the electronic voting machines that are the current mania, we could reduce this sort of annoying error down to almost nothing.
Why can't you just have a ballot with 2 squares "Check here if you want Bush for president" and "Check here if you want Kerry for president"? I dare you to misinterpret that.
There are more than just two candidates.
In Florida, 10 candidates were on the ballot as running for US President. In some other states, that list can get up to 30 or so candidates.
There's also more than one item on the ballot. Imagine that there are 20-30 or so offices on the ballot, and another 20-30 or so resolutions or ammendments to vote on.
With 30 candidates for just one of the offices, and over 50 things to vote on in total--well, you simply can't do that with a simple two-checkbox ballot.
Heh, so when a spammer has a SPF record that states the IP sending the spam (some Chinese proxy) is valid, what will that get us? Proof that they really are sending it from China?
It means that we can look up the domain in a reputation/blacklist server, and reject the message if it's from a known spammer or spamming domain.
Or we can accept the message if it had been from a reputable company or domain.
Or we can greylist the message if it's from a new or unproven domain, delaying message acceptabnce by a few hours, (by that time spamming domains may have gotten into the blocklists.)
The end result is that new domains that haven't proved themselves yet will have to deal with some greylisting-related delays while they build up their reputations.
Eventually accreditation providers will probably spring up, vouching for new domains that haven't yet built up a reputation, but who have met the accreditation provider's requirements and have posted some sort of bond. If you trust this accreditation provider, you can use the reputation whitelist he maintains and avoid greylisting his bonded customers.
If you don't, then you can just still greylist domains you don't yet trust, but who haven't (yet?) proved themselves to be spammy.
In the end, spammers using SPF is just a great result. It would be like burglers leaving copies of their drivers licenses, or in-person scammers giving you copies of their drivers license, so that you could more easily query for reports from their previous, displeased victims.
(As a side note, one long-term advantage would be the likely transition from IP-based to domain-based block lists, such that email providers need only protect against forgeries, and cross-customer forgeries, and not be sooo overburdened about possible is-it-or-isn't-it a spamming customer. I personally think it would be fantastic if email service providers could simply provide good technical service, not have to spend so much effort and resources in protecting against possible spammy customers, and yet their spammy customers simply get blocked by everyone anyway without their non-spammy customers getting blocked. I'm hoping that SPF and SenderID checks can eventually make that possible.)
Obviously there are evolutionary advantages in being able to avoid accidents and disease, but when they do occur, aren't there also evolutionary advantages in being able to somehow avoid them killing you, (or causing your offspring or kin to die or suffer)?
I think you're focusing purely on the former, and completely ignoring the latter.
In dogs, a broken leg makes them walk on three legs. This is compensation, not evolution toward bipedal posture. The broken-legged puppy is LESS likely to survive and reproduce (its weaker bones mayhap?).
But a broken-legged puppy that can still manage to walk (on three legs, or however) is MORE likely to survive than a broken-legged puppy that can't manage to walk at all.
In monkeys, a broken or weak arm (eg. from illness) makes them prefer to walk on two legs, but again the arm problem makes them LESS likely to survive
But a broken-arm monkey that can still manage to walk (on two legs, or however) is MORE likely to survive than a broken-arm monkey that can't manage to walk at all.
The biggest problem with all this is that we get no notification that something was blocked. I find out later when someone asks why I havent responded.
I guess my question is--does the sender of the mail get notified? It sounds like that's not happening.
If someone tries to send you mail, and your mailserver accepts the message but just drops it on the floor without telling anyone, then that's not a very useful mailserver.
It's like placing an order for pizza delivery, and then a few minutes later after you're off the phone with them, the pizza place cancels your order because they don't approve of your pizza topping combination--and they don't call you back to tell you.
If your mailserver would just do its tests during the SMTP transaction, then it could REJECT the incoming email then and there and the senders would be notified. (The server needn't accept-then-bounce, which would risk bouncing messages to forged addresses if the mailserver doesn't do SPF tests. But it also shouldn't accept-then-maybe-delete.)
Now after it's delivered to you, *you* might decide to not read the mail, perpahps just leaving it unread in a spam folder you control and deleting it automatically. Your sender won't find out about that either, but at least it's something under your control.
But if your mailserver accepts the mail and then sometimes just deletes it--that's awful.
Further, no email program now would recognize the new header - it wouldn't be displayed, categorized, etc.
You don't need a new header for this purpose. There's already one out there that's very well-defined: "Keywords:".
A requirement of a keyword of "ftc.gov:SEXUALLY-EXPLICIT" wouldn't pollute the Subject line and it would be easily filterable by those who wanted to so filter it. Even if every government in the world required a thousand keywords for various types of messages, you could include enough Keywords: headers to account for every one of them.
But if even if merely 50 governments required one keyword each for this type "sexually-explicit" type of content, the subject line would becomes unusable if the keyword had to be put in the Subject, though again not if the keywords were all put into Keywords: body headers.
Furthermore, with specific keywords being required, folks who are worried about over-zeleous prosecution can set their MTA's to simply add that keyword to the Keywords: headers of *all* messages sent through it. That way they'd not have to worry about being in violation of the rule because they or their customers refered to something sexually explicit such as, say, women with bare arms or exposed faces.
(Yes, I have put the project on the back burner for a bit, but there are certain important parts I've had to work on all this time--it's not in any sense of the word abandonded,)
I want to have provisions to allow people to exit out of the license in some ways within some initial startup time period. Adding in a critical-mass of patents before that time period starts ticking is a good idea. Thanks!
Of course I can't really ask people to commit to submitting patents until the license is finalized though, and the license is still in its beginning stages. But I had been attempting to do something else, similar to what you suggest: In talking with people about the notion of so licensing their patents, I tried to persuade interested parties to allow their interest to be public knowledge, once a certain number of companies so similarly agreed, and that number reached some critical mass.
The closed-source companies I talked with were not so interested in their interest being known, publicly or privately, not until many other closed-source companies were known to be interested--so stalemate. (I was aiming to be able to publicize a set of closed and open source companies interested in such a license--not just one or the other, and concentrated on the "harder" side of things first. Perhaps I was wrong in that and should just talk with the Open Source folks first.)
(In a breath of fresh air, on asking him about this, Raph Levien quite nonchalantly said I could publicize his interest in considering something like the Open Patent License once it became solid. I'm guessing the more Open Source companies may respond similarly--again, I was probably wrong in my thoughts about the order of doing things, and should possibly go ahead and contact them.)
In any event, the license still needs a lot of work. If you're willing to put some time in helping there, that would be great! And if you're a lawyer, that's even better!:-)
We can still get people to publicize their interest independently of the license state, but it should go considerably faster as the license becomes more solid.
If you're emailing all 600 people in one batch, that's idiotic and email each person individually.
I don't see how it's idiotic--assumming mailservers are blocking mail merely because they consider emails with more than a handful of recipients to mean the mail is spam, I can't see how pandering to such silly behavior is in the long-term best interest of anyone.
In fact, RFC2128 even discourages such behavior:
recipients buffer
The minimum total number of recipients that must be buffered is 100 recipients. Rejection of messages (for excessive recipients) with fewer than 100 RCPT commands is a violation of this specification. The general principle that relaying SMTP servers MUST NOT, and delivery SMTP servers SHOULD NOT, perform validation tests on message headers suggests that rejecting a message based on the total number of recipients shown in header fields is to be discouraged. A server which imposes a limit on the number of recipients MUST behave in an orderly fashion, such as to reject additional addresses over its limit rather than silently discarding addresses previously accepted. A client that needs to deliver a message containing over 100 RCPT commands SHOULD be prepared to transmit in 100-recipient "chunks" if the server declines to accept more than 100 recipients in a single message.
I would worry that if I tried to work around broken behavior of other ISPs, that I'd be continually running around in circles implementing workaround on top of workaround and causing even more problems. (I might want to alert my customers entering in hotmail or aol addresses that they will likely not receive my emails, suggesting they use another addresss if possible.)
I don't think you can't force the give-back requirement as much as you seem to me to want. At least, I don't think you can do that in a patent license as far as it applies to all Open Source code. (However, I think you'll end up getting much the same effect indirectly anyway, so I don't worry about the fact that I'm trying to do option (1) below instead of option (2).)
The reason your options are limited is that you have to work within the bounds of the Open Source definition and at least some Open Source licenses.
For instance:
From the Open Source Definition, item 7:
7. Distribution of License
The rights attached to the program must apply to all to whom the program is redistributed without the need for execution of an additional license by those parties.
Rationale: This clause is intended to forbid closing up software by indirect means such as requiring a non-disclosure agreement.
There may already be a conflict between having the OPL grant patent rights for use in Open Source software and OSD7 saying you can't be dependent on another license--I don't really know. However, I'm not really overly worried about it, as I imagine that if such a conflict exists that the OSD board would probably be willing to look at changing OSD7 enough to accomodate such patent grants. (This is an assumption--I haven't asked.)
However, I don't think they'd be willing to modify it much more than what would be necessary to attain our mutual goals. Things like implicitely putting into the OSD requirements for specific patent poision-pills or joining specific interlicense patent pools is a bit much to ask, to put it mildly! (I think this is effectively what you're getting at--not sure though.)
Other Open Sorce Licenses:
While the OPL can be written to exclude Open Source licenses with particularly onerous patent requirements, there will still be requirements that remain which prevent the sort of thing I think you're thinking about.
For instance, if you're a patent owner and via the OPL submit a patent such that it can be used in all Open Source code, after which someone writes and GPLs code that uses that patent and 10 others, and those other 10 patents have been (with permission) used in GPL code before, then which result are you looking for?
All 11 patents can be used in GPL code. The one patent can also be used in code licensed under some subset of Open Source licenses, that subset described in the OPL. (And the subset is bigger than just the GPL.:-) )
All 11 patents can be used in code licensed under the same subset of Open Source license listed above.
(1) is what I'm trying to have the OPL do. (2) is what I think you're suggesting, and I don't see how it would be practical. (You'd have to either contact the owners of all 10 patents to get them to explicitely relicense their patents, or you'd have to have already convinced the FSF to have changed the GPL accordingly and similarly, before the owners allowed their patents to be used in GPLv? products. You would have to do much the same thing for most any other Open Source license as well. Very difficult, if not impossible. (1) IMHO is more likely to be much more practical.)
Please let me know if I'm totally misunderstanding you, as is probably likely the case.
But what I think you mean by forcing aggregation is more than what can be done. (I'm more assertive with closed-source code, by the way, which I think makes sense and is the more desirable and practical way to go about things for all concerned.)
(And as I plan to get back to work on all this in the next month or so, I welcome any comments, suggestions, and diffs.:-) )
I dunno, I kinda liked having cars talk to me:). I think the problem arises when people start talking back and expecting another answer.
I liked it too, as did everyone I knew who heard our family's car talk.
And, well, we also all talked back to it too: "Door is ajar? The door is not a jar; the door is a door!":-)
(Amazingly, no one ever tired of that joke. But we were all disappointed that cars stopped coming out with speech as an option.)
As for me, I was wanting future versions to let you easily upload your own sound samples for every error condition! Sort of like ringtones for your car.
"Are you really going to leave me here with my lights on again? You remember all the trouble that caused last time, don't you?"
Or maybe a Britney Spears-like: "Oops, you did it again...." might be more popular.
Or maybe for service-required warnings for the Hitchhiker's Guide fans: "I've got this terrible pain in all the diodes down my left hand side."
You quote the article's description of one power supposedly granted to theaters with this Ohio law when the theater suspects people of videotaping movies, point out how taking pictures of friends is hardly videotaping a movie, then conclude that this doesn't affect anyone.
Here are some mismatches in your logic:
The one power quoted is not the whole of the new law.
It's not just the theater who can do this. Police can do it too, and unfortunately police are often used for political reasons. I think of this law as just another potential method for someone in power to use to cause hardship for someone who might be embarassing.
You don't have to have "videotaped a movie" to be "supected of videotaping a movie", but now it's far easier to be so suspected. When they find out later that I haven't done what they think I've done, will they recompensate me for my time?
Why should videotaping parts of the movie not be considered fair use and be considered *very* socially acceptable in the first place? (Meaning restrictions on such activity would affect people.)
You may have missed the part in the article where an EFF lawyer wonders whether he'd be committing a felony to take a 5-second clip in the theatre to mail to some friends to point out how bad the movie is. To frame the issue in terms of your reply: Would any "theatre owner/security guard in their right mind" bother someone doing this? Would they bother if he did this with something that looked more like a traditional camcorder?
In the examples I gave, people were in the theatre taking picutures and even videotaping parts of the director question/answers--even a few times when someone involved in making the film went up to the screen to point at something of interest *during the film*. (So now when a director goes up to the screen during his own film to point out a hilarious/embarassing error, the people videotaping that, (probably the people paying for his trip there) could potentially be arrested for it. Is that not insane?)
Two words: "Rocky Horror" Who would not think it would be cool to have videos of their friends in a theatre doing the time warp in sync with the move? Something to show their grandkids, heh.:-)
The biggest common effect I would imagine though, would be if searches on entering/leaving theatres became common--I can only hope that people wouldn't put up with such silly extra lines and just leave.
In any even, having to go through police checkpoints on entering a movie is definitely something that would affect people.
Unrelated to your reply but:
As far as the bit about guards walking through the theatre wearing night-vision goggles: I'm sorry, but the first thing I thought of on reading that was how hilarious it would be to get one of those night-mode camcorders and record the guards walking around the theater, then sell that recording to a local news station.
I'm guessing the camcorder would pick up the IR illumination the night-vision goggles use, and the news station could show "how your ticket money at thus-and-such theatre goes to pay for the guards to spy on you and your date watching a movie," and how the station is of course only doing the same thing themselves to expose the whole thing. (Who guards the guards, as it were.)
So if the California law were passed years ago, would it have made "Last Action Hero" illegal? Remember the film shows a person watching a movie in a theatre. Come to think of it, any movie now made where at some point the characters are themselves watching a movie in a theatre could run afowl of the law, unless the MPAA exempted themselves from their own law, of course.
Ok so it's now punishable by law to record in a theater, does this affect you?
Yes, it affects me.
Here are a couple examples:
Others have already given examples of common everyday folks "recording" things in a theatre--I mean, is it not absolutely *insane* to say that if you and 20 or so friends go see a movie as part of a group social outing, that *no one* will bring any sort of camera?
I would think by now that it would be considered perfectly normal for weblogs and the like to include pictures of friends in situations like that!
If the theatre owner sees that
Your group is buying lots of snacks and
exudes such fun and exitement in simply getting together that it rubs off on the other guests and makes them enjoy their time there more,
and gives you a couple dozen free passes so as to get you back more often,
would you think that he'd be likely to want to confiscate the cameras, cell-phones, camcorders, laptops, etc, of the group and make them want to leave?
Do you think he'd be likely to be in favor of laws that not only require he confiscate such items, but also require that he have these customers arrested?
I'm thinking the answer to both questions is "no".
Not being able to record such social outings would make them comparatively boring.
Fortunately, it's legal in my state to go to the theatre with a bunch of friends and have everyone enjoy the experience. It's too bad for the residents of California and Ohio that they're legally restrained from having as much fun.
A local film society group regularly rents out theatre space to show independent/little-known films.
At the beginning of the show, a few people representing the group get up and talk about the film, random historical tidbits related to it, etc., the story about how they heard about it and were able to "get in line" to have the film and rights to show it for a few days, and so forth, plus a bit of a status update on films they're trying to get.
During the yearly film festivals, producers, directors, actors, and others involved in the making of the films are often there as well. (Realize these are not-very-well-known independents, and for most of them it's still a novelty for them to talk about their work in front of an interested audience.)
Anyway, they're brought up on stage before and after the films, relating their viewpoint on things, answering questions, and basically doing a live and not-during-the-film version of a dvd's "directors soundtrack."
Naturally, multiple people take pictures and record all this, (including and especially the folks involved in making the films--or at least they'll ask for copies later.)
It can be a great social event, very interactive, and fun for everyone.
But if you can't legally record any of it, then it won't be as much fun: The film society can't have pictures of their events in their newsletter or website, the people involved in the making of the films won't have hardcopy records of their memories, and the general mood of the event will be diminished given the shortsighted legal restrictions.
Fortunately I live in a state that has not passed such a law.
If a director of a small film is invited to a film festival in my state, and he has to decide between going to this film festival versus a similar event in California or Ohio, the fact that we don't legally limit him having fun like that means he's more likely to come to our event. Woohoo!
I think it's quite magnanimous of the residents of California and Ohio to
have their elected representitives limit their theatre experiences so that I might enjoy mine all the more--all without me having to vote for these representatives or contribute a dime to their compaigns!
Make free space map efficiently reuse empty index pages, and other free space management improvements
In previous releases, B-tree index pages that were left empty because of deleted rows could only be reused by rows with index values similar to the rows originally indexed on that page. In 7.4, "VACUUM" records empty index pages and allows them to be reused for any future index rows.
The following is an example from running VACUUM on a table in the regression database:
regression=# VACUUM VERBOSE ANALYZE onek;
INFO: vacuuming "public.onek" INFO: index "onek_unique1" now contains 1000 tuples in 14 pages
DETAIL: 3000 index tuples were removed.
0 index pages have been deleted, 0 are currently reusable.
CPU 0.01s/0.08u sec elapsed 0.18 sec.
[...]
These improvements seem to address your complaint.
We have tables with 90% daily turnover, and without daily reindexes, those indexes reach gigs if we don't catch them. Regardless of any improvements, I will continue to suggest against Postgresql until that little problem is fixed.
Given that you wouldn't be needing to run the reindex or reindexdb commands on a regular basis anymore, I guess you might be changing your suggestions.
I've use punch card voting systems, and the problem with them is that you get almost no feedback after you've punched the card. The card itself is hidden under a template full of little holes (which in the case of a butterfly ballot, don't quite line up with the names off to the sides), and it's hard to see down into those holes to tell if you actually punched the hole. There isn't much mechanical feedback, either. Once you pull the card out of the template, it looks like just a bunch of random holes.
That may be the case for the punchcard systems you've used, but that doesn't have to always be the case. In the system I'm familiar with, each potential ballot choice is not only labelled with the person running for the office, but also a number. The punch card itself is also labelled/printed with the same numbers next to its perforated holes.
(And after punching out all my votes I would of course always pull out the card and use those numbers to check that the punched-out holes & numbers matched up with the numbers of the people/referrendums I was voting for.)
So I'm amazed that anyone would have ever made a punchcard system with *blank* cards! (Other than registration marks I guess.) How do the elections people expect voters to verify their votes after they pull the cards out, I wonder?
Anyway, I live in Georgia, and this whole fairly sensible way of voting was thrown out recently, to be replaced with the touchscreen nonsense everyone's read about--which of course is something far worse!
Interesting side note: In the election before the punchards were replaced, there were demo units of the touchscreen system in each polling place, along with a person there showing off the new glitzy machines we'd all be using the next election.
From the demo-person's reaction to my questions/complaints, I was apparently the only person all day who had been less that wildly enthusiastic about the soon-to-be "upgrade" of the voting equipment! And even more shocking, even though the purpose of the demo was supposedly to determine voter reaction to the coming change, the only thing they were actually set up to do is inform the electorate of what changes would be imposed--there was no process and no way to submit comments or questions back upstream. I thought that was a pretty unusual way of doing a market survey!
I still see unstated assumptions in this version of the "we need patents for drugs" argument.
Let me snip your argument down a bit, (mainly removing supporting data that isn't really in dispute):
The reason is that our health care industry needs drugs to be patented is because the cost of developing drugs is so high.
[...]
By far the biggest cost of developing the drug is the R&D.
These are costs that post-patent generic companies simply don't have to pay.
Without patent protection, we'd end up with a world where the developer of a drug would incur *all* of the costs to develop that drug and the generic manufacturer would gain all of the profits. The drug developer would simply not be able to recover their costs without patents.
Patents enable the drug developer to earn the profits from the sale of that drug to cover:
[...]
Inventing is a hard business, and without some sort of legal protection, we simply wouldn't have it.
I think you might be confusing the likely failure of a particular unchanged business model under a changed regulatory environment with a claim of failure of *any* business model under a changed regulatory environment.
Let me know if I'm misrepresenting your argument, but it basically seems to boil down to something like the following:
Drug development costs big money.
Drug development costs are paid by companies specializing in drug development.
Drug development companies recoup these costs through sales of the resultant drugs.
The drug development companies sell their drugs at a price much higher than the comparatively small cost of manufacture.
Sales of their drugs would be undercut by competitors if just anyone were could make cheaper generic versions of these drugs and sell them.
Patents keep these would-be competitors from making and selling cheap generic versions of these drugs (for a time).
This temporary monopoly allows for the drug development companies to recoup their costs by selling at a price higher than would be economical if they didn't have a monopoly on the drug production/sale.
Without patents, this particular business model would be unsustainable.
I am unable to think of any other viable business model, so I conclude that it is absolutely clear that we need a strong patent system to make sure drug development continues.
Okay, maybe the wording for the first part of #9 above was a bit strong.:-)
However, you still concluded that there can be no viable business model for drug development in a no-patent world.
Below I give the same line of reasoning for a different field. After examining that argument, I wonder if you would also conclude that there is no viable model for development in that field under less monopolistic conditions.
The arguments are the same--or do you consider one argument valid and the other invalid?
(NOTE: I do not claim here that a business model that may fit the case above will necessarily have an analagous business model that will fit the case below, or vice versa; I merely claim that both lines of reasoning that conclude with an impossibility of the existence of a viable business model are flawed.)
So here is the second, similar, and IMHO similarly flawed line of reasoning:
Software development costs big money.
Software development costs are paid by companies specializing in software development.
The software development companies recoup these costs through sales of their software.
The software development companies sell their software at a price much higher than the comparatively small cost of manufacture.
Sales of their software would be undercut by competitors if just anyone could make cheaper copies of their software and sell those copies.
Say what you want about drug companies, but IMHO, it is absolutely clear that we would not have the breadth and depth of health care if we didn't have patents protecting drug companies.
It may be clear to you, but it isn't clear to me.
Can you explain your line of reasoning in more detail?
You make a short, extremely strong claim: "Lack of patent protection would KILL the pharm industry."
You also make a longer version of the same claim:
Go after the analysis, not just polio - I listed other ailments as well. I imagine you aren't familiar with the pharm industry, but it wouldn't exist without patents. When you invest $1B in a drug, and a generic competitor can then start making it for under a $1M, NO R&D would take place. You MIGHT get cures for the most mass-life-threatening diseases, but the idea that universities could carry on the R&D that is currently done in the private sector is absolutely laughable. I'm a chemist at a top-5 research school, so you're not trumping on this one. Lack of patent protection would KILL the pharm industry.
People often make this sort of claim, but I don't think I've ever seen it actually backed up.
Can you explain why it would be utterly point-blank impossible for old and new cures to be available and to be continually developed over the long-term in a reasonable fashion (that is, not at a snail's pace for the entire medical field), if we were living in a world without patents? Two notes:
I'm asking you to prove something slightly different from what you actually claimed above, though I imagine it's basically similar to what you mean to claim.
I admit that this may not be a fair question, in that I am in effect asking you to prove a negative.
In any event, I don't think I've ever seen anyone actually go through, step-by-step, and prove why medical progress would necessarily slow to a snails pace were we suddenly in a no-patent world. Instead it's more or less claimed to be obvious without too much real backing to the argument. Generally the arguments as I've seen them have all sorts of subtle assumptions. For instance:
When you invest $1B in a drug, and a generic competitor can then start making it for under a $1M, NO R&D would take place.
There are all sorts of assumptions in the above statement. Here are a couple:
There is an assumption that the group spending the $1B has an interest in selling the drug. Why does that have to be the case?
There is an assumption that another group that could make and sell this new drug for just $1M is necessarily a (feared or potential) "competitor" that would eat away from the first group's business. Why would this have to be the case? Why would it, for instance, be utterly impossible for them to be ecstatic that someone else can make the drug for $1M, (even assuming no royalties go to the first group), perhaps even talking with this so-called "competitor" to find ways of making production cheaper for them, or sponsoring some of the costs?
Perhaps those assumptions would be considered to be valid for businesses operating in world with pharmaceutical patents, but I don't see how they must necessarily be indisputably valid for for-profit businesses operating in a world without patents.
It is far from clear to me that the fact that the box which has been transported to my facility is rented means that the software is not distributed. I don't see why the nature of the financial transaction is important to the concept of distribution.
And from a previous comment:
If you rent me a piece of software at my site or by placing it under my control, you're distributing the software and GPL 2 applies just fine.
Okay, then let me present a hypothetical example.
Suppose I make a soft drink vending machine, that I refill myself, and pay you for the privilege of putting outside (or maybe even inside) your place of business. Let's say I pay you five cents per purchased drink.
The vending machine runs linux internally, and it's connected to a network, and I use that connection to make sure it's optimally filled up with drinks. (I come around and fill it up when things run low.)
With this situation, GPL'd software is in a machine at your facility.
Would you say I am obligated via the GPL to provide you with a copy of the source of all the GPL'd programs in the vending machine? Would I be obligated to provide a copy to all the vending machine's customers?
What if I allowed you some access to the machine. Say I allowed you to ssh in and run a single command that displayed logfiles of when people bought different drinks. Would that change anything in the above answers?
Or, remembering your comment that the nature of the financial transaction doesn't matter, if I rented you the machine instead, so that you were checking it's current load and refilling it up regularly, (ie, using the software and "renting" the machine and software, as opposed to neither renting nor owning as in the cases above). Would that change anything?
I believe Debian is still the distro of choice, but this is simply a second Linux offering. Previous to now HP would not provide support for Red Hat on its systems.
Not entirely true. Not all hardware was supported, but that's a different issue from whether any OS or distribution was.
I used to work in the Response Center at HP. I took linux support calls right along with HP-UX calls; we sold support contracts for various linux distributions as far back as 1999.
And I'm not trying to flame or be rude with this comment... bear with me... but why do you need ethernet surge protection? In my situation I have 2 lines run out to the garage (detached) so I can play drums/guitar etc, but they're buried deep (6 feet) and grounded....
I've always wondered if surge protection would help....
It's for a bunch of different machines set up at client locations. Your ethernet cables might be burried 6 feet deep, but theirs could be wrapped around a lightning rod somewhere for all I know.
BTW the need for both phone and ethernet surge protection is because I'll be using phone lines as a backup for data syncs, if my client's network is down, or if they don't have a network.
The dual surge protection requirement is one of the less serious requirements--you can always buy an additional phone line/ethernet surge protection wall-wart type of thing to cover for one or the other not being there--but it's kind of ugly.
However, the big thing for me is to make sure the UPS behavior actually really does work correctly. For instance, if the UPS looses power and goes battery-low, and the system shuts down but the power comes back on during the shutdown process, the UPS had better still cut power for a short while, so the machines have an actual powerloss and then poweron to start them up again. Not every ups and software combination does this correctly all the time.
To learn more about the recall action and the process for replacing the affected units, users should visit www.apc.com or call 866 APC-RELY (866 272-7359).
You could try their ironically spelled phone number.
You mean perhaps the ironic spelling of 866-ARCS-FLY?:-)
As an aside, does anyone know of any very small (in size, va rating, and cost) ups's that have both serial and usb connections, spacing for 2 transformers plus 1 non-transformer plug, avr of some sort, separate phone line and ethernet surge protection built in, (that is, you can plug in a phone line and an ethernet connection at the same time), have some sort of smart signaling ("line power down and battery low--you better shut down now"; "okay, give me 2 minutes, then shut off my power whether or not wall power has come up, and keep me shut down until you've had reliable power for a while"), *and* that are properly documented by the manufacturer and work with some sort of open source ups software under linux?
I'm finding it hard to find any ups's that meet even half of those criterion, much less all of them.
Granted, I was very late (1.5 hours) to the Atlanta meetup, but no one else was there when I did finally arrive.
I had expected that this sort of event would follow other geek meeting-at-a-place-with-food traditions and last well into the night, but not only were there not any people still there with the slashdot meetup crowd, the employees at the coffee shop the meetup was scheduled to occur at didn't remember any large group or any "slashdot" group showing up earlier.
Given that there were at least 65 people rsvp'd, from what I remember, (now that the meeting start time has passed, slashdot.meetup.com doesn't allow you to see who all rsvp'd or read their comments), I would have though that at least 20-30 people would still be there.
I'm wondering if this might have been a hoax--Did other folks have similar experiences in other cities?
Slashdot Mirror
Suppose your mailserver gets a piece of email that's fully authenticated. Before accepting the message for delivery, it can ask a reputation service, (or domain name block list, using today's limited version of such), whether this email is from a reputable place, or whether it's from such a spammy or untrustable place that we should reject it, or whether it's from a "new" domain without any real info and thus should greylist it, temporarily rejecting the message for a few hours.
Your mailserver will then use the answer to help decide how to handle the email.
Presumably, a market will spring up to help these new domains and businesses that don't want to put up with many-hours of delays on their mail getting delivered, namely; accreditation agencies.
An accreditation agency can put its own reputation behind new domains. Companies can go to this accreditation agency, sign some forms, pay a fee, pay a bond, and become accredited.
Accreditation companies that do a good job will get a good following, with people paying attention to their claims of trustworthyness and non-spaminess. Of course, no one will listen to accreditation services that say good things of domains that turn out to be spammers, so they'll presumably go out of business.
In any event, this "buy 100's of domains" worry isn't something I'm concerned about being a problem in the long term.
You can set that sort of record up with an "exists:" mechanism. (Though $USER has to be turned into xtext, to handle special characters, and of course you have to handle both ipv4 and ipv6.)With 300 checkboxes, you simply won't get clear, unambiguous results 100% of the time.
If voters were given one big sheet of paper to write checkmarks on, you'll get some ballots back with random marks on them. I know I'd probably hold the uncapped pen in one hand while checking that I hadn't missed any offices/referrendums I knew of or that were on my cheat-sheet I had brought with me, and I'd be likely to make a random mark somewhere on the ballot. Even if I didn't mess up and write an unnoticed stray mark somewhere, someone else probably will do so.
I live in Georgia, where we've moved to the no-paper-trail electronic voting machines, so it's now not possible for a voter to submit such an ambiguous ballot to the voting machinery. (Of course, it's now possible for the voting machinery to undetectably log a completely different ballot than the one submitted, so that one improvement in the process is more than outweighed by the fact that Georgia election results can no longer be trusted, but that's a different story.)
But before we switched over, most counties used a punchcard like system, similar to Florida's. After I punched my entries, I'd always pull out my punchcard from the older punch-machine, and run through the entire list again, making sure that the punched holes matched up with the numbers on the printed ballot, and that they were fully punched out. (It wasn't until after the Florida debacle that I found out that not everyone does this.) Everyone is now aware of the different sorts of errors this system can introduce.
So the fact of the matter is that neither make-a-mark-in-boxes, or punch-holes-in-boxes is going to be 100% foolproof. So saying "just make checkboxes" isn't a real solution.
I think that about the only way you can mostly get rid of these technical problems is to have a paper ballot (possibly automatically printed with electronic voting machinery), that is validated when the voter submits it to a central box.
But if you don't have such a system, you will get ambiguous ballots, and that means that you will need a rule to deal with them. If you throw them out, then in a sense you're disenfanchising voters. If you go by an "intent of the voter", you wind up with a subjective mess.
At least if we ever get voter-verified, machine-validated paper trails added into the electronic voting machines that are the current mania, we could reduce this sort of annoying error down to almost nothing.
In Florida, 10 candidates were on the ballot as running for US President. In some other states, that list can get up to 30 or so candidates.
There's also more than one item on the ballot. Imagine that there are 20-30 or so offices on the ballot, and another 20-30 or so resolutions or ammendments to vote on.
With 30 candidates for just one of the offices, and over 50 things to vote on in total--well, you simply can't do that with a simple two-checkbox ballot.
Or we can accept the message if it had been from a reputable company or domain.
Or we can greylist the message if it's from a new or unproven domain, delaying message acceptabnce by a few hours, (by that time spamming domains may have gotten into the blocklists.)
The end result is that new domains that haven't proved themselves yet will have to deal with some greylisting-related delays while they build up their reputations.
Eventually accreditation providers will probably spring up, vouching for new domains that haven't yet built up a reputation, but who have met the accreditation provider's requirements and have posted some sort of bond. If you trust this accreditation provider, you can use the reputation whitelist he maintains and avoid greylisting his bonded customers.
If you don't, then you can just still greylist domains you don't yet trust, but who haven't (yet?) proved themselves to be spammy.
In the end, spammers using SPF is just a great result. It would be like burglers leaving copies of their drivers licenses, or in-person scammers giving you copies of their drivers license, so that you could more easily query for reports from their previous, displeased victims.
(As a side note, one long-term advantage would be the likely transition from IP-based to domain-based block lists, such that email providers need only protect against forgeries, and cross-customer forgeries, and not be sooo overburdened about possible is-it-or-isn't-it a spamming customer. I personally think it would be fantastic if email service providers could simply provide good technical service, not have to spend so much effort and resources in protecting against possible spammy customers, and yet their spammy customers simply get blocked by everyone anyway without their non-spammy customers getting blocked. I'm hoping that SPF and SenderID checks can eventually make that possible.)
I think you're focusing purely on the former, and completely ignoring the latter.
But a broken-legged puppy that can still manage to walk (on three legs, or however) is MORE likely to survive than a broken-legged puppy that can't manage to walk at all. But a broken-arm monkey that can still manage to walk (on two legs, or however) is MORE likely to survive than a broken-arm monkey that can't manage to walk at all.If someone tries to send you mail, and your mailserver accepts the message but just drops it on the floor without telling anyone, then that's not a very useful mailserver.
It's like placing an order for pizza delivery, and then a few minutes later after you're off the phone with them, the pizza place cancels your order because they don't approve of your pizza topping combination--and they don't call you back to tell you.
If your mailserver would just do its tests during the SMTP transaction, then it could REJECT the incoming email then and there and the senders would be notified. (The server needn't accept-then-bounce, which would risk bouncing messages to forged addresses if the mailserver doesn't do SPF tests. But it also shouldn't accept-then-maybe-delete.)
Now after it's delivered to you, *you* might decide to not read the mail, perpahps just leaving it unread in a spam folder you control and deleting it automatically. Your sender won't find out about that either, but at least it's something under your control.
But if your mailserver accepts the mail and then sometimes just deletes it--that's awful.
A requirement of a keyword of "ftc.gov:SEXUALLY-EXPLICIT" wouldn't pollute the Subject line and it would be easily filterable by those who wanted to so filter it. Even if every government in the world required a thousand keywords for various types of messages, you could include enough Keywords: headers to account for every one of them.
But if even if merely 50 governments required one keyword each for this type "sexually-explicit" type of content, the subject line would becomes unusable if the keyword had to be put in the Subject, though again not if the keywords were all put into Keywords: body headers.
Furthermore, with specific keywords being required, folks who are worried about over-zeleous prosecution can set their MTA's to simply add that keyword to the Keywords: headers of *all* messages sent through it. That way they'd not have to worry about being in violation of the rule because they or their customers refered to something sexually explicit such as, say, women with bare arms or exposed faces.
(Yes, I have put the project on the back burner for a bit, but there are certain important parts I've had to work on all this time--it's not in any sense of the word abandonded,)
:-)
I want to have provisions to allow people to exit out of the license in some ways within some initial startup time period. Adding in a critical-mass of patents before that time period starts ticking is a good idea. Thanks!
Of course I can't really ask people to commit to submitting patents until the license is finalized though, and the license is still in its beginning stages. But I had been attempting to do something else, similar to what you suggest: In talking with people about the notion of so licensing their patents, I tried to persuade interested parties to allow their interest to be public knowledge, once a certain number of companies so similarly agreed, and that number reached some critical mass.
The closed-source companies I talked with were not so interested in their interest being known, publicly or privately, not until many other closed-source companies were known to be interested--so stalemate. (I was aiming to be able to publicize a set of closed and open source companies interested in such a license--not just one or the other, and concentrated on the "harder" side of things first. Perhaps I was wrong in that and should just talk with the Open Source folks first.)
(In a breath of fresh air, on asking him about this, Raph Levien quite nonchalantly said I could publicize his interest in considering something like the Open Patent License once it became solid. I'm guessing the more Open Source companies may respond similarly--again, I was probably wrong in my thoughts about the order of doing things, and should possibly go ahead and contact them.)
In any event, the license still needs a lot of work. If you're willing to put some time in helping there, that would be great! And if you're a lawyer, that's even better!
We can still get people to publicize their interest independently of the license state, but it should go considerably faster as the license becomes more solid.
Oops, that should be RFC2821, not RFC2128.
In fact, RFC2128 even discourages such behavior:
I would worry that if I tried to work around broken behavior of other ISPs, that I'd be continually running around in circles implementing workaround on top of workaround and causing even more problems. (I might want to alert my customers entering in hotmail or aol addresses that they will likely not receive my emails, suggesting they use another addresss if possible.)The reason your options are limited is that you have to work within the bounds of the Open Source definition and at least some Open Source licenses .
For instance:
- From the Open Source Definition, item 7:
There may already be a conflict between having the OPL grant patent rights for use in Open Source software and OSD7 saying you can't be dependent on another license--I don't really know. However, I'm not really overly worried about it, as I imagine that if such a conflict exists that the OSD board would probably be willing to look at changing OSD7 enough to accomodate such patent grants. (This is an assumption--I haven't asked.)
-
Other Open Sorce Licenses:
-
All 11 patents can be used in GPL code. The one patent can also be used in code licensed under some subset of Open Source licenses, that subset described in the OPL. (And the subset is bigger than just the GPL.
:-) )
- All 11 patents can be used in code licensed under the same subset of Open Source license listed above.
Please let me know if I'm totally misunderstanding you, as is probably likely the case.However, I don't think they'd be willing to modify it much more than what would be necessary to attain our mutual goals. Things like implicitely putting into the OSD requirements for specific patent poision-pills or joining specific interlicense patent pools is a bit much to ask, to put it mildly! (I think this is effectively what you're getting at--not sure though.)
While the OPL can be written to exclude Open Source licenses with particularly onerous patent requirements, there will still be requirements that remain which prevent the sort of thing I think you're thinking about.
For instance, if you're a patent owner and via the OPL submit a patent such that it can be used in all Open Source code, after which someone writes and GPLs code that uses that patent and 10 others, and those other 10 patents have been (with permission) used in GPL code before, then which result are you looking for?
(1) is what I'm trying to have the OPL do. (2) is what I think you're suggesting, and I don't see how it would be practical. (You'd have to either contact the owners of all 10 patents to get them to explicitely relicense their patents, or you'd have to have already convinced the FSF to have changed the GPL accordingly and similarly, before the owners allowed their patents to be used in GPLv? products. You would have to do much the same thing for most any other Open Source license as well. Very difficult, if not impossible. (1) IMHO is more likely to be much more practical.)
But what I think you mean by forcing aggregation is more than what can be done. (I'm more assertive with closed-source code, by the way, which I think makes sense and is the more desirable and practical way to go about things for all concerned.)
(And as I plan to get back to work on all this in the next month or so, I welcome any comments, suggestions, and diffs. :-) )
And, well, we also all talked back to it too: "Door is ajar? The door is not a jar; the door is a door!" :-)
(Amazingly, no one ever tired of that joke. But we were all disappointed that cars stopped coming out with speech as an option.)
As for me, I was wanting future versions to let you easily upload your own sound samples for every error condition! Sort of like ringtones for your car.
"Are you really going to leave me here with my lights on again? You remember all the trouble that caused last time, don't you?"
Or maybe a Britney Spears-like: "Oops, you did it again...." might be more popular.
Or maybe for service-required warnings for the Hitchhiker's Guide fans: "I've got this terrible pain in all the diodes down my left hand side."
You quote the article's description of one power supposedly granted to theaters with this Ohio law when the theater suspects people of videotaping movies, point out how taking pictures of friends is hardly videotaping a movie, then conclude that this doesn't affect anyone. Here are some mismatches in your logic:
- The one power quoted is not the whole of the new law.
- It's not just the theater who can do this. Police can do it too, and unfortunately police are often used for political reasons. I think of this law as just another potential method for someone in power to use to cause hardship for someone who might be embarassing.
- You don't have to have "videotaped a movie" to be "supected of videotaping a movie", but now it's far easier to be so suspected. When they find out later that I haven't done what they think I've done, will they recompensate me for my time?
- Why should videotaping parts of the movie not be considered fair use and be considered *very* socially acceptable in the first place? (Meaning restrictions on such activity would affect people.)
- You may have missed the part in the article where an EFF lawyer wonders whether he'd be committing a felony to take a 5-second clip in the theatre to mail to some friends to point out how bad the movie is. To frame the issue in terms of your reply: Would any "theatre owner/security guard in their right mind" bother someone doing this? Would they bother if he did this with something that looked more like a traditional camcorder?
- In the examples I gave, people were in the theatre taking picutures and even videotaping parts of the director question/answers--even a few times when someone involved in making the film went up to the screen to point at something of interest *during the film*. (So now when a director goes up to the screen during his own film to point out a hilarious/embarassing error, the people videotaping that, (probably the people paying for his trip there) could potentially be arrested for it. Is that not insane?)
- Two words: "Rocky Horror" Who would not think it would be cool to have videos of their friends in a theatre doing the time warp in sync with the move? Something to show their grandkids, heh.
:-)
- The biggest common effect I would imagine though, would be if searches on entering/leaving theatres became common--I can only hope that people wouldn't put up with such silly extra lines and just leave.
Unrelated to your reply but:In any even, having to go through police checkpoints on entering a movie is definitely something that would affect people.
I'm guessing the camcorder would pick up the IR illumination the night-vision goggles use, and the news station could show "how your ticket money at thus-and-such theatre goes to pay for the guards to spy on you and your date watching a movie," and how the station is of course only doing the same thing themselves to expose the whole thing. (Who guards the guards, as it were.)
Here are a couple examples:
I would think by now that it would be considered perfectly normal for weblogs and the like to include pictures of friends in situations like that!
If the theatre owner sees that
and gives you a couple dozen free passes so as to get you back more often, would you think that he'd be likely to want to confiscate the cameras, cell-phones, camcorders, laptops, etc, of the group and make them want to leave?
Do you think he'd be likely to be in favor of laws that not only require he confiscate such items, but also require that he have these customers arrested?
I'm thinking the answer to both questions is "no".
Not being able to record such social outings would make them comparatively boring.
Fortunately, it's legal in my state to go to the theatre with a bunch of friends and have everyone enjoy the experience. It's too bad for the residents of California and Ohio that they're legally restrained from having as much fun.
At the beginning of the show, a few people representing the group get up and talk about the film, random historical tidbits related to it, etc., the story about how they heard about it and were able to "get in line" to have the film and rights to show it for a few days, and so forth, plus a bit of a status update on films they're trying to get.
During the yearly film festivals, producers, directors, actors, and others involved in the making of the films are often there as well. (Realize these are not-very-well-known independents, and for most of them it's still a novelty for them to talk about their work in front of an interested audience.)
Anyway, they're brought up on stage before and after the films, relating their viewpoint on things, answering questions, and basically doing a live and not-during-the-film version of a dvd's "directors soundtrack."
Naturally, multiple people take pictures and record all this, (including and especially the folks involved in making the films--or at least they'll ask for copies later.)
It can be a great social event, very interactive, and fun for everyone.
But if you can't legally record any of it, then it won't be as much fun: The film society can't have pictures of their events in their newsletter or website, the people involved in the making of the films won't have hardcopy records of their memories, and the general mood of the event will be diminished given the shortsighted legal restrictions.
Fortunately I live in a state that has not passed such a law. If a director of a small film is invited to a film festival in my state, and he has to decide between going to this film festival versus a similar event in California or Ohio, the fact that we don't legally limit him having fun like that means he's more likely to come to our event. Woohoo!
I think it's quite magnanimous of the residents of California and Ohio to have their elected representitives limit their theatre experiences so that I might enjoy mine all the more--all without me having to vote for these representatives or contribute a dime to their compaigns!
An example in the Postgresql 7.4 documentation on the vacuum command shows the new lines of output:
These improvements seem to address your complaint. Given that you wouldn't be needing to run the reindex or reindexdb commands on a regular basis anymore, I guess you might be changing your suggestions.That may be the case for the punchcard systems you've used, but that doesn't have to always be the case. In the system I'm familiar with, each potential ballot choice is not only labelled with the person running for the office, but also a number. The punch card itself is also labelled/printed with the same numbers next to its perforated holes.
(And after punching out all my votes I would of course always pull out the card and use those numbers to check that the punched-out holes & numbers matched up with the numbers of the people/referrendums I was voting for.)
So I'm amazed that anyone would have ever made a punchcard system with *blank* cards! (Other than registration marks I guess.) How do the elections people expect voters to verify their votes after they pull the cards out, I wonder?
Anyway, I live in Georgia, and this whole fairly sensible way of voting was thrown out recently, to be replaced with the touchscreen nonsense everyone's read about--which of course is something far worse!
Interesting side note: In the election before the punchards were replaced, there were demo units of the touchscreen system in each polling place, along with a person there showing off the new glitzy machines we'd all be using the next election.
From the demo-person's reaction to my questions/complaints, I was apparently the only person all day who had been less that wildly enthusiastic about the soon-to-be "upgrade" of the voting equipment! And even more shocking, even though the purpose of the demo was supposedly to determine voter reaction to the coming change, the only thing they were actually set up to do is inform the electorate of what changes would be imposed--there was no process and no way to submit comments or questions back upstream. I thought that was a pretty unusual way of doing a market survey!
Let me snip your argument down a bit, (mainly removing supporting data that isn't really in dispute):
I think you might be confusing the likely failure of a particular unchanged business model under a changed regulatory environment with a claim of failure of *any* business model under a changed regulatory environment.
Let me know if I'm misrepresenting your argument, but it basically seems to boil down to something like the following:
Okay, maybe the wording for the first part of #9 above was a bit strong. :-)
However, you still concluded that there can be no viable business model for drug development in a no-patent world.
Below I give the same line of reasoning for a different field. After examining that argument, I wonder if you would also conclude that there is no viable model for development in that field under less monopolistic conditions.
The arguments are the same--or do you consider one argument valid and the other invalid?
(NOTE: I do not claim here that a business model that may fit the case above will necessarily have an analagous business model that will fit the case below, or vice versa; I merely claim that both lines of reasoning that conclude with an impossibility of the existence of a viable business model are flawed.)
So here is the second, similar, and IMHO similarly flawed line of reasoning:
Can you explain your line of reasoning in more detail?
By the way, I posted basically the same question in another comment in response to someone else's comment making a similar claim.
Feel free to respond either here or there. :-)
You also make a longer version of the same claim:
People often make this sort of claim, but I don't think I've ever seen it actually backed up.Can you explain why it would be utterly point-blank impossible for old and new cures to be available and to be continually developed over the long-term in a reasonable fashion (that is, not at a snail's pace for the entire medical field), if we were living in a world without patents? Two notes:
- I'm asking you to prove something slightly different from what you actually claimed above, though I imagine it's basically similar to what you mean to claim.
- I admit that this may not be a fair question, in that I am in effect asking you to prove a negative.
In any event, I don't think I've ever seen anyone actually go through, step-by-step, and prove why medical progress would necessarily slow to a snails pace were we suddenly in a no-patent world. Instead it's more or less claimed to be obvious without too much real backing to the argument. Generally the arguments as I've seen them have all sorts of subtle assumptions. For instance: There are all sorts of assumptions in the above statement. Here are a couple:- There is an assumption that the group spending the $1B has an interest in selling the drug. Why does that have to be the case?
- There is an assumption that another group that could make and sell this new drug for just $1M is necessarily a (feared or potential) "competitor" that would eat away from the first group's business. Why would this have to be the case? Why would it, for instance, be utterly impossible for them to be ecstatic that someone else can make the drug for $1M, (even assuming no royalties go to the first group), perhaps even talking with this so-called "competitor" to find ways of making production cheaper for them, or sponsoring some of the costs?
Perhaps those assumptions would be considered to be valid for businesses operating in world with pharmaceutical patents, but I don't see how they must necessarily be indisputably valid for for-profit businesses operating in a world without patents.Suppose I make a soft drink vending machine, that I refill myself, and pay you for the privilege of putting outside (or maybe even inside) your place of business. Let's say I pay you five cents per purchased drink.
The vending machine runs linux internally, and it's connected to a network, and I use that connection to make sure it's optimally filled up with drinks. (I come around and fill it up when things run low.)
With this situation, GPL'd software is in a machine at your facility.
Would you say I am obligated via the GPL to provide you with a copy of the source of all the GPL'd programs in the vending machine? Would I be obligated to provide a copy to all the vending machine's customers?
What if I allowed you some access to the machine. Say I allowed you to ssh in and run a single command that displayed logfiles of when people bought different drinks. Would that change anything in the above answers?
Or, remembering your comment that the nature of the financial transaction doesn't matter, if I rented you the machine instead, so that you were checking it's current load and refilling it up regularly, (ie, using the software and "renting" the machine and software, as opposed to neither renting nor owning as in the cases above). Would that change anything?
I used to work in the Response Center at HP. I took linux support calls right along with HP-UX calls; we sold support contracts for various linux distributions as far back as 1999.
BTW the need for both phone and ethernet surge protection is because I'll be using phone lines as a backup for data syncs, if my client's network is down, or if they don't have a network.
The dual surge protection requirement is one of the less serious requirements--you can always buy an additional phone line/ethernet surge protection wall-wart type of thing to cover for one or the other not being there--but it's kind of ugly.
However, the big thing for me is to make sure the UPS behavior actually really does work correctly. For instance, if the UPS looses power and goes battery-low, and the system shuts down but the power comes back on during the shutdown process, the UPS had better still cut power for a short while, so the machines have an actual powerloss and then poweron to start them up again. Not every ups and software combination does this correctly all the time.
As an aside, does anyone know of any very small (in size, va rating, and cost) ups's that have both serial and usb connections, spacing for 2 transformers plus 1 non-transformer plug, avr of some sort, separate phone line and ethernet surge protection built in, (that is, you can plug in a phone line and an ethernet connection at the same time), have some sort of smart signaling ("line power down and battery low--you better shut down now"; "okay, give me 2 minutes, then shut off my power whether or not wall power has come up, and keep me shut down until you've had reliable power for a while"), *and* that are properly documented by the manufacturer and work with some sort of open source ups software under linux?
I'm finding it hard to find any ups's that meet even half of those criterion, much less all of them.
Granted, I was very late (1.5 hours) to the Atlanta meetup, but no one else was there when I did finally arrive.
I had expected that this sort of event would follow other geek meeting-at-a-place-with-food traditions and last well into the night, but not only were there not any people still there with the slashdot meetup crowd, the employees at the coffee shop the meetup was scheduled to occur at didn't remember any large group or any "slashdot" group showing up earlier.
Given that there were at least 65 people rsvp'd, from what I remember, (now that the meeting start time has passed, slashdot.meetup.com doesn't allow you to see who all rsvp'd or read their comments), I would have though that at least 20-30 people would still be there.
I'm wondering if this might have been a hoax--Did other folks have similar experiences in other cities?