Show me a standards-compliant page that renders differently in Opera.
It all depends on which subset of the standards you use.
A while back, I ran into some bugs with:first-line in Opera 8. Fully validating code, but styles would persist incorrectly and, in one case, text would disappear. One's fixed in Opera 9 TP2, the other is still around. Sure, this was real edge-case stuff, but it happens.
Here's a more practical example going the other way: creating frames with generated content. It works handily in Opera, not at all in IE, and can be made to almost-work in Firefox (it comes down to positioning on generated content based on different versions of the CSS spec).
Standards compliance in a browser isn't a matter of working down a checklist from item #1 to item #1000, and counting complaince based on whether you get to item #800, #900, or #999. You look through the standards and prioritize. Maybe both browsers fulfill 900 of those items -- but maybe item #723 is only implemented in one of them, because one browser picked it as a priority, and the other put it off.
Stepping away from standards-compliance for a moment, you also have to take into account things like XMLHTTPRequest, contentEditable, and so on. Most rich text editors (popular in webmail apps) couldn't run on Opera until they released the Opera 9 previews, because the required features weren't there. And every minor version of Opera from 8 on has made more adjustments for AJAX apps.
Even within standards -- both de facto and de jure -- you need to find a common subset. Validating the code doesn't guarantee that other browsers will behave the way you expect.
Plus there's the fact that Firefox depends almost entirely on its users for marketing. Mozilla provides ad banners of various sizes, a website to discuss marketing, and comes up with campaigns, but they don't have a full-on marketing department to do their own TV commercials, take out advertisements in magazines, etc. Instead they do fundraisers to buy an ad, or hold video contests.
The result is a strange chimera that's part low-budget advertising, part grass-roots campaign. (I wouldn't call it astroturfing, since that would mean Mozilla was mobilizing a small group to impersonate a larger group, and I see no indication of that.) On one hand, with lots of volunteers, you get people with lots of different skills -- it's very open source in that regard. But that also means you have lots of people who aren't trained in advertising, so yes, you're going to get "third-class promotional tactics."
Unfortunately, it also means you get people who prefer to engage in flame wars, or set up their website to block users of other browsers, etc. Spread Firefox actively tries to discourage these tactics, but they do happen, and I think they're more of a concern than any "controversy" over download numbers.
I'd say they aren't meaningless, they just don't mean what some people want them to mean.
If the numbers continue to go up at a comparable rate, and you have any indication that it's not just people upgrading or DoS bots repeatedly downloading the installers, then it indicates continued interest in the product. It means people are still trying it out. You can assume from past experience that at least some of those people will stick around.
Sure, it doesn't tell you how many people are actually using it, but the nature of the web is such that you can't count users accurately, and people will always contest your claims because the stats in their region/audience/etc. disagree. Something like "X copies were downloaded from our official mirrors" is pretty unambiguous.
IMO, as long as they're not trying to claim that the download count means something it doesn't (like users) and the numbers aren't inflated, I see no problem with it appearing in the PR.
While we're at it, what makes them think that "only a meteorite" could have triggered a sonic boom in the area?
Not that meteorites are uncommon, it's just that I tend to see red flags anytime someone pops up with "X is the only explanation for Y" rather than "X appears to explain Y."
Interesting that all this vitriol is directed at AOL, yet Yahoo! is planning on using the same service.
As far as I can tell, the main difference seems to be that Yahoo only plans to use them for transactional emails -- order confirmations and the like -- while AOL hasn't gone into much detail on that issue.
Depends entirely on what companies get onto the list. According to Goodmail, they'll only certify mailers who stick to opt-in list management. They explicitly prohibit list-sharing, harvesting, etc. The only drawback is that while they require lists to be opt-in, they don't require confirmed opt-in. That could be a potential loophole, but it's still stricter than CAN-SPAM.
Assuming Goodmail does their job correctly, the result is a net decrease in false positives, not a net increase in spam.
Everyone is forgetting buisnesses like say EBAY and PAYPAL. I am sure just the shear size that they have a big enough aol email list that aol will want money from them as well.
They paid for Bonded Sender, so I see no reason they wouldn't pay for Goodmail.
Of course, if they don't care about getting that extra "Certified!" icon, they can stick with the whitelists AOL already has. But considering that mail that's really from Ebay would have that seal, and phishing mails wouldn't, I can imagine they might be interested.
If it doesn't work, they will lose customers, if it works, I doubt I'll ever see a story about its success on Slashdot.
Absolutely correct. Hotmail's deal with Bonded Sender seems to have worked fine, despite all the fuss when it was announced. It's been almost two years, and I don't think I've seen a single mention on Slashdot since the "Hotmail will allow pay-to-spam!!!!11111one" stories.
I mean if mass-emailers (spammers) are not, in fact, going to be able to pay to insure that their targets get their emails, then why are mass-emailers going to pay for this privelege?
You're confusing mass-mailers with spammers. All spammers are mass-mailers, but not all mass-mailers are spammers.
A mass-mailer who practices responsible list management -- i.e. only sends mail to people who opt in, and verifies that in some way (so you can't sign up someone else), and responds to unsubscribe requests in a timely fashion, and doesn't change things out from under you (like collecting a subscriber list for one purpose and then using it for another) is not a spammer by any reasonable definition.
They're the target audience. They're rightly concerned about false positives, because automatically classifying email into spam and non-spam categories is a difficult task, prone to a noticeable error rate. This way they get to bypass the filters and get a "Certified!" icon in the mail client. (And I think AOL will let them show images in the message by default.)
If Goodmail does its job right, when a spammer tries to sign up for the service, they'll find that in a background check and refuse to let them sign up.
So certified mail allows senders to pay for what privilege, again?
As I understand it, AOL's mail client will display images in certified mail by default (rather than hiding them by default) and will add some sort of icon to the UI to indicate that it's been "certified."
Anyone who pays AOL to send me a certified email has just got to be someone I don't want to talk to.
I heard the same arguments two years ago when Hotmail started using Bonded Sender. Of course, given that sites like Ebay were signed up with Bonded Sender, that would mean not getting your outbid notices. And with some of the names I've seen attached this time around, blocking certified email could be a good way to filter out any real messages from your bank so that you only get the phish.
Actually, the entire controversy has a sense of deja vu about it. I guess it just goes to show two things: (1) People have short memories. (2) Bonded Sender didn't cause the massive problems predicted for Hotmail, or people would remember it anyway.
You do realize that they're adding a whitelist to bypass filters, not blocking mail that isn't on the list, right?
This is the equivalent of opening a VIP lane and leaving the other lanes unchanged -- not requiring people to sign up for the VIP list to get in in the first place.
As long as MoveOn and other organizations practice responsible mailing list management, their delivery will be unchanged from the way it is today. So they're not fighting what they think they're fighting.
This is a whitelist that bypasses filters, not a whitelist that is the only way to get through. Bulk mailers who don't pay up will still be able to send to AOL, and can still participate in AOL's other whitelists.
And Goodmail's service isn't a matter of "pay and we'll let you in" so much as it's "pay and we'll do a background check to see if you're a spammer, and if you pass our criteria we'll put you in the fast lane." Hmm, that sounds a lot like Bonded Sender and Habeas. Remember the controversy here on/. when Hotmail started using Bonded Sender two years ago? How exactly did that play out?
But why let the facts get in the way of a good knee-jerk reaction? We like placing AOL as the big corporate enemy. They often are, of course, but in this case? It's all overreaction and misinformation stemming from mistakes in the initial press.
Check out some of the commentary at Planet Antispam to get some views from the anti-spam community. You'll be surprised to find most of them siding with AOL on this one.
People who can quote baseball statistics out of their heads are just as geeky but THEY don't get made fun of.
Agreed. Somehow society has decided it's OK to set up a fantasy football league, but not to get into a debate about the Hulk vs. Superman. It's OK to paint your face blue or wear a giant piece of cheese on your head when you go to watch a game five or more times a season, but it's not OK to paint yourself blue and dress up as a Farscape character when you go to a convention to meet other science-fiction fans once a year.
One obsession is accepted as just some sports fan. The other gets told to get a life.
The talk of encryption is what worries me. Given that it's regularly used for secure remote access (SSH), used for secure communications (S/MIME and PGP), and essential to commerce over the internet (SSL), I'd expect there's quite a bit of legitimate encrypted traffic flying around already.
Sure, it's buried amid the flood of email (80% or more of which is spam), web traffic, and P2P traffic. But encryption isn't a rare thing mostly used by bad guys, as the article suggests.
The attitude reminds me of one of the five or so episodes of Enterprise I saw, in which T'Pol got an letter from home and the crew spent the whole episode trying to decrypt it. The theme was very anti-privacy, with one of the characters actually saying to her, "Do you know how suspicious that looked?" It made as much sense as claiming that closed curtains were a challenge to look inside.
I'd guess that even without encrypted torrents, most encrypted traffic on the net is business traffic of one sort or another. So the bad guys using encryption are already lost in the noise.
Nanotech just refers to anything dealing with really really small technology. Grey goo is only an issue with self-replicating machines -- often referred to as Von Neumann machines (or possibly other machines that alter the environment around them). The fear there is that microscopic self-replicating machines could, molecule by molecule, transform everything around them into more self-replicating machines.
If the nanotech doesn't alter matter around it, grey goo isn't a risk.
Meanwhile, charity groups, e-zines, and other legitimate free mailing lists that people sign up for will be screwed.
How?
No, really, how?
Where has AOL said that people who don't sign up for this list will be blocked?
This is a whitelist that bypasses filters, not a whitelist that is the only way to get through. Bulk mailers who don't pay up will still be able to send to AOL. Their mail will be subject to more scrutiny, sure. It'll be subject to as much scrutiny as... well, as it is today.
And as I recall, the Certified Email whitelist isn't a matter of "pay and we'll let you in" so much as it's "pay and we'll do a background check to see if you're a spammer, and if you pass our criteria we'll put you in the fast lane." Why, that sounds like Bonded Sender and Habeas. Remember the controversy here on/. when Hotmail started using Bonded Sender two years ago? How exactly did that play out?
Of course, there are easier ways to trick people into landing on your site. Apparently just taking out an ad for your site with the keywords "download firefox" is enough to confuse some people and get them to think you're the official Firefox download site.
This, I think, lends a bit more credence to Jakob Nielsen's anti-search diatribe earlier this year.
All that said, I agree with your point that in the long run, it's easier and more effective to write good content and do the necessary promotion than it is to try to cheat the system.
People will set up a site that just aggregates RSS feeds based on a few keywords, use those keywords as tags, then ping Technorati. This clutters up search results until someone at Technorati notices the site is junk, or someone reports it to Technorati as junk.
The odd thing is, the keywords they choose aren't always what you'd expect to be popular spam terms. I've found some of my own posts showing up in scraped sites about things as mundane as coffee.
Slashdot Mirror
Then you get into other sci-fi fandoms, like Farscape, where the women seem to outnumber the men.
Show me a standards-compliant page that renders differently in Opera.
:first-line in Opera 8. Fully validating code, but styles would persist incorrectly and, in one case, text would disappear. One's fixed in Opera 9 TP2, the other is still around. Sure, this was real edge-case stuff, but it happens.
It all depends on which subset of the standards you use.
A while back, I ran into some bugs with
Here's a more practical example going the other way: creating frames with generated content. It works handily in Opera, not at all in IE, and can be made to almost-work in Firefox (it comes down to positioning on generated content based on different versions of the CSS spec).
Standards compliance in a browser isn't a matter of working down a checklist from item #1 to item #1000, and counting complaince based on whether you get to item #800, #900, or #999. You look through the standards and prioritize. Maybe both browsers fulfill 900 of those items -- but maybe item #723 is only implemented in one of them, because one browser picked it as a priority, and the other put it off.
Stepping away from standards-compliance for a moment, you also have to take into account things like XMLHTTPRequest, contentEditable, and so on. Most rich text editors (popular in webmail apps) couldn't run on Opera until they released the Opera 9 previews, because the required features weren't there. And every minor version of Opera from 8 on has made more adjustments for AJAX apps.
Even within standards -- both de facto and de jure -- you need to find a common subset. Validating the code doesn't guarantee that other browsers will behave the way you expect.
Plus there's the fact that Firefox depends almost entirely on its users for marketing. Mozilla provides ad banners of various sizes, a website to discuss marketing, and comes up with campaigns, but they don't have a full-on marketing department to do their own TV commercials, take out advertisements in magazines, etc. Instead they do fundraisers to buy an ad, or hold video contests.
The result is a strange chimera that's part low-budget advertising, part grass-roots campaign. (I wouldn't call it astroturfing, since that would mean Mozilla was mobilizing a small group to impersonate a larger group, and I see no indication of that.) On one hand, with lots of volunteers, you get people with lots of different skills -- it's very open source in that regard. But that also means you have lots of people who aren't trained in advertising, so yes, you're going to get "third-class promotional tactics."
Unfortunately, it also means you get people who prefer to engage in flame wars, or set up their website to block users of other browsers, etc. Spread Firefox actively tries to discourage these tactics, but they do happen, and I think they're more of a concern than any "controversy" over download numbers.
I'd say they aren't meaningless, they just don't mean what some people want them to mean.
If the numbers continue to go up at a comparable rate, and you have any indication that it's not just people upgrading or DoS bots repeatedly downloading the installers, then it indicates continued interest in the product. It means people are still trying it out. You can assume from past experience that at least some of those people will stick around.
Sure, it doesn't tell you how many people are actually using it, but the nature of the web is such that you can't count users accurately, and people will always contest your claims because the stats in their region/audience/etc. disagree. Something like "X copies were downloaded from our official mirrors" is pretty unambiguous.
IMO, as long as they're not trying to claim that the download count means something it doesn't (like users) and the numbers aren't inflated, I see no problem with it appearing in the PR.
While we're at it, what makes them think that "only a meteorite" could have triggered a sonic boom in the area?
Not that meteorites are uncommon, it's just that I tend to see red flags anytime someone pops up with "X is the only explanation for Y" rather than "X appears to explain Y."
Yes, block all email from the American Red Cross, including donation receipts.
I wonder how popular that'll be.
Interesting that all this vitriol is directed at AOL, yet Yahoo! is planning on using the same service.
As far as I can tell, the main difference seems to be that Yahoo only plans to use them for transactional emails -- order confirmations and the like -- while AOL hasn't gone into much detail on that issue.
So that's a net gain of spam.
Depends entirely on what companies get onto the list. According to Goodmail, they'll only certify mailers who stick to opt-in list management. They explicitly prohibit list-sharing, harvesting, etc. The only drawback is that while they require lists to be opt-in, they don't require confirmed opt-in. That could be a potential loophole, but it's still stricter than CAN-SPAM.
Assuming Goodmail does their job correctly, the result is a net decrease in false positives, not a net increase in spam.
Everyone is forgetting buisnesses like say EBAY and PAYPAL. I am sure just the shear size that they have a big enough aol email list that aol will want money from them as well.
They paid for Bonded Sender, so I see no reason they wouldn't pay for Goodmail.
Of course, if they don't care about getting that extra "Certified!" icon, they can stick with the whitelists AOL already has. But considering that mail that's really from Ebay would have that seal, and phishing mails wouldn't, I can imagine they might be interested.
Yes, they have told me they will be charging for "bulk" mailers.
Did they say they were going to flat-out charge bulk mailers to send to AOL?
Or did they say there were going to charge bulk mailers to get onto this list?
Even if an airline charges extra for first-class seats, people in coach still make it to their destination.
Let me ask them, "So why are you dropping the enhanced whitelist?"
Um... they're not. Early reports said that they were, but those turn out to be incorrect.
If it doesn't work, they will lose customers, if it works, I doubt I'll ever see a story about its success on Slashdot.
Absolutely correct. Hotmail's deal with Bonded Sender seems to have worked fine, despite all the fuss when it was announced. It's been almost two years, and I don't think I've seen a single mention on Slashdot since the "Hotmail will allow pay-to-spam!!!!11111one" stories.
I mean if mass-emailers (spammers) are not, in fact, going to be able to pay to insure that their targets get their emails, then why are mass-emailers going to pay for this privelege?
You're confusing mass-mailers with spammers. All spammers are mass-mailers, but not all mass-mailers are spammers.
A mass-mailer who practices responsible list management -- i.e. only sends mail to people who opt in, and verifies that in some way (so you can't sign up someone else), and responds to unsubscribe requests in a timely fashion, and doesn't change things out from under you (like collecting a subscriber list for one purpose and then using it for another) is not a spammer by any reasonable definition.
They're the target audience. They're rightly concerned about false positives, because automatically classifying email into spam and non-spam categories is a difficult task, prone to a noticeable error rate. This way they get to bypass the filters and get a "Certified!" icon in the mail client. (And I think AOL will let them show images in the message by default.)
If Goodmail does its job right, when a spammer tries to sign up for the service, they'll find that in a background check and refuse to let them sign up.
So certified mail allows senders to pay for what privilege, again?
As I understand it, AOL's mail client will display images in certified mail by default (rather than hiding them by default) and will add some sort of icon to the UI to indicate that it's been "certified."
Anyone who pays AOL to send me a certified email has just got to be someone I don't want to talk to.
I heard the same arguments two years ago when Hotmail started using Bonded Sender. Of course, given that sites like Ebay were signed up with Bonded Sender, that would mean not getting your outbid notices. And with some of the names I've seen attached this time around, blocking certified email could be a good way to filter out any real messages from your bank so that you only get the phish.
Actually, the entire controversy has a sense of deja vu about it. I guess it just goes to show two things: (1) People have short memories. (2) Bonded Sender didn't cause the massive problems predicted for Hotmail, or people would remember it anyway.
You do realize that they're adding a whitelist to bypass filters, not blocking mail that isn't on the list, right?
This is the equivalent of opening a VIP lane and leaving the other lanes unchanged -- not requiring people to sign up for the VIP list to get in in the first place.
No, it means they only certify mailers who practice opt-in list management.
Spam kings who chip in won't get certified because they aren't using opt-in lists.
Average joe mailing lists get the same treatment they receive today.
As long as MoveOn and other organizations practice responsible mailing list management, their delivery will be unchanged from the way it is today. So they're not fighting what they think they're fighting.
/. when Hotmail started using Bonded Sender two years ago? How exactly did that play out?
This is a whitelist that bypasses filters, not a whitelist that is the only way to get through. Bulk mailers who don't pay up will still be able to send to AOL, and can still participate in AOL's other whitelists.
And Goodmail's service isn't a matter of "pay and we'll let you in" so much as it's "pay and we'll do a background check to see if you're a spammer, and if you pass our criteria we'll put you in the fast lane." Hmm, that sounds a lot like Bonded Sender and Habeas. Remember the controversy here on
But why let the facts get in the way of a good knee-jerk reaction? We like placing AOL as the big corporate enemy. They often are, of course, but in this case? It's all overreaction and misinformation stemming from mistakes in the initial press.
Check out some of the commentary at Planet Antispam to get some views from the anti-spam community. You'll be surprised to find most of them siding with AOL on this one.
isn't this particular thread entitled "Encryption"? you know? the post? anyone got anything to say about the bittorrent article? anyone?
I'd been wondering about that. The Star Trek reference was kind of an afterthought, but that seems to be the main thing people are responding to.
At least we know people have their priorities straight...
People who can quote baseball statistics out of their heads are just as geeky but THEY don't get made fun of.
Agreed. Somehow society has decided it's OK to set up a fantasy football league, but not to get into a debate about the Hulk vs. Superman. It's OK to paint your face blue or wear a giant piece of cheese on your head when you go to watch a game five or more times a season, but it's not OK to paint yourself blue and dress up as a Farscape character when you go to a convention to meet other science-fiction fans once a year.
One obsession is accepted as just some sports fan. The other gets told to get a life.
The talk of encryption is what worries me. Given that it's regularly used for secure remote access (SSH), used for secure communications (S/MIME and PGP), and essential to commerce over the internet (SSL), I'd expect there's quite a bit of legitimate encrypted traffic flying around already.
Sure, it's buried amid the flood of email (80% or more of which is spam), web traffic, and P2P traffic. But encryption isn't a rare thing mostly used by bad guys, as the article suggests.
The attitude reminds me of one of the five or so episodes of Enterprise I saw, in which T'Pol got an letter from home and the crew spent the whole episode trying to decrypt it. The theme was very anti-privacy, with one of the characters actually saying to her, "Do you know how suspicious that looked?" It made as much sense as claiming that closed curtains were a challenge to look inside.
I'd guess that even without encrypted torrents, most encrypted traffic on the net is business traffic of one sort or another. So the bad guys using encryption are already lost in the noise.
Nanotechnology != Nanites.
Nanotech just refers to anything dealing with really really small technology. Grey goo is only an issue with self-replicating machines -- often referred to as Von Neumann machines (or possibly other machines that alter the environment around them). The fear there is that microscopic self-replicating machines could, molecule by molecule, transform everything around them into more self-replicating machines.
If the nanotech doesn't alter matter around it, grey goo isn't a risk.
Meanwhile, charity groups, e-zines, and other legitimate free mailing lists that people sign up for will be screwed.
/. when Hotmail started using Bonded Sender two years ago? How exactly did that play out?
How?
No, really, how?
Where has AOL said that people who don't sign up for this list will be blocked?
This is a whitelist that bypasses filters, not a whitelist that is the only way to get through. Bulk mailers who don't pay up will still be able to send to AOL. Their mail will be subject to more scrutiny, sure. It'll be subject to as much scrutiny as... well, as it is today.
And as I recall, the Certified Email whitelist isn't a matter of "pay and we'll let you in" so much as it's "pay and we'll do a background check to see if you're a spammer, and if you pass our criteria we'll put you in the fast lane." Why, that sounds like Bonded Sender and Habeas. Remember the controversy here on
But looking at it logically like this gets in the way of a good knee-jerk reaction.
Of course, there are easier ways to trick people into landing on your site. Apparently just taking out an ad for your site with the keywords "download firefox" is enough to confuse some people and get them to think you're the official Firefox download site.
This, I think, lends a bit more credence to Jakob Nielsen's anti-search diatribe earlier this year.
All that said, I agree with your point that in the long run, it's easier and more effective to write good content and do the necessary promotion than it is to try to cheat the system.
People will set up a site that just aggregates RSS feeds based on a few keywords, use those keywords as tags, then ping Technorati. This clutters up search results until someone at Technorati notices the site is junk, or someone reports it to Technorati as junk.
The odd thing is, the keywords they choose aren't always what you'd expect to be popular spam terms. I've found some of my own posts showing up in scraped sites about things as mundane as coffee.