no one uses them, if they did, search engines would ignore them

Posted Jun 22, 2014 14:43 UTC (Sun) by khim (subscriber, #9252)
In reply to: no one uses them, if they did, search engines would ignore them by anselm
Parent article: Microformats turn 9 years old

Easy. If people will actually start actively using such microformats then spammers could produce legitimately-looking pages of various companies (with correct address information presented on pages!) which will send users to spammer-provided location only when they will try to actually use entry from their address book.

Microformats could only exist in obscurity or in controlled conditions (e.g. on the intranet). Otherwise they will be exploited sooner or later.

no one uses them, if they did, search engines would ignore them

Posted Jun 22, 2014 15:56 UTC (Sun) by anselm (subscriber, #2796) [Link] (19 responses)

Unlikely. I would expect a »microformat address extractor« to present the extracted address record to me before actually adding it to the address book, with an opportunity to delete or edit any field, or to reject the record outright. Obviously doctored URLs and such wouldn't make it into actual address books. (The extractor could also check blacklists like Spamhaus DBL to see whether the domains of any of the URLs in the record show up there, and emit suitable warnings or reject a record automatically.)

This is a quality-of-implementation issue on the part of software that deals with microformats. By your reasoning, e-mail should have been abolished long ago because most e-mail messages are spam. Yet strangely, for many of us e-mail is still a useful resource.

no one uses them, if they did, search engines would ignore them

Posted Jun 22, 2014 18:46 UTC (Sun) by bronson (subscriber, #4806) [Link] (7 responses)

> I would expect a »microformat address extractor« to present the extracted address record to me before actually adding it to the address book

I remember people trying that a decade ago. It's too much work -- you need easy-to-use software to discover the record, present it without irritating the user, correct it, and save it somewhere useful. And it must be upgraded as the formats evolve. It sounds great in theory but the end result is always a mess that only appeals to the tiny nerd demographic.

I've got to admit, I'm surprised that not even the calendar event microformat has caught on. That seems so simple and so useful that it's almost cheating. But nope, basically zero uptake.

no one uses them, if they did, search engines would ignore them

Posted Jun 22, 2014 22:50 UTC (Sun) by anselm (subscriber, #2796) [Link] (6 responses)

This is really a chicken/egg problem. Few people bother to publish data using microformats because there isn't a lot of software that uses them. On the other hand, few people feel the need to write software that uses microformats because they are so little used.

I don't think we need to appeal to the »nerd factor« to explain why we don't see more microformat-enabled web sites.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 2:02 UTC (Mon) by bronson (subscriber, #4806) [Link] (5 responses)

Well, almost everything on the network starts as a chicken/egg problem (email, Facebook, RSS, HTML5, JavaScript, etc etc etc).

When there's demand, it's no problem. When there's very little demand, though...

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 8:10 UTC (Mon) by anselm (subscriber, #2796) [Link] (4 responses)

Fair enough.

However, let's not forget that the initial question was how a SEO scammer would actually game microformats such as vcard information, to which khim posted a frankly ridiculous and easily refuted answer.

So suppose for a moment that microformats were indeed popular and widespread. How would a spammer arrange to, say, subvert an event microformat when (a) getting data from the page with the event on it into someone's calendar requires an explicit action (unlike spam e-mail, where the spam shows up in your inbox without an explicit action on your part), and (b) the microformat extraction tool implements reasonable and straightforward safeguards similar to those found in most web browsers or MUAs (e.g., the tool would enforce that an event listing in a microformat can only contain URLs with the same domain as the page it is on)?

no one uses them, if they did, search engines would ignore them

Posted Jun 24, 2014 6:41 UTC (Tue) by bronson (subscriber, #4806) [Link] (3 responses)

It's a fair question and I don't have a good answer. I would guess it depends on the user experience of the microformat tools. They've been pretty terrible so far but maybe they just haven't received much attention.

Now a bit of speculatin' on the problems that a good, general purpose microformat tool might have to overcome...

First problem might be that people will click absolutely anything with an OK button. When adding calendar events that's no big deal (worst case, alarm at 3:30 in the morning: Buy \/ia9ra!) Other formats would need to be more careful, but how do you do that without harming convenience and usability?

Another problem, seems like users would probably get desensitized to notifications. (on every page: "Add Slate to your address book!" "Add Slate to your address book!") This sounds like a problem like browser popups: only solution is to keep tweaking heuristics until you find an adequate compromise between advertisers and users.

Also, scope. How does one microformat tool integrate with all your other applications? Contacts (email, mail, phone, IM, etc), calendar, social graphs, outlines, CV, and all the other ones? Seems like it's got to be built into the browser?

Those seem fairly insurmountable to me. But then again, so did Linux graphics drivers, Wikipedia, and YouTube's business model. I'd be more than happy to be proven wrong.

no one uses them, if they did, search engines would ignore them

Posted Jun 24, 2014 9:53 UTC (Tue) by krake (guest, #55996) [Link] (2 responses)

> Other formats would need to be more careful, but how do you do that without harming convenience and usability?

It could be something unobstrusive, like the RSS icon that appears when there is an RSS source available.
E.g. a "Contact" icon appearing when there is an embedded contact information.

This way the user would initiate the data transfer.

> Seems like it's got to be built into the browser?

I guess that makes most sense, the difficulty would be to find some kind of plugin API so that browsers could easily hook into the paltform's data services.

no one uses them, if they did, search engines would ignore them

Posted Jun 24, 2014 18:26 UTC (Tue) by bronson (subscriber, #4806) [Link] (1 responses)

That icon has been widely blamed for the death of RSS. (citations are easy to google, here's the first: http://camendesign.com/blog/rss_is_dying ).

no one uses them, if they did, search engines would ignore them

Posted Jun 25, 2014 6:58 UTC (Wed) by krake (guest, #55996) [Link]

Well, any UI that is reduced to just an icon is less obvious than a more verbose one, yet we do have them.

They are a trade-off between spacial needs and clarity.

I find the RSS icon useful, I've added many feeds to my reader that way.

But such an icon is just one option, if there are better unobstrusive ways to make the user aware of the additional content then they are obviously also valid and probably preferably choices.

This was just an example that enabling users to access additional content does not require interrupting their present task via a popup.

no one uses them, if they did, search engines would ignore them

Posted Jun 22, 2014 19:29 UTC (Sun) by khim (subscriber, #9252) [Link] (10 responses)

Obviously doctored URLs and such wouldn't make it into actual address books.

They wouldn't make in your address book, sure. But people like you and me are insignificant minority. Even if only just half of users will add the spoofed address it'll still be a will.

I would expect a »microformat address extractor« to present the extracted address record to me before actually adding it to the address book, with an opportunity to delete or edit any field, or to reject the record outright.

What will it change? People will just not look there!

By your reasoning, e-mail should have been abolished long ago because most e-mail messages are spam. Yet strangely, for many of us e-mail is still a useful resource.

E-mail is kept alive by it's network effect. New service with such a huge percentage of spam will have no chance. We've seen it happening with Jabber: Google tried to “make it work” for years but eventually gave up since most users who tried to talk with GTalk users using not official client but something else were spammers.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 9:32 UTC (Mon) by zlynx (guest, #2285) [Link] (1 responses)

> E-mail is kept alive by it's network effect. New service with such a huge percentage of spam will have no chance. We've seen it happening with Jabber: Google tried to “make it work” for years but eventually gave up since most users who tried to talk with GTalk users using not official client but something else were spammers.

Really? I suppose that this could be true.

I rather think that the majority of users were using multi-protocol chat clients like Pidgin, Empathy, Adium, etc.

Out of the people I know personally, the only official GTalk client we use is the one included on Android phones. *Everyone* else uses a unified IM client.

How else would we talk to people on Yahoo, AOL, MSN and Google without running multiple chat clients? Which is just silly.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 14:05 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link]

> How else would we talk to people on Yahoo, AOL, MSN and Google without running multiple chat clients?
The answer is "we don't", of course.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 16:07 UTC (Mon) by krake (guest, #55996) [Link] (7 responses)

> We've seen it happening with Jabber: Google tried to “make it work” for years but eventually gave up since most users who tried to talk with GTalk users using not official client but something else were spammers.

What?
How on earth would that possibly work?

I've been using Jabber for years, including two gmail accounts and I have not received a single spam messsage. ever.

With email I receive several on each account every day.

Since I have never used GTalk I can only assume that this is a flaw of that particular client software.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 18:21 UTC (Mon) by raven667 (subscriber, #5198) [Link] (6 responses)

The thing which is being specifically referred to is the loss of XMPP/Jabber federation with the move from Google Talk to Hangouts, even though both systems use Jabber. Google used to be the largest destination for federated access, where you set up Jabber on your own server but users on your server can add Google users as buddies because Google accepted server-to-server communications. This was abused by spammers who would set up their own servers and send spammy messages to Google Jabber users. SMTP mail is just too big to shut down and de-federate like that but XMPP was not.

I don't think federation became super popular because users are used to using multiple clients, or multi-protocol clients, and having accounts on each IM system they use, the problem has been worked around sufficiently that it is in no ones particular interest to consolidate more even if that would be better for the system as a whole, like a Nash Equilibrium.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 18:53 UTC (Mon) by krake (guest, #55996) [Link] (2 responses)

Aside from federation still working (four of my currently online contacts who are on gmail accounts are reachable from by non-gmail account), this is not what I was referring to.

I was puzzled by the claim that there was rampant spam on Jabber.
In my 6 years of using Jabber (probably longer, the current client's logs go back to 2008) I've never received a single spam message on any of the accounts on any of the servers (including gmail).

So I was wondering if that was somehow a flaw in the GTalk client the other poster referred to.

Sure, I could have been just very lucky but given the amount of spam I've received in the same time frame via email, I don't consider that very likely.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 23:01 UTC (Mon) by mathstuf (subscriber, #69389) [Link] (1 responses)

I've noticed that my G+ contacts use an obsfuscated Jabber ID and still works from Xabber and finch. It's not the email address, but some alphanumeric string @talk.google.com. I think federation might still work, but to find the address to add, you need to be in the network (at least to some extent).

no one uses them, if they did, search engines would ignore them

Posted Jun 24, 2014 7:10 UTC (Tue) by krake (guest, #55996) [Link]

I haven't tried adding anyone new, but the already established contacts work without change.

Good to know though, thanks!

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 23:09 UTC (Mon) by dlang (guest, #313) [Link] (2 responses)

google did not de-federate, the jabber community decided to mandate server-server encryption and google decided not to implement this, so any part of the community that followed the encryption-only suggestion de-federated google.

no one uses them, if they did, search engines would ignore them

Posted Jun 24, 2014 2:35 UTC (Tue) by raven667 (subscriber, #5198) [Link] (1 responses)

I dunno, maybe I misremembered

https://www.eff.org/deeplinks/2013/05/google-abandons-ope...

no one uses them, if they did, search engines would ignore them

Posted Jun 24, 2014 5:57 UTC (Tue) by dlang (guest, #313) [Link]

up until a few weeks ago the XMPP interface and federation were still running (as far as I know, and based on other comments here, they still are), but most of the servers that they were talking to switched to encrypted-only communication, and since Google didn't switch, they no longer talk.

no one uses them, if they did, search engines would ignore them

Posted Jun 23, 2014 21:27 UTC (Mon) by ballombe (subscriber, #9523) [Link]

There is a very important middle ground between the internet and the intranet: the trusted net, i.e. a set of public site that some organization
trust. For example, I trust *.debian.org hosts and they do publish RDFa records, which I use.

The fact that these records follow a well-known specification means that I can process them using generic RDFa tools instead of debian.org-specific tools.