## August 24, 2005

Ever since hep-th was founded in 1991 (or maybe a couple of years later, when it sprouted a web interface), people have talked about trying to overlay some sort of “discussion” or “commentary” layer. For understandable reasons, Paul Ginsparg has been reluctant to host any such endeavour at arxiv.org itself, wishing to maintain it purely as an e-print server.

Two and a half years ago, I started this blog, in the belief that weblogs (and weblogging software) could provide a distributed mechanism for such an overlay, without the implied endorsement and other drawbacks of a centralized service. Blogs, as you know, provide a wonderful sense of an interwoven, distributed conversation. But the missing ingredient was the ability to ‘plug into’ that conversation from the arXivs. How do you discover that I or some other physicist-blogger, has commented on a given paper?

The obvious answer is, of course … Trackbacks. I’d been intermittently bugging Ginsparg to implement Trackbacks on the arXivs for a couple of years, now. Finally, at Sidneyfest, he sidled up to me and said, “So, about this Trackback thing …”

I’m happy to report that my pestering has finally paid off. The arXiv abstract pages are now Trackback-enabled. At the bottom of every abstract page, you’ll find a link, “N trackbacks” (when N>0). Click on it, and you’ll get a list of weblog entries commenting on that paper.

So does that mean that every crackpot and trackback spammer on the internet can now get linked-to from the arXivs? Well, … no. Just as you need to be a registered author to submit papers, your weblog needs to be on an approved list, in order for your Trackbacks to appear. Going forward, the precise mechanism for getting on that list is yet to be determined. But, in the short term, the list of serious physicist-bloggers is short enough to handle by hand.

You know who you are …

Posted by distler at August 24, 2005 4:50 PM

TrackBack URL for this Entry:   http://golem.ph.utexas.edu/cgi-bin/MT-3.0/dxy-tb.fcgi/638

Weblog: iChris.ws
Excerpt: On his blog, Jacques Distler wrote of how arXiv.org e-Print archive starts to support trackbacks for individual articles. Which I think is very cool. Of course, we have systems that track paper citations, but one doesn't always want to write another paper
Tracked: August 24, 2005 11:04 PM

### Re: Trackbacks and the ArXivs

Nice effort. Maybe I could bug the Cryptology ePrint guys to implement something similar.

Posted by: Srijith on August 25, 2005 1:36 AM | Permalink | Reply to this

### Re: Trackbacks and the ArXivs

On a related note: I have seen that on CosmicVariance there is sort of an auto-trackbacking in operation, in that entries that refer to previous entries in the same blog automatically generate a trackback notice in the comment section of that previous entry. Would that be hard to implement here or, say, on the String Coffee Table?

Posted by: Urs on August 25, 2005 4:09 AM | Permalink | Reply to this

Yup. A bad habit (and not just because of trackbacks). I count 16 entries where you’ve done that.

I have seen that on CosmicVariance there is sort of an auto-trackbacking in operation, in that entries that refer to previous entries in the same blog automatically generate a trackback notice in the comment section of that previous entry.

You can always send yourself a trackback manually. It would be a poor design decision, IMHO, to automatically generate a trackback for every self-reference.

Posted by: Jacques Distler on August 25, 2005 8:08 AM | Permalink | PGP Sig | Reply to this

### More links to arXiv mirrors

Yup. A bad habit (and not just because of trackbacks). I count 16 entries where you’ve done that.

Even worse, you have 34 entries where you link to the mirror at xxx.uni-augsburg.de.

Posted by: Jacques Distler on August 30, 2005 9:51 AM | Permalink | PGP Sig | Reply to this

### Re: Trackbacks and the ArXivs

I do have one query. The arXiv’s page on the system describes a sensible sounding moderation system:

Trackbacks will not be immediately visible. Because of widespread Trackback spam we have a semi-automated editorial process that approves trackbacks for display. Trackbacks from known blogs should become visible in a few minutes, but it may take longer for us to recognize new blogs.

But your post suggests that crackpots will be blocked. In that case, it’s not clear what “known blogs” means in the arXiv’s description. Who’s right? Who are the “serious physicist-bloggers”?

Posted by: Michael Williams on August 25, 2005 8:23 AM | Permalink | Reply to this

### Moderation

Trackbacks to the arXivs are, indeed, moderated. The details are, obviously, still evolving. I expect they will continue to evolve, as more people start to avail themselves of the service. (The criteria for posting papers to the arXivs underwent a similar evolution from 1991 to the present.)

Since the service wasn’t expected to be announced publicly till next week (Sean Carroll’s eagle-eyes were not taken into account), it would not surprise me that some of the documentation could be better-worded.

All in good time …

Posted by: Jacques Distler on August 25, 2005 8:46 AM | Permalink | PGP Sig | Reply to this
Weblog: e pur si muove
Excerpt: [^] Big news in the scientific blogging community: arXiv.org, arguably the world’s largest preprint repository, now supports trackbacks. Jacques Distler has some news on how it happened. « Peter Woit I think there’s some very nice cl...
Tracked: August 25, 2005 11:39 AM
Weblog: semantics etc.
Excerpt: As reported here, here, here, here, and here, the giant open research archive arxiv.org for physics, math, and computer science now supports not just RSS feeds for newly added papers but also trackbacks, which means that if someone posts comments...
Tracked: August 25, 2005 2:21 PM
Weblog: Library clips
Excerpt: arxiv are enabling trackbacks for all papers. This is a great idea, now when you view a paper you can see who as blogged about this paper in the blogosphere…kind of like citations from other papers (like CiteBase), but blogs instead. Does this ...
Tracked: August 26, 2005 2:02 AM
Read the post At Musings today distler notes
Weblog: screaming-penguin.com
Excerpt: At Musings today distler notes how useful trackbacks are for commenting externally on web content in. . .
Tracked: August 26, 2005 7:56 PM
Read the post Last of the Summer Whine
Weblog: Upon Reflection
Excerpt: It also means weblog entries will increase in frequency again, given that I am in the office all the time. Several things are on the horizon, not the least of which is a system upgrade to Tiger at some point. Also, Movable Type 3.2 has been released,...
Tracked: August 29, 2005 12:14 PM

### Re: Trackbacks and the ArXivs

Dear Jacques,

Please allow me the liberty of sending my first impressions on the subject of backtracking in the hope that if the problems I raise from the depths of my ignorance are at all sensible, they might be the beginnings of an FAQ and might even get answers from the arXiv itself.

For quite some time, you have been linking to my readings from quant-ph page

I therefore presume that you might allow me into your category of “serious physicist-bloggers”.

If so, it might be appropriate for links to be established between my comments on papers and the original abstracts of those papers.

This brings us to the first of my questions for you:

Take my comment on quant-ph/0506083. In full, this reads

****
N.P. Landsman, “Between Classical and Quantum” quant-ph/0506082

An excellent review of a variety of ideas about the relationship between classical theories and quantum theories. Landsman begins with a well-informed discussion of the significance of such ideas in the early development of quantum mechanics and in the Copenhagen interpretation. He then turns to three relevant areas of mathematical physics and discusses theories of quantization, theories of classical limits, and some aspects of infinite system theories. I think his descriptions of these difficult but important theories are of exemplary clarity. He concludes with some remarks about decoherence and consistent histories.
****

Now, what do you see as the purpose of the backtracking system in relation to this comment?

Is it merely to make a link from Landsman’s abstract to my comment, or is it to place at the arXiv a note of my approval of Landsman’s paper?

More particularly, what would you recommend as the relevant “excerpt” from this comment, and do you expect this to appear at the arXiv?

Here are three possible excerpts:

a) The whole comment.

Is there a length limit?

b) “An excellent review”.

Note that the comment I picked just happens to start appropriately. My next item begins “Members of a criminal gang are being held in separate cells”!

c) Something that I can generate automatically, such as “Matthew Donald comments on this paper here”.

d) No excerpt.

c and d are the easy options for me.

I have two further problems with the idea of trackbacking, both of which apply to many people who have useful public comments on arXiv papers.

I am not a blogger in the strict sense. Indeed, my “readings” page is just a single, fairly long, occasionally-updated, “hand-written” HTML page which now has comments on around 145 papers. This means that:

1) I don’t have internal html tags indexing the separate items on my page. I could put them in, but I’m not sure whether this would be essential. I’m inclined to think that if you want to find a comment on a paper then you should be prepared to search a page when you get to it.

Would you agree?

2) I don’t use blogging software and I don’t particularly want to; although if backtracking takes off and there are no work-arounds, I might eventually do so.

According to the arXiv help page, “Trackbacks from known blogs should become visible in a few minutes, but it may take longer for us to recognize new blogs.”

This suggests (assuming that “from” really means “from” and not “about”) that the arXiv is going to check that there is a link between the IP address from which they receive a ping and the IP address of a comment.

Does this mean that I will be able to use a standalone TrackBack tool downloaded from here and placed on my site to ping? If so, what are the precise placement restrictions?

Is there a simpler mechanism?

The technical specification page says that pings are made using “standard HTTP calls”. This suggests that I might be able to automate the process using essentially a self-generated HTML form and appropriate text processing. Is this right? (I’m way out of my depth here.)

Am I right to believe that pinging is something which only need be done once, and that it is subsequently up to the arXiv to check for dead links?

If the arXiv recognises my readings page as a legimate source of comments, but does not check the source of pings then it will have no way of telling whether a ping about my readings actually came from me, or from your “recognised” blog, or even from anyone using one of several sites including (for example and without any endorsement) this one which offer public and free “pinging” services.

Perhaps however, such a service is something that the arXiv itself could offer (behind its password protection) to its registered submitters. In such a case, I would be able to ping the arXiv each time I wanted to notify it that I had a comment on a paper which I wanted to share with other readers and the arXiv would already be reasonably sure of my identity.

This is a fairly tedious (and difficult to automate) mechanism from my point of view, but that also means that such comments are less likely to be spam. It would be most useful for those with only the occasional comment, but again, the arXiv would have to decide whether it wanted to require/check that the linker was linking only to his own site.

With best wishes,

Matthew Donald.

web site: a many-minds interpretation of quantum theory .
***************************************************

Posted by: Matthew J. Donald on August 30, 2005 8:14 AM | Permalink | Reply to this

### Re: Trackbacks and the ArXivs

For quite some time, you have been linking to my readings from quant-ph page

I therefore presume that you might allow me into your category of “serious physicist-bloggers”.

Well …

1. My sidebar contains many links which are not to “serious physicist-bloggers.”
2. There are many, presumably “serious physicist-bloggers” who aren’t linked-to.
3. Who cares what I think anyway?

But, yes, as a registered author at quant-ph, I expect you would be able to send trackbacks (not ‘backtracks,’ which is something else) to abstracts on quant-ph.

Now, what do you see as the purpose of the backtracking system in relation to this comment?

I don’t see much purpose for that particular one. But, then, given that you would be manually submitting trackbacks, you could simply not send one for that entry of yours.

In general, I tend to agree that the fully automated use of trackback auto-discovery (which generates a trackback for every link you make) is a scourge, generating lots of superfluous trackbacks.

On the other hand, I think it’s bettter, on balance, to have some superfluous trackbacks, if this insures that the useful ones also get sent (instead of being omitted due to oversight or laziness).

More particularly, what would you recommend as the relevant “excerpt” from this comment, and do you expect this to appear at the arXiv?

Currently, the excerpt isn’t published. In the future, that might change, so I would craft a good one: a two-sentence summary ( no HTML) of your comment.

I don’t have internal html tags indexing the separate items on my page. I could put them in, but I’m not sure whether this would be essential.

Essential? Maybe not. Would it make your page vastly more useful? Absolutely.

I don’t use blogging software and I don’t particularly want to; although if backtracking takes off and there are no work-arounds, I might eventually do so.

I don’t think trackbacks (by themselves) are a reason to switch to using a CMS from hand-crafted HTML. Producing better, more usefully marked-up pages with much less effort on your part is the reason to switch.

… arXiv is going to check that there is a link between the IP address from which they receive a ping and the IP address of a comment.

No it’s the trackback-URL (which points to the your entry about the paper in question) which is significant.

This suggests that I might be able to automate the process using essentially a self-generated HTML form and appropriate text processing. Is this right?

Yes. Though, rather than have everyone reinvent the wheel for themselves, a free-standing tool for generating trackbacks to the arXivs is under development.

Perhaps however, such a service is something that the arXiv itself could offer (behind its password protection) to its registered submitters.

Indeed …

Posted by: Jacques Distler on August 30, 2005 9:04 AM | Permalink | PGP Sig | Reply to this

### Re: Trackbacks and the ArXivs

My two cents, for self styled pingers. I’m not sure of the way to add the title, but it should work.

 $cabeceras=array();$cabeceras['Content-Type']="application/x-www-form-urlencoded"; $arxivtrack="http://arxiv.org/trackback/".$matches[0]; $contenido="url=http://www.physcomments.org/node/".$node->nid."&title=PhysComments+Collective+Blog"; $request=drupal_http_request($arxivtrack,$cabeceras ,'POST',$contenido ); 

Posted by: Alejandro Rivero on September 3, 2005 6:40 AM | Permalink | Reply to this

### Re: Trackbacks and the ArXivs

About not having blogging software, well, it seems that the implementation is very trivial, the only thing one needs is an utility able to send POSTs (and I am not sure if ArXiV admits GETs instead, does it). Lynx or some variant should work. But The ArXiV should make clear which of the non compulsory fields are compulsory for them (eg title? excerpt?)

Posted by: Alejandro Rivero on September 1, 2005 2:53 PM | Permalink | Reply to this

### Re: Trackbacks and the ArXivs

To be precise, the lynx option is -post_data, the data is taken from stdin

So from unix
 lynx http://arxiv.org/trackback/blah -post_data 
with input
 url=http://myurl/&title=my+title --- 
should work. Untested.

Posted by: Alejandro Rivero on September 3, 2005 6:48 AM | Permalink | Reply to this