Joi Ito made a post to his weblog yesterday where his iTunes post received a TrackBack ping that when he checked the site did not link back or mention his post. So, maybe I feel a bit hurt, but nothing illegal going on here,
he writes. Obviously it makes sense to try to direct people to more information about a topic and sending a trackback to an entry about the same topic makes sense. It just felt weird. I had been looking at trackback more as a two-way thing, but I guess they are technically one way.
He asks for thoughts and an excellent thread ensues.
Being a proponent of TrackBack and the exploration of RESTful interfaces I jumped in with my commentary. (This is a reworking and expansion of that post.)
The issue of legitimate one-way TrackBacking backing has come up on a few occasions in the past. (Take this thread on Sam Ruby's site for instance.) I documented a legitimate use of one-way TrackBack pinging around that time also. In that scenario I had read a post by Steven Noels where he documented two different ways of encoding HTML in RSS and asked Now, what should be the good practice?
I had written such a piece months earlier where I offered my opinion on the topic so I went into MT and pinged his entry with that post. It doesn't mention Steve or link to him because I had written that post months earlier. That ping was, in my opinion, highly appropriate because it answered the question he asked and continued that conversation. (I just so happened to have done it months earlier since I am a Master Jedi that is one with The Force.)
Unlike the destruction of Usenet in which Kevin Marks raised the specter of, we as publishers have the ability to control and managed the situation with TrackBack and wikis. We will have to (and should) actively maintain the quality of such things. I'd advocate deleting low quality or spam pings. As a publisher of a weblog that uses TrackBack for post-to-post linking, I have the right to delete a ping I don't think furthers the conversation. I've deleted pings because they just lead to a page with a link back to the post they where pinging. Christine
posted a similar comment while I was writing the same thing. (Good karma I suppose.)
Its not like you are deleting their comment where that content is gone – you are deleting the link, the content lives on. If the job of managing pings is too big (ala Slashdot) then perhaps a karma points/ratings system could be layered on top of TrackBack. Right now I don't see anyone receiving that many pings in their weblog.
There was discussion of how new two-way mediums, such as TrackBack or wikis (or mailing lists? – momentary sarcasm), could be abused by spammers though generally they have not. In some sense I agree with the notion that TrackBack and any other forms of the two-way web are ripe to be abused. Isaac
noted manually or programmatically setting off a TrackBack ping is easy – its just an HTTP POST to a URI. (RESTful Web services at their greatest.)
The bottom line is that we will eventually need to a mechanism such as IP blocking or other optional means of filtering rubbish like any other interactive medium. I don't see the specter of abuse being a reason not to move forward if something like TrackBack is useful and effective in some context.
With that the thread moved onto to discussing authentication systems with Bill Kearney raising his suggestion of using PKI. Not a bad idea, but as Bill notes unless we start using them we won't be able to grapple with the next layers of complexity.
It has to get a lot easier to understand and use. At the same time the next layer of complexity seems premature at best. We need to give it some time to see what emerges.
<p>Joi Ito made <a href="http://joi.ito.com/archives/2003/04/29/one_way_trackbacking.html">a post</a> to his weblog yesterday where his iTunes post received a TrackBack ping that when he checked the site did not link back or mention his post. <q>So, maybe I feel a bit hurt, but nothing illegal going on here,</q> he writes. <q>Obviously it makes sense to try to direct people to more information about a topic and sending a trackback to an entry about the same topic makes sense. It just felt weird. I had been looking at trackback more as a two-way thing, but I guess they are technically one way.</q> He asks for thoughts and an excellent thread ensues.</p>
<p>Being a proponent of TrackBack and the exploration of RESTful interfaces I jumped in with my commentary. (This is a reworking and expansion of that post.)</p>
<p>The issue of legitimate one-way TrackBacking backing has come up on a few occasions in the past. (Take <a href="http://intertwingly.net/blog/1186.html">this thread</a> on Sam Ruby's site for instance.) I documented <a href="http://www.intertwingly.net/blog/1184.html#c1044729239">a legitimate use of one-way TrackBack pinging</a> around that time also. In that scenario I had read <a href="http://blogs.cocoondev.org/stevenn/archives/000602.html">a post by Steven Noels</a> where he documented two different ways of encoding HTML in RSS and asked <q>Now, what should be the good practice?</q> I had written such a piece months earlier where I offered <a href="http://www.timaoutloud.org/archives/000094.html">my opinion on the topic</a> so I went into MT and pinged his entry with that post. It doesn't mention Steve or link to him because I had written that post months earlier. That ping was, in my opinion, highly appropriate because it answered the question he asked and continued that conversation. (I just so happened to have done it months earlier since I am a Master Jedi that is one with <a href="http://www.starwars.com/databank/organization/thejediorder/index.html">The Force</a>.) </p>
<p>Unlike the destruction of Usenet in which Kevin Marks raised the specter of, we as publishers have the ability to control and managed the situation with TrackBack and wikis. We will have to (and should) actively maintain the quality of such things. I'd advocate deleting low quality or spam pings. As a publisher of a weblog that uses TrackBack for post-to-post linking, I have the right to delete a ping I don't think furthers the conversation. I've deleted pings because they just lead to a page with a link back to the post they where pinging. <a href="http://joi.ito.com/archives/2003/04/29/one_way_trackbacking.html#n006368"><q>Christine</q> posted a similar comment</a> while I was writing the same thing. (Good karma I suppose.) </p>
<p>Its not like you are deleting their comment where that content is gone – you are deleting the link, the content lives on. If the job of managing pings is too big (ala Slashdot) then perhaps a karma points/ratings system could be layered on top of TrackBack. Right now I don't see anyone receiving that many pings in their weblog.</p>
<p>There was discussion of how new two-way mediums, such as TrackBack or wikis (or mailing lists? – momentary sarcasm), could be abused by spammers though generally they have not. In some sense I agree with the notion that TrackBack and any other forms of the two-way web are ripe to be abused. <q>Isaac</q> <a href="http://joi.ito.com/archives/2003/04/29/one_way_trackbacking.html#n006353">noted</a> manually or programmatically setting off a TrackBack ping is easy – its just an HTTP POST to a URI. (RESTful Web services at their greatest.)</p>
<p>The bottom line is that we will eventually need to a mechanism such as IP blocking or other optional means of filtering rubbish like any other interactive medium. I don't see the specter of abuse being a reason not to move forward if something like TrackBack is useful and effective in some context.</p>
<p>With that the thread moved onto to discussing authentication systems with Bill Kearney <a href="http://joi.ito.com/archives/2003/04/29/one_way_trackbacking.html#n006401">raising his suggestion of using PKI</a>. Not a bad idea, but as Bill notes <q>unless we start using them we won't be able to grapple with the next layers of complexity.</q> It has to get a lot easier to understand and use. At the same time the next layer of complexity seems premature at best. We need to give it some time to see what emerges.</p>
Leave a comment