Marty wrote a great, thoughtful essay about some of the problems with webmention right now, and I agree with it.
By that I mean, a sender attempts to send a mention and if there is no endpoint or the send fails, then as a back up, the sender parses the receiving site’s page for an h-card and if an email address exists, sends an email notification there instead?
Might be helpful for those who don’t yet have Webmention set up, but could act as a backup. Then when they have things working later, they could force manual mentions to recollect them? Also useful for those who’d like notifications, but don’t want to build infrastructure or who might not want to show comments on their site either.
Pingbacks are essentially dead and in personal experience some of the few sites that still support them are in academia, but they’re relatively rare and have horrible UI in the best of times. Webmention is a much better evolutionary extension of the pingback idea and have been rapidly growing since before the spec was standardized by the W3C.
I’ve sketched out how individual academics could use their own websites and publish pre-prints and syndicate them to pre-print servers and even to their final publications while still leveraging Webmentions to allow their journal articles, books, other works, to accept and receive webmentions from other web publications as well as social media platforms that reference them.
I think the Microformats process is probably the best standardized way of doing this with classes and basic HTML and there is a robust offering of parsers that work in a variety of programming languages to help get this going. To my mind the pre-existing
h-cite is probably the best route to use along with the well-distributed and oft-used
<cite> tag with authorship details easily fitting into the
As an example, if Zeynep were to cite Tessie, then she could write up her citation in basic HTML with a few microformats and include a link to the original paper (with a rel=”canonical” or copies on pre-print servers or other journal repositories with a rel=”alternate” markup). On publishing a standard Webmention would be sent and verified and Tessie could have the option of displaying the citation on her website in something like a “Citation” section. The Post Type Discovery algorithm is reasonably sophisticated enough that I think a “citation” like this could be included in the parsing so as to help automate the way that these are found and displayed while still providing some flexibility to both ends of the transaction.
Ideally all participants would also support sending salmentions so that the online version of the “officially” published paper, say in Nature, that receives citations would forward any mentions back to the canonical version or the pre-print versions.
Since most of the basic citation data is semantic enough in mark up the receiver with parsing should be able to designate any of the thousands of journal citation formats that they like to display any particular flavor on the receiving website, which may be it’s own interesting sub-problem.
Of course those wishing to use schema.org or JSON-LD could include additional markup for those as well as parsing if they liked.
Perhaps I ought to write a longer journal article with a full outline and diagrams to formalize it and catch some of the potential edge cases.
If you have questions/problems with them or want to chat with the developers on potentially improving them, I’ll invite you to join the IndieWeb WordPress chat: https://chat.indieweb.org/wordpress/.
Webmentions are strange, at least in how the WordPress plugin handles them, as they contain far less context about the pinging post — which is to say none whatsoever. Old-fashioned trackbacks and pingbacks at least include a snippet of the post which sent the ping. Webmentions are presented simply with, “This post was mentioned by whomever.” This does not seem especially helpful when such inter-blog links are meant to serve not just conversation but context on the web.
“If your software fails you,” writes Evgeny Kuznetsov, “it’s not always an indication that something is wrong with the protocol.” Which might be why my webmentions post this replies to begins, “Webmentions are strange, at least in how the WordPress plugin handles them.” Although, upon doing some reading, it’s my understanding that trackback does send a post excerpt, etc., although it’s optional; only pingback and webmention send just the URL.
Feel free to email or call me directly and we can set up a time to chat.
While some of these ideas sound romantic at present with minimal penetration and implementation, we’ll definitely need to be cognizant of how they grow and building tools to mitigate abuses in the future as they become more common. No one wants Webmention to become a vector for spam and harassment the way it’s poorly designed and implemented predecessors like Pingback or Trackbacks were.
While the IndieWeb seems to be the largest hub of this conversation so far, especially for the technical portions, it’s also been distributed across multiple platforms and personal websites and wikis. If you haven’t come across the IndieWeb you may appreciate their wiki and bridged chat channels.
Lately I’ve noticed a big spillover into the wiki space primarily by way of Tom Critchlow, Kicks Condor, some from TiddlyWiki and the Roam Research spaces, and many of your colleagues at egghead.io. I’m personally looking forward to the convergence of the website, blog, personal wiki, commonplace book, etc. in a single platform.
As I notice that you’re in Brighton, if you haven’t been before, you might consider joining in one of the local Homebrew Website clubs either there, in other parts of the UK, or across the world. I see events for Nottingham and London coming up on the schedule, but I’m sure Jeremy Keith or other organizers will do another in Brighton soon.
In any case, you’re on the web, and we can “see” and “hear” you. Thanks for drawing up a campfire to create a discussion.
Sadly, it seems like too many in the thread completely got lost on the “why” portion which was the best part of the question.
/wp-admin/admin.php?page=webmention) and make sure your self-ping settings will allow it.
If you wanted, you could also modify the Webmention type and/or the excerpt that shows in the comment section, though you’d need to do it manually.
I’m not aware of anyone using it “only” for this purpose. I think David Shanske also has built some whitelisting settings for Webmention moderation so that you can automatically approve ones from certain domains. I would suspect you could use some of those portions to , but it may require a few lines of code to do it.</p
My wiki currently doesn’t have very many webmentions or incoming links, but after writing about a Bookmarklet for pasting content into TiddlyWiki, I got an email from Anne-Laure Le Cunff that she’d used some of the code to write a bookmarklet for Roam Research. Since her article didn’t send a webmention, I used Telegraph to manually force her article to send my wiki a Webmention so my account would have a record of it for the future for potential exporting or other use.
Now I’d like to display this webmention on that tiddler. Doing it automatically would be great, but I thought, since I don’t expect to receive many on my wiki that I ought to try out a manual set up to see how things might work and how I might display them if they were automated.
Since I had created that bookmarklet, I used it to copy and paste the text from Anne-Laure Le Cunff‘s website into a new tiddler. I then massaged it a bit to format it to look like a response and I’ve transcluded it into the original post under a heading of Responses.
The side benefit of doing this is that the stand alone tiddler that has the link and the context from her post also sits in my wiki as a bookmark of her post as well. As a result, I get a two-for-one deal: I get the bookmark of her post with some context I’m interested in, but my original post can now also display it as a response! Now I can also use that bookmark in other places in my website as well. If only one could do this so easily in other CMSes?!
I’ve yet to hear of another example of this in the wild, so unless I’m missing something, this may be the first displayed Webmention on a TiddlyWiki in the wild.
TiddlyWiki has lots of ways to display data in Tiddlers, so perhaps one might use various fields in a bookmark tiddler to create the necessary comment display. This could give a more standardized method of displaying them as well. It could be particularly useful if someone was using a microformats parser to import the data of such mentions. If this were the case, then the tiddler that is being commented on could do a filter/search for all tiddlers in the wiki that mention it and transclude the appropriate pieces in a list format with the appropriate mark up as well as links back to the individual tiddlers and/or the links to the sending site.
I’m curious if others have ideas about how to best/easiest implement the display portion of webmentions on a public TiddlyWiki? Since I’m also hosting my entire TiddlyWiki on GitHub pages, there might be other potential considerations if I were to be hosting it statically instead. This may require some experimentation.
I’ve got a few mental models about how one might implement showing Webmentions in TiddlyWiki, but it may take some more thinking to figure out which way may be the best or most efficient.
I don’t anticipate a lot of incoming webmentions to my wiki at present, but if they become more prevalent, I’ll want to automate the display of these notifications somehow. This will take some thought and coding as well as more knowledge of the internals of TiddlyWiki than I’ve currently got. If someone with the coding chops is interested, I could probably brainstorm a set up fairly quickly.
It would also be nice to be able to have full microformats support in TiddlyWiki so that the stand alone “bookmark” mention works properly, but also so that the transcluded version might have the correct mark up. This may rely on the two things to be properly nested to make them work in both contexts.