Uhhh...I think the idea is that, if you want backward or multiplatform compatibility, you ooften get an enormous help by building on a rich widely-deployed interface, *even* if that interface sucks in various ways. We've seen this a thousand times. Rewriting one of these legacy components sounds like a good idea, and the result is a much cleaner implementation, but you inevitably lose the bug-for-bug compatibility needed to run in these legacy environments. NOBODY is prepared to reimplement all the gory mess that was there before -- anyway, if that's what you want, why rewrite it?
Think about how many people stick with a buggy tool, wishing they could switch to the new improved model, but knowing they have to stick with what's already in use. (This is how WordPerfect, Adobe, and Autodesk, and of course Microsoft have been able to dominate various markets. I mean, why did MS/DOS last for so long? Exactly for this reason.)
I have the explanation for how this gizmo works. As you lean in the direction you want to go, gravity acting on your body tries to pull you in that direction. The scooter, using its internal smarts, rolls that way without tipping over, maintaining balance, and thus you move off efficiently. It's the same propulsion system used in many perpetual motion machines. You can go as far and as fast as you want, powered by the acceleration of gravity.
(Please realise this is tongue-in-cheek. Though on a flat surface you can do alot with gravity and momentum -- that's why scooters work.)
> You're the guy that I'm worried about...No offense
Ha ha. Well, actually, I'm not that guy, and I basically agree with everything you've said in this post. I think I was answering a different question, maybe one that you didn't ask. I had concluded that your customer list had been randomly accumulated, with no opt-in process, and that your customers weren't necessarily expecting to hear from you except in response to their own messages. (Example: When I place an order with amazon.com, I don't want to start receiving ads, and I damn sure don't want them to sell my name to somebody else. An amazon.com party invitation? I think in that case, given the number of their customers and what a small part of their business I represent, it would be an inappropriate use of my address. There's no reasonable expectation that ordering a $10.95 book would somehow put me onto their A-list.) It sounds like you have a much closer relationship with your customers, and so it's not black-and-white. Well, again, as I said earlier, if it were black-and-white then it wouldn't be worth discussing.
To use your restaurant analogy, rather than collecting business cards I thought you were saving the telephone numbers left when people made reservations -- in that case, there's NO expectation that they'll start gettting telemarketing calls.
I totally agree that this spam blocking approach -- the one this thread is putatively about -- has real weaknesses, and that (depending on how it's implemented) one individual might have the ability to block legitimate mailings. I suppose one approach would be to withhold action until some number of complaints are received -- just like how your cable company won't send out a service truck until 3 calls come in.
But returning to my original post, I stand by my belief that a bulk mailing is spam unless there's a clear opt-in by the recipients. This opt-in could come from several means: via an explicit opt-in form; by the user manually submitting an email address; or as part of your published terms of service. (Given the abuses of list sharing, I feel that there should be NO way to resell an email address obtained from an outside source -- the only address that you should be able to share is one that you've been told first-hand can be shared. In fact, if this were done, and we got rid of the faked sender addresses etc., opt-out would have a better chance of working. But today, once you get on one list being distributed, you're screwed -- you can never stop the barrage.)
I hope my position doesn't sound as draconian as my original post made it. I was responding to the philosophical question of "What is spam?" rather than the practical questions "Is this spam blocker a good idea?" or "Were we good guys with this party mailing?" You were good guys; but if I were you, I'd institute procedures to maintain the customer list more carefully, so that NO customer could ever be surprised by an invitation in the future. They'll know whether or not such mailings are likely at the time they leave their email addresses. And that's good customer service.
> I may have been sending unsolicited email to a large number of > people, but I had an established, existing relationship with those people already.
I totally agree. It's not the size of the mailing, it's how you got the email addresses. Sending bulk email to people who expect to hear from you is, of course, A-OK.
I'd like it to be illegal to sell an email list without showing the source of each entry, and without documentation of an opt-in for each entry. There is no situation where I'd want company A to make my information available to company B, unless I explicitly told them they could release my name.
But that law could never happen, of course, and even if it could it wouldn't protect against international sources.
I think this issue is important enough that public discussion is worthwhile, and you've made your points well -- again, as it says in my sig, if both sides of the argument didn't have some validity, the whole issue would have been resolved long ago. And, although many/. readers will dismiss what you're saying because you're a sender of UCE, I think your points merit discussion.
But here's why I think this situation is black-and-white.
Bad spammers have swamped the medium. Therefore it doesn't matter that there may be good spammers; I can't opt out to *ANY* UCE, because there's no way that I can distinguish between the white hats and the black hats. Therefore opt-out does not help. If we could shut down the spoofers and list churners, and opt-out had a reasonable chance of working, then I'd be singing a different song. But so far we can't.
For the same reason, I don't equate catalogues and other unsolicited bulk snail mail with UCE. Those mailings are selective, because there's a substantial cost in sending it out. But UCE has been taken over by the bad guys. Therefore, in my opinion, a good guy sending UCE is actually contributing to the spam problem, by further muddying the water. I will NEVER reply to any UCE.
If we could cut out the anonymous and spoofed crap, and UCE always had a valid sender, opt-out mechanism, etc., I might become a UCE supporter. But at the moment the amount of spam I receive grows steadily each month, the amount of time I spend dealing with it grows accordingly, and the amount of virus-infected spam goes up as well.
So my view is that if you're not against UCE, then you're part of why we have a spam problem.
But again, this is just my opinion, and you've made one of the better cases for 'white hat' UCE that I've heard.
> some people don't see relevance anywhere, whilst others see everything as interconnected issues
Well, I think the point here is that if you are running a site with a particular focus, and you take what you're doing seriously, then it's your call what's inside the fence and what isn't. If you see relevance everywhere, or hardly anywhere, it's up to you to make the call. If the folks who visit the site don't agree with your judgment, they'll let you know, and if you are consistenly an idiot then they'll split, and you'll wind up conducting a monologue. But if your judgment is reasonably close to theirs, then the community with thrive.
The bottom line is whether it's your site or an unmoderated public site. If it's your site, then you have the right, ability, and duty to keep the discussion on topic. In my experience, a well-moderated site has very little trouble with abuse, yet still maintains an open forum with no perception of censorship, because the posts that get removed are the ones that piss everybody off.
But again, this works in direct proportion to how focused the topic is for the site. If you're discussing banana farming, to use your example, and everybody on the site is interested in banana farming, it will be obvious to everybody if some newcomer starts talking trash and should get booted.
This issue gets argued both ways. I've seen lots of claims like yours, that bulk mail is the profit-maker. But I've also seen figures that show first-class mail to be subsidizing bulk mail. I think there's a lot of deliberate obfuscation here. (I've had direct professional involvement with the USPS and with direct marketers at various points, and so I'm talking about well-researched statistics, not the types of claims you'd find in The Onion.) As usual, a good clue is to see who is in the best position to influence the associated policy and laws. The direct marketers have a strong lobby and incredible influence on the postal service. First class mail is less and less important from a political standpoint. Therefore, I choose to believe the statistics that show first class mail to be the victim, because it's in a good position to be a victim.
But here's another way to look at it. If you got rid of all that bulk mail, you could do the postal service's job with 1/10th the resources. There's a reason that FedEx and UPS can make a profit: We don't mind paying for efficient delivery of the stuff we really want.
> On the other hand, it's my responsibility to take them off those lists at their request
I forgot to comment on this very important point. Opting-out is not a valid mechanism, because as you know many spammers use opt-out responses as a way to maintain their lists of valid email addresses. So opting-out may work for a customer/vendor relationship (but opting-in is just as easy); but it will not work for bulk email received from an unknown party. And even with a known vendor, opting-out can be dangerous, because it would be easy to forge an opt-out mechanism with a legit business name for the purpose of collecting email addresses. "Do you want any more bulk pr0n emails from Microsoft? Click here to remove your name from our list." I've received these, by the way.
But just to clarify: I don't think these guys were being bad, I just think we need to have a very clear definition of spam that is not dependent on the spammer's intent. I thought that issue was at the heart of the original post.
I don't think I'm taking this to an extreme, but perhaps I was responding to a different point from the one the original author intended. Let me try an entirely different tack:
The problem here was that, when this company assembled its email mailing list, it should have been very precise about how the list would be used. When customers provided their email addresses, they should have been able to say "Only respond to my emails," "Send me new product info," "Tell me about parties," "Sell my email address to bulk emailers in Malaysia," etc. So the fault was in how the list was created. By the time they were sending out party invitations, they didn't have a valid mailing list for sending such invitations.
Now, I agree that it sounds like they proceeded in a reasonable way, and I bet they'll be quite clear in their future list maintenance. But at a philosophical level, was this action spam or not? If you say it wasn't, you've carved out an important exception to our definition of spam. You'd have to say that any vendor can send bulk email if they have a reasonable expectation that you might want those emails. Well, every spammer would claim this.
This party invitation problem is an extreme case; but it's in the extreme cases that we find the places to draw the line. I'd say that that line was crossed, because email addresses were used for a purpose that hadn't been agreed upon in advance. I think that 'opt in' is the only valid approach to bulk email, because the situation is so badly abused at the moment.
I hope this position makes more sense to you. I don't think that unsolicted bulk email is at all like talking to people in a room. The assumptions and responsibilities are totally different.
One other point, though, about sending invoices. If you send an invoice via email, but email hadn't been discussed as an acceptable medium, then no, I don't think your customer has any obligation to pay. IANAL, but there's always lots of steps to go through to set up electronic billing, and that must be for a good reason.
I think you may have misunderstood what I consider spam. If I've given you my email address and have agreed to receive email from you, then of course it's not spam. The question was about unsolicited unexpected email, e.g. messages sent using a mailing list from a third-party source.
> I wonder if you have ever ran a business
Yes, I've run my own business since 1980. I've also done marketing, promotion, and support plans for businesses that use direct mail, so I'm familiar with the issues.
> email is a very effective way of keeping your customers in touch > Most of my vendors...I also trust with my email address
Yes, indeed, email is a great way to keep vendors and customers in touch. But if your customers give you their email addresses and opt-in for mailings, it's not unsolicited email. I agree that this is a great use of technology.
The whole point of this thread was to discuss whether sending 'nice' spam to people who have not agreed to receive your email is still spam. In my opinion, there is no 'nice' spam.
> [What if] I sent you a nice, fancy invitation to my New Years party via snail mail...?
The point is that email is different from snail mail. You wouldn't spend the money to mail me an invitation unless you knew me, or there was some legit reason to invite me (e.g. we live on the same block). Email is swamped with spam -- in my case, 90% of what I receive is spam. Anybody adding to that burden is 'over the line' as far as I'm concerned. I didn't quite get your final comment about talkin' out my ass, but my point is that I follow a strict policy about spam and about telemarketing. I simply will not do business with any spammer, nor will I respond to any telemarketing offer. So no matter what wonderful party invitation might arrive by spam, I wouldn't consider it, as a matter of principle (not that Playboy mansion invitations would be sent via spam!).
> Just wondering, does your firm find snail mail spam as offensive?
Good question. It's annoying as hell, and as far as I'm concerned bulk mail is the reason our postal system sucks. It concentrates most of the postal system's resources on the things we usually throw in the trash.
However, we don't penalize vendors who use bulk mailings for a few reasons.
1. It costs them (considerable) money to do a mailing. There's thus a built-in incentive to send appropriate, useful, legit information to people who have been selected on some plausible basis. Otherwise they're wasting their money. So there's a chance that a piece of bulk mail might actually be of interest. (Unlike email spam.)
2. The majority of bulk mailers are legit businesses, and many of them are companies we've actually heard about before, or even done business with. (Unlike email spam.)
3. There's a long tradition in this country of promotion through bulk mailing. Even though countless AOL mailings may piss me off, it's not sensible to regard bulk mailers as intrinsically evil. (Unlike email spam.)
4. A good chunk of the bulk mail is coming from companies where we've done business before (due to their selection criteria used in (1) above). So although it may be unsolicited it's not from an unknown source. (Unlike email spam.)
5. Bulk mail generally identifies the sender, including a valid mailing address. (Unlike email spam.)
So though a) I waste plenty of time each month going through unwanted bulk mail, b) I can't remember the last time a bulk mailing from an unknown vendor actually generated any business from me, and c) we always opt out from bulk mailing lists and form any sharing of customer data, we still don't penalize vendors who use bulk snail mail. (Unlike email spam.) Not counting expected catalogues from vendors (which though it's bulk mail is not spam), we throw out 99% of the bulk mail that arrives. One piece out of 100 is either funny enough or relevant enough to avoid an instant trip to the dustbin.
If you're running a site that has a particular mission, e.g. to discuss wristwatches or Earley's Algorithm, then I think it's totally appropriate to remove off-topic posts. It's not censorship, it's moderation, and it's part of the site's intent. Often, such moderation is the difference between an interesting/useful site and yet another cheesy slugfest.
In general I abhor censorship, but censorship is only relevant IMO when you're talking about an open public forum. Editorial judgment is quite appropriate and even necessary in a forum with a specific purpose.
I get a ton of spam addressed directly to me. (I think this all stems from a time years ago when I foolishly posted my email address on a bulletin board with a very limited user group. Normally, all the messages scrolled away after a week or so, so I thought I was safe. But unfortunately somebody archived the posts and they wound up somewhere permanent, and the harvesters have had me since then.):( So though this is a good approach, in my case it doesn't help much.
You make an interesting point, but I think you have to lose this one. If there were absolutely no legitimate basis for unsolicited bulk email then we'd already have laws against it. The problem, as usual, is that there are pros and cons. But I think we (on/.) have concluded that the cons totally outweigh the pros in this matter. Even though you and your organization may have been on the side of the angels, you could just as well have said this: "We wanted to send party invitations, so we hacked into each of our customers' servers and put a message on their home page. Yeah, lots of crackers are evil, but we were doing this with good intentions, so it was OK."
If you want to send out party invitations, get your customers to opt-in through your normal commercial channels.
I would never go to a party announced via spam, even if it were at the Playboy mansion with hot and cold running blondes. If you were my vendor, what you did would have put you on the 'prohibited vendor list' -- we have a strict no-spam policy.
Others may not agree, but that's how I see it. Needless to say, of course, I'm not flaming you about this -- it sounds like you tried to do the right thing. I'm just saying that, in my company, this would have had direct commercial repercussions against your firm.
Your point is absolutely right, of course, and my post wasn't an argument that all software belongs under such a license (even though it does have the best name) -- I was being slightly tongue-in-cheek. But I think there's an important place for nonrestrictive licenses. They're desirable for software that should be as widely distributed and used as possible. Consider some widget that is a) highly useful, b) widely applicable, c) well-implemented, d) not worth reimplimenting so long as a good public version exists, and e) not so intrinsically valuable that there's an incentive to build expensive derivative works. Device drivers are good examples. So are certain function libraries and algorithm implementations. Compilers for novel languages often fall in this class. A nonrestrictive license might be the best means for encouraging wide distribution and use. Once you have a huge user community, open-source economics will mitigate against somebody getting rich off your work -- or, if they do, it's because they've added enough value that they arguably would deserve it.
I think the heart of the matter is sociological rather than legalistic or economic. It's a question of how to create and sustain a user community. You might regard a piece of code as 'your baby,' and only be prepared to share it with people who promise not to make money off your baby. In that case, it's 'all about you.' But you might instead want to get as many people excited about your baby as possible -- in which case giving new users the ability to make some money off it is a positive inducement. To use an extreme example: Suppose you designed a kewl language, wrote an efficient compiler, and got lots of praise from your initial users. But you craft a license agreement that not only restricts sale of the compiler code, but further restricts users from selling any applications built with your language. You might get praise from RMS and others who feel that all software should in principle be free; but you won't win the hearts and minds of developers whose salaries depend on the ability to build software products. It's so hard to promulgate a new language anyway; and this extra restriction would cut out the very people most likely to have an open mind about new technology. Even end-user organizations would balk at building their custom apps with your language, if they must give up the option of ever reselling them, and if their software vendors aren't embracing it.
So I guess my point is that I see a need for several licensing regimes, appropriate for different kinds of software and software users.
There are certain applications that are too specialized and expensive to get built through an open source community, and will thus require a commercial R&D team that can only be funded through proprietary licensing.
There are widely-used and widely-needed applications that are best served by communities working under something like GPL or LGPL.
And there are certain components that should be as widely-distributed as possible, where everybody benefits through standardization even outside the open source community; and if these components have very nonrestrictive licenses, it's easier to proselytize effectively and reduce the temptation or need for anybody to roll their own solution.
The Artistic License is one I like. I'm always suspicious of an open source license that either a) has a polemical preamble that tries to coerce your behavior, b) reads like the team of lawyers who wrote the license are making a lot more money than the developers, or c) presumes that the only good programmer is one who either programs as a hobby, is an academic, works for a big company that can afford to subsidize the programmer's time, or works for an end-user company that can afford to build complex systems strictly for internal use -- in other words, that there's no moral way to be a software vendor.
Yeah, I know there's plenty of room for argument all around, but my sympathies are with small software vendors who need some way to get enough revenue from 100-5000 licenses to pay salaries. The Artistic License strikes me as compact and commonsensical, and a good model for many situations. And of course it has the coolest name.:-)
I agree with your concern about how useful or fair this will prove to be. I think there's a good fundamental goal of improving the quality of hits; but without really strong metamoderation, I can't see how this could be made fair. (I suspect it's probably more important to weed out irrelevant/bad hits, which might perhaps be easier to police.)
I wonder whether a bunch of volunteer moderators might be a better solution -- give them some freebie benefit for spending a few hours a day reviewing other people's searches and picking out the best and worst responses. But of course, those moderators won't know what original search was after, soo all they could do is weed out the pr0n sites etc.
Several of the posts here imply that Josh's intent is to find a way to sell FLAC to vendors for use in embedded systems. Those posts suggest a mix of open-source and fee-based licensing, and seem to view the embedded system vendors as folks who should essentially be paying a tax to subsidize development.
But my take on the original question is quite different. I believe that Josh's goal is to make this software available for free, to all, including embedded system vendors; but he sees(correctly) that such vendors generally have a hard time using open source components, because of the nature of how their systems are sold, distributed, and supported. It's not necessarily practical or reasonable to compel a hardware vendor to provide the means for recompilation/relinking etc. -- even though all parties may feel that an open-source component is the best technical solution, and that using it would help to establish good industry-wide standards. (This is also true of device drivers and other components where, even if the vendors don't conform 100% to open source requirements, a watered-down standarization is still far better than if each vendor rolls his own solution. Many 'benign' proprietary software vendors would love to use open source tools and components in a non-exploiting way, i.e. as a way to improve industry-wide interoperability and standardization -- as opposed to 'evil' vendors who would seek to put a private label on open source components and sell the result -- but of course even with good intents this can't be done, since a software vendor is committing the sin of trying to get paid for its R&D costs. But that's another argument.)
There are many good suggestions here. I think this is an extremely interesting/. thread, and it has raised some key issues.
The bottom line for me is a basic problem with many open source licenses: Licensing restrictions often prevent a developer from making software available to folks who common sense says should be able to use it; and they often prevent a product vendor from incorporating software in ways that common sense says is consistent with the open source philosophy and would improve the overall quality of products and range of choices available to end-users.
Thanks to all who have made thoughtful posts here.
Quite right. I do understand why AT&T did the deal that way. But I was infuriated that, between AT&T and Worldcom, there was zero notice of pending troubles. My connections suddenly went dark, and as you can imagine it took weeks before I could provision new service. Given AT&T's prominence and strength, I'd have thought they could have negotiated with the RBOC's for some kind of transition period for the stranded customers -- even a day or two would have made a huge difference. And of course UUNET knew trouble was in the wind for weeks, but gave us no warning. (Even though the industry's CLEC troubles were well-known to/. folks, my contracts were with a tier-1 provider, not with a fly-by-night CLEC, so I thought I would be protected.)
So anyway, my comment was a simple warning that folks should not blithely assume that AT&T's purchase of assets would necessarily mean continued service. Caveat emptor and all that.
When AT&T acquired the Northpoint DSL assets, they stranded all the affected DSL users. I had multiple UUNET DSL connections using Northpoint as a CLEC, and all were dropped overnight with zero notice -- suddenly one day, there was no service. All those emails, pending web transactions, etc. went into the dustbin of history. UUNET failed utterly to provide any advance warning, or any reasonable recovery strategy. (This is despite all those Worldcom TV ads about safety and security. Don't you believe it!) AT&T had (presumably) no legal obligation to keep existing Northpoint customers operating, since they just bought the assets; but if AT&T had given us a few days to get off our active connections, it would have been nice. Every vendor involved screwed the end-user.
I now have contracts with an RBOC who happens to provide broadband servce, so there's a single vendor. I'm still not totally safe, but at least I only have one set of schmucks to deal with.
Bottom line: Have contingency plans that assume your ISP and all intermediaries will fail and will leave you high and dry. (UUNET's contract had a force majeure clause that excluded service interruption due to their suppliers -- one assumed this was to cover the case where a tornado hit a central office, but it apparently was much broader. Here's a similar situation: "Oh, CISCO has stopped supporting the principal router model in our network, so we're terminating your contract rather than upgrading our hardware.")
One accelerometer can't. You need accelerometers on all three axes.... However, without an occasional speed or position sample, your calculated speed will always be drifting. -- seanadams
Precisely. The original suggestion was for putting a (single) accelerometer on-chip. If you require several accelerometers, solid state gyros, and other sensors, the thing gets to be quite a bit more than a single chip.
It's useful to remember that correcting drift of intertial guidance systems is a core feature of commercial aircraft waypoint navigation. GPS lets us eliminate this complex method.
Uhh... I can't see how this would do anything helpful without a gyroscope. How would an accelerometer distinguish between rotation, acceleration, deceleration, etc.? You might convince me you can put an accelerometer on a chip (surprising but what the hell) but I don't expect to see gyros!
Anyway, commercial aircraft have inertial navigation to back up their other navigation aids. (Recall that GPS is still a newcomer to nonmilitary aviation.)
The rest of the world pays the US (say) $3B, and the US relinquishes any capability for military control of the system. As was pointed out elsewhere, there's no strong US military argument for retaining control over GPS transmissions. This would avoid most of that wasted EU investment, and the US would get some of its R&D investment back at a time when cash would come in handy. There's no reason I can think of that we need multiple incompatible geopositioning systems around this planet.
Come to think of it, bundle in a satellite warning system while we're at it, and make the whole package available to all nations.:)
Slashdot Mirror
Uhhh...I think the idea is that, if you want backward or multiplatform compatibility, you ooften get an enormous help by building on a rich widely-deployed interface, *even* if that interface sucks in various ways. We've seen this a thousand times. Rewriting one of these legacy components sounds like a good idea, and the result is a much cleaner implementation, but you inevitably lose the bug-for-bug compatibility needed to run in these legacy environments. NOBODY is prepared to reimplement all the gory mess that was there before -- anyway, if that's what you want, why rewrite it?
Think about how many people stick with a buggy tool, wishing they could switch to the new improved model, but knowing they have to stick with what's already in use. (This is how WordPerfect, Adobe, and Autodesk, and of course Microsoft have been able to dominate various markets. I mean, why did MS/DOS last for so long? Exactly for this reason.)
I have the explanation for how this gizmo works. As you lean in the direction you want to go, gravity acting on your body tries to pull you in that direction. The scooter, using its internal smarts, rolls that way without tipping over, maintaining balance, and thus you move off efficiently. It's the same propulsion system used in many perpetual motion machines. You can go as far and as fast as you want, powered by the acceleration of gravity.
(Please realise this is tongue-in-cheek. Though on a flat surface you can do alot with gravity and momentum -- that's why scooters work.)
> You're the guy that I'm worried about...No offense
Ha ha. Well, actually, I'm not that guy, and I basically agree with everything you've said in this post. I think I was answering a different question, maybe one that you didn't ask. I had concluded that your customer list had been randomly accumulated, with no opt-in process, and that your customers weren't necessarily expecting to hear from you except in response to their own messages. (Example: When I place an order with amazon.com, I don't want to start receiving ads, and I damn sure don't want them to sell my name to somebody else. An amazon.com party invitation? I think in that case, given the number of their customers and what a small part of their business I represent, it would be an inappropriate use of my address. There's no reasonable expectation that ordering a $10.95 book would somehow put me onto their A-list.) It sounds like you have a much closer relationship with your customers, and so it's not black-and-white. Well, again, as I said earlier, if it were black-and-white then it wouldn't be worth discussing.
To use your restaurant analogy, rather than collecting business cards I thought you were saving the telephone numbers left when people made reservations -- in that case, there's NO expectation that they'll start gettting telemarketing calls.
I totally agree that this spam blocking approach -- the one this thread is putatively about -- has real weaknesses, and that (depending on how it's implemented) one individual might have the ability to block legitimate mailings. I suppose one approach would be to withhold action until some number of complaints are received -- just like how your cable company won't send out a service truck until 3 calls come in.
But returning to my original post, I stand by my belief that a bulk mailing is spam unless there's a clear opt-in by the recipients. This opt-in could come from several means: via an explicit opt-in form; by the user manually submitting an email address; or as part of your published terms of service. (Given the abuses of list sharing, I feel that there should be NO way to resell an email address obtained from an outside source -- the only address that you should be able to share is one that you've been told first-hand can be shared. In fact, if this were done, and we got rid of the faked sender addresses etc., opt-out would have a better chance of working. But today, once you get on one list being distributed, you're screwed -- you can never stop the barrage.)
I hope my position doesn't sound as draconian as my original post made it. I was responding to the philosophical question of "What is spam?" rather than the practical questions "Is this spam blocker a good idea?" or "Were we good guys with this party mailing?" You were good guys; but if I were you, I'd institute procedures to maintain the customer list more carefully, so that NO customer could ever be surprised by an invitation in the future. They'll know whether or not such mailings are likely at the time they leave their email addresses. And that's good customer service.
> I may have been sending unsolicited email to a large number of
> people, but I had an established, existing relationship with those people already.
I totally agree. It's not the size of the mailing, it's how you got the email addresses. Sending bulk email to people who expect to hear from you is, of course, A-OK.
I'd like it to be illegal to sell an email list without showing the source of each entry, and without documentation of an opt-in for each entry. There is no situation where I'd want company A to make my information available to company B, unless I explicitly told them they could release my name.
But that law could never happen, of course, and even if it could it wouldn't protect against international sources.
But here's why I think this situation is black-and-white.
Bad spammers have swamped the medium. Therefore it doesn't matter that there may be good spammers; I can't opt out to *ANY* UCE, because there's no way that I can distinguish between the white hats and the black hats. Therefore opt-out does not help. If we could shut down the spoofers and list churners, and opt-out had a reasonable chance of working, then I'd be singing a different song. But so far we can't.
For the same reason, I don't equate catalogues and other unsolicited bulk snail mail with UCE. Those mailings are selective, because there's a substantial cost in sending it out. But UCE has been taken over by the bad guys. Therefore, in my opinion, a good guy sending UCE is actually contributing to the spam problem, by further muddying the water. I will NEVER reply to any UCE.
If we could cut out the anonymous and spoofed crap, and UCE always had a valid sender, opt-out mechanism, etc., I might become a UCE supporter. But at the moment the amount of spam I receive grows steadily each month, the amount of time I spend dealing with it grows accordingly, and the amount of virus-infected spam goes up as well.
So my view is that if you're not against UCE, then you're part of why we have a spam problem.
But again, this is just my opinion, and you've made one of the better cases for 'white hat' UCE that I've heard.
Ha ha!
> some people don't see relevance anywhere, whilst others see everything as interconnected issues
Well, I think the point here is that if you are running a site with a particular focus, and you take what you're doing seriously, then it's your call what's inside the fence and what isn't. If you see relevance everywhere, or hardly anywhere, it's up to you to make the call. If the folks who visit the site don't agree with your judgment, they'll let you know, and if you are consistenly an idiot then they'll split, and you'll wind up conducting a monologue. But if your judgment is reasonably close to theirs, then the community with thrive.
The bottom line is whether it's your site or an unmoderated public site. If it's your site, then you have the right, ability, and duty to keep the discussion on topic. In my experience, a well-moderated site has very little trouble with abuse, yet still maintains an open forum with no perception of censorship, because the posts that get removed are the ones that piss everybody off.
But again, this works in direct proportion to how focused the topic is for the site. If you're discussing banana farming, to use your example, and everybody on the site is interested in banana farming, it will be obvious to everybody if some newcomer starts talking trash and should get booted.
This issue gets argued both ways. I've seen lots of claims like yours, that bulk mail is the profit-maker. But I've also seen figures that show first-class mail to be subsidizing bulk mail. I think there's a lot of deliberate obfuscation here. (I've had direct professional involvement with the USPS and with direct marketers at various points, and so I'm talking about well-researched statistics, not the types of claims you'd find in The Onion.) As usual, a good clue is to see who is in the best position to influence the associated policy and laws. The direct marketers have a strong lobby and incredible influence on the postal service. First class mail is less and less important from a political standpoint. Therefore, I choose to believe the statistics that show first class mail to be the victim, because it's in a good position to be a victim.
But here's another way to look at it. If you got rid of all that bulk mail, you could do the postal service's job with 1/10th the resources. There's a reason that FedEx and UPS can make a profit: We don't mind paying for efficient delivery of the stuff we really want.
> On the other hand, it's my responsibility to take them off those lists at their request
I forgot to comment on this very important point. Opting-out is not a valid mechanism, because as you know many spammers use opt-out responses as a way to maintain their lists of valid email addresses. So opting-out may work for a customer/vendor relationship (but opting-in is just as easy); but it will not work for bulk email received from an unknown party. And even with a known vendor, opting-out can be dangerous, because it would be easy to forge an opt-out mechanism with a legit business name for the purpose of collecting email addresses. "Do you want any more bulk pr0n emails from Microsoft? Click here to remove your name from our list." I've received these, by the way.
But just to clarify: I don't think these guys were being bad, I just think we need to have a very clear definition of spam that is not dependent on the spammer's intent. I thought that issue was at the heart of the original post.
I don't think I'm taking this to an extreme, but perhaps I was responding to a different point from the one the original author intended. Let me try an entirely different tack:
The problem here was that, when this company assembled its email mailing list, it should have been very precise about how the list would be used. When customers provided their email addresses, they should have been able to say "Only respond to my emails," "Send me new product info," "Tell me about parties," "Sell my email address to bulk emailers in Malaysia," etc. So the fault was in how the list was created. By the time they were sending out party invitations, they didn't have a valid mailing list for sending such invitations.
Now, I agree that it sounds like they proceeded in a reasonable way, and I bet they'll be quite clear in their future list maintenance. But at a philosophical level, was this action spam or not? If you say it wasn't, you've carved out an important exception to our definition of spam. You'd have to say that any vendor can send bulk email if they have a reasonable expectation that you might want those emails. Well, every spammer would claim this.
This party invitation problem is an extreme case; but it's in the extreme cases that we find the places to draw the line. I'd say that that line was crossed, because email addresses were used for a purpose that hadn't been agreed upon in advance. I think that 'opt in' is the only valid approach to bulk email, because the situation is so badly abused at the moment.
I hope this position makes more sense to you. I don't think that unsolicted bulk email is at all like talking to people in a room. The assumptions and responsibilities are totally different.
One other point, though, about sending invoices. If you send an invoice via email, but email hadn't been discussed as an acceptable medium, then no, I don't think your customer has any obligation to pay. IANAL, but there's always lots of steps to go through to set up electronic billing, and that must be for a good reason.
I think you may have misunderstood what I consider spam. If I've given you my email address and have agreed to receive email from you, then of course it's not spam. The question was about unsolicited unexpected email, e.g. messages sent using a mailing list from a third-party source.
> I wonder if you have ever ran a business
Yes, I've run my own business since 1980. I've also done marketing, promotion, and support plans for businesses that use direct mail, so I'm familiar with the issues.
> email is a very effective way of keeping your customers in touch
> Most of my vendors...I also trust with my email address
Yes, indeed, email is a great way to keep vendors and customers in touch. But if your customers give you their email addresses and opt-in for mailings, it's not unsolicited email. I agree that this is a great use of technology.
The whole point of this thread was to discuss whether sending 'nice' spam to people who have not agreed to receive your email is still spam. In my opinion, there is no 'nice' spam.
> [What if] I sent you a nice, fancy invitation to my New Years party via snail mail...?
The point is that email is different from snail mail. You wouldn't spend the money to mail me an invitation unless you knew me, or there was some legit reason to invite me (e.g. we live on the same block). Email is swamped with spam -- in my case, 90% of what I receive is spam. Anybody adding to that burden is 'over the line' as far as I'm concerned. I didn't quite get your final comment about talkin' out my ass, but my point is that I follow a strict policy about spam and about telemarketing. I simply will not do business with any spammer, nor will I respond to any telemarketing offer. So no matter what wonderful party invitation might arrive by spam, I wouldn't consider it, as a matter of principle (not that Playboy mansion invitations would be sent via spam!).
Good question. It's annoying as hell, and as far as I'm concerned bulk mail is the reason our postal system sucks. It concentrates most of the postal system's resources on the things we usually throw in the trash.
However, we don't penalize vendors who use bulk mailings for a few reasons.
1. It costs them (considerable) money to do a mailing. There's thus a built-in incentive to send appropriate, useful, legit information to people who have been selected on some plausible basis. Otherwise they're wasting their money. So there's a chance that a piece of bulk mail might actually be of interest. (Unlike email spam.)
2. The majority of bulk mailers are legit businesses, and many of them are companies we've actually heard about before, or even done business with. (Unlike email spam.)
3. There's a long tradition in this country of promotion through bulk mailing. Even though countless AOL mailings may piss me off, it's not sensible to regard bulk mailers as intrinsically evil. (Unlike email spam.)
4. A good chunk of the bulk mail is coming from companies where we've done business before (due to their selection criteria used in (1) above). So although it may be unsolicited it's not from an unknown source. (Unlike email spam.)
5. Bulk mail generally identifies the sender, including a valid mailing address. (Unlike email spam.)
So though a) I waste plenty of time each month going through unwanted bulk mail, b) I can't remember the last time a bulk mailing from an unknown vendor actually generated any business from me, and c) we always opt out from bulk mailing lists and form any sharing of customer data, we still don't penalize vendors who use bulk snail mail. (Unlike email spam.) Not counting expected catalogues from vendors (which though it's bulk mail is not spam), we throw out 99% of the bulk mail that arrives. One piece out of 100 is either funny enough or relevant enough to avoid an instant trip to the dustbin.
If you're running a site that has a particular mission, e.g. to discuss wristwatches or Earley's Algorithm, then I think it's totally appropriate to remove off-topic posts. It's not censorship, it's moderation, and it's part of the site's intent. Often, such moderation is the difference between an interesting/useful site and yet another cheesy slugfest.
In general I abhor censorship, but censorship is only relevant IMO when you're talking about an open public forum. Editorial judgment is quite appropriate and even necessary in a forum with a specific purpose.
I get a ton of spam addressed directly to me. (I think this all stems from a time years ago when I foolishly posted my email address on a bulletin board with a very limited user group. Normally, all the messages scrolled away after a week or so, so I thought I was safe. But unfortunately somebody archived the posts and they wound up somewhere permanent, and the harvesters have had me since then.) :( So though this is a good approach, in my case it doesn't help much.
You make an interesting point, but I think you have to lose this one. If there were absolutely no legitimate basis for unsolicited bulk email then we'd already have laws against it. The problem, as usual, is that there are pros and cons. But I think we (on /.) have concluded that the cons totally outweigh the pros in this matter. Even though you and your organization may have been on the side of the angels, you could just as well have said this: "We wanted to send party invitations, so we hacked into each of our customers' servers and put a message on their home page. Yeah, lots of crackers are evil, but we were doing this with good intentions, so it was OK."
If you want to send out party invitations, get your customers to opt-in through your normal commercial channels.
I would never go to a party announced via spam, even if it were at the Playboy mansion with hot and cold running blondes. If you were my vendor, what you did would have put you on the 'prohibited vendor list' -- we have a strict no-spam policy.
Others may not agree, but that's how I see it. Needless to say, of course, I'm not flaming you about this -- it sounds like you tried to do the right thing. I'm just saying that, in my company, this would have had direct commercial repercussions against your firm.
I think the heart of the matter is sociological rather than legalistic or economic. It's a question of how to create and sustain a user community. You might regard a piece of code as 'your baby,' and only be prepared to share it with people who promise not to make money off your baby. In that case, it's 'all about you.' But you might instead want to get as many people excited about your baby as possible -- in which case giving new users the ability to make some money off it is a positive inducement. To use an extreme example: Suppose you designed a kewl language, wrote an efficient compiler, and got lots of praise from your initial users. But you craft a license agreement that not only restricts sale of the compiler code, but further restricts users from selling any applications built with your language. You might get praise from RMS and others who feel that all software should in principle be free; but you won't win the hearts and minds of developers whose salaries depend on the ability to build software products. It's so hard to promulgate a new language anyway; and this extra restriction would cut out the very people most likely to have an open mind about new technology. Even end-user organizations would balk at building their custom apps with your language, if they must give up the option of ever reselling them, and if their software vendors aren't embracing it.
So I guess my point is that I see a need for several licensing regimes, appropriate for different kinds of software and software users.
There are certain applications that are too specialized and expensive to get built through an open source community, and will thus require a commercial R&D team that can only be funded through proprietary licensing.
There are widely-used and widely-needed applications that are best served by communities working under something like GPL or LGPL.
And there are certain components that should be as widely-distributed as possible, where everybody benefits through standardization even outside the open source community; and if these components have very nonrestrictive licenses, it's easier to proselytize effectively and reduce the temptation or need for anybody to roll their own solution.
But of course, that's only my opinion.
The Artistic License is one I like. I'm always suspicious of an open source license that either a) has a polemical preamble that tries to coerce your behavior, b) reads like the team of lawyers who wrote the license are making a lot more money than the developers, or c) presumes that the only good programmer is one who either programs as a hobby, is an academic, works for a big company that can afford to subsidize the programmer's time, or works for an end-user company that can afford to build complex systems strictly for internal use -- in other words, that there's no moral way to be a software vendor.
:-)
Yeah, I know there's plenty of room for argument all around, but my sympathies are with small software vendors who need some way to get enough revenue from 100-5000 licenses to pay salaries. The Artistic License strikes me as compact and commonsensical, and a good model for many situations. And of course it has the coolest name.
I agree with your concern about how useful or fair this will prove to be. I think there's a good fundamental goal of improving the quality of hits; but without really strong metamoderation, I can't see how this could be made fair. (I suspect it's probably more important to weed out irrelevant/bad hits, which might perhaps be easier to police.)
I wonder whether a bunch of volunteer moderators might be a better solution -- give them some freebie benefit for spending a few hours a day reviewing other people's searches and picking out the best and worst responses. But of course, those moderators won't know what original search was after, soo all they could do is weed out the pr0n sites etc.
Several of the posts here imply that Josh's intent is to find a way to sell FLAC to vendors for use in embedded systems. Those posts suggest a mix of open-source and fee-based licensing, and seem to view the embedded system vendors as folks who should essentially be paying a tax to subsidize development.
/. thread, and it has raised some key issues.
But my take on the original question is quite different. I believe that Josh's goal is to make this software available for free, to all, including embedded system vendors; but he sees(correctly) that such vendors generally have a hard time using open source components, because of the nature of how their systems are sold, distributed, and supported. It's not necessarily practical or reasonable to compel a hardware vendor to provide the means for recompilation/relinking etc. -- even though all parties may feel that an open-source component is the best technical solution, and that using it would help to establish good industry-wide standards. (This is also true of device drivers and other components where, even if the vendors don't conform 100% to open source requirements, a watered-down standarization is still far better than if each vendor rolls his own solution. Many 'benign' proprietary software vendors would love to use open source tools and components in a non-exploiting way, i.e. as a way to improve industry-wide interoperability and standardization -- as opposed to 'evil' vendors who would seek to put a private label on open source components and sell the result -- but of course even with good intents this can't be done, since a software vendor is committing the sin of trying to get paid for its R&D costs. But that's another argument.)
There are many good suggestions here. I think this is an extremely interesting
The bottom line for me is a basic problem with many open source licenses: Licensing restrictions often prevent a developer from making software available to folks who common sense says should be able to use it; and they often prevent a product vendor from incorporating software in ways that common sense says is consistent with the open source philosophy and would improve the overall quality of products and range of choices available to end-users.
Thanks to all who have made thoughtful posts here.
Quite right. I do understand why AT&T did the deal that way. But I was infuriated that, between AT&T and Worldcom, there was zero notice of pending troubles. My connections suddenly went dark, and as you can imagine it took weeks before I could provision new service. Given AT&T's prominence and strength, I'd have thought they could have negotiated with the RBOC's for some kind of transition period for the stranded customers -- even a day or two would have made a huge difference. And of course UUNET knew trouble was in the wind for weeks, but gave us no warning. (Even though the industry's CLEC troubles were well-known to /. folks, my contracts were with a tier-1 provider, not with a fly-by-night CLEC, so I thought I would be protected.)
So anyway, my comment was a simple warning that folks should not blithely assume that AT&T's purchase of assets would necessarily mean continued service. Caveat emptor and all that.
When AT&T acquired the Northpoint DSL assets, they stranded all the affected DSL users. I had multiple UUNET DSL connections using Northpoint as a CLEC, and all were dropped overnight with zero notice -- suddenly one day, there was no service. All those emails, pending web transactions, etc. went into the dustbin of history. UUNET failed utterly to provide any advance warning, or any reasonable recovery strategy. (This is despite all those Worldcom TV ads about safety and security. Don't you believe it!) AT&T had (presumably) no legal obligation to keep existing Northpoint customers operating, since they just bought the assets; but if AT&T had given us a few days to get off our active connections, it would have been nice. Every vendor involved screwed the end-user.
I now have contracts with an RBOC who happens to provide broadband servce, so there's a single vendor. I'm still not totally safe, but at least I only have one set of schmucks to deal with.
Bottom line: Have contingency plans that assume your ISP and all intermediaries will fail and will leave you high and dry. (UUNET's contract had a force majeure clause that excluded service interruption due to their suppliers -- one assumed this was to cover the case where a tornado hit a central office, but it apparently was much broader. Here's a similar situation: "Oh, CISCO has stopped supporting the principal router model in our network, so we're terminating your contract rather than upgrading our hardware.")
One accelerometer can't. You need accelerometers on all three axes.... However, without an occasional speed or position sample, your calculated speed will always be drifting. -- seanadams
Precisely. The original suggestion was for putting a (single) accelerometer on-chip. If you require several accelerometers, solid state gyros, and other sensors, the thing gets to be quite a bit more than a single chip.
It's useful to remember that correcting drift of intertial guidance systems is a core feature of commercial aircraft waypoint navigation. GPS lets us eliminate this complex method.
I meant bundle this all into the deal with the EU.
Uhh... I can't see how this would do anything helpful without a gyroscope. How would an accelerometer distinguish between rotation, acceleration, deceleration, etc.? You might convince me you can put an accelerometer on a chip (surprising but what the hell) but I don't expect to see gyros!
Anyway, commercial aircraft have inertial navigation to back up their other navigation aids. (Recall that GPS is still a newcomer to nonmilitary aviation.)
Well, OK, sell it to the EU instead of the UN.... My point is, avoid paying for a duplicate system by sharing the cost of the first one.
The rest of the world pays the US (say) $3B, and the US relinquishes any capability for military control of the system. As was pointed out elsewhere, there's no strong US military argument for retaining control over GPS transmissions. This would avoid most of that wasted EU investment, and the US would get some of its R&D investment back at a time when cash would come in handy. There's no reason I can think of that we need multiple incompatible geopositioning systems around this planet.
:)
Come to think of it, bundle in a satellite warning system while we're at it, and make the whole package available to all nations.