Neil Mather (doubleloop)My online social experience is mostly through the indieweb. For following people and blogs, I use Aperture, a Microsub server, to subscribe to various social feeds. And then I read and interact with those feeds in various clients – e.g. Indigenous on Android, and Monocle on the web. Although I don...
Jonathan Gill (Nieman Lab)One way to meet the many needs that most if not all publishers share would be to collaboratively develop their digital products. Specifically, they should build for interoperability. One publisher’s CMS, another’s content APIs, a third company’s data offering — they might one day all work together to allow all ships to rise and to reclaim advertising and subscription revenue from the platforms. This might allow publishers to refocus on differentiating where it truly matters for the user: in the quality of their content. ❧
I don’t have quite the crazy analysis that Jeremy Keith has done of his posts, and I initially thought that there was no way I’d posted as much as he had. Perhaps it might be worth delving deeper into the numbers to see exactly what is going on?
Possibly worse(?!), that total posting number is up from 1,762 public posts in 2017. I can only attribute the increase in quantity to the ability to increasingly easily post to my site via micropub clients and some simple bookmarklets I use in conjunction with David Shanske’s brilliant Post Kinds plugin. G-d bless the IndieWeb and its tremendously helpful community for helping me take back ownership of my digital online life. I can only imagine how much higher that number goes this coming year if I can manage to build a Microsub set up and indie reader into my website and make the entire processes even more friction-less.
I unwittingly spent a few minutes last night on cleaning up some plumbing on my back end that will make it easier to follow up (when necessary) on likes, reads, and bookmarks that I collect.
I can’t bear to go through and count the number of private posts for the year, but I will say that having my own online searchable database of things I’ve written, replied to, bookmarked, read, listened to, watched, annotated, etc. has been incredibly useful over the past few years.
James King (Twitter)Started tinkering with building a feed reader again.
— Les Orchard (@lmorchard) December 28, 2018
Baffled to find that some of the feeds I subscribe to have 300-500 items in them for no good reason.
(e.g. a web comic or a news site with like the last 2-4 years worth of items listed)Have you ever considered just using Newsboat and then building a web interface on top of it for mobile reading? I have been thinking about doing that myself recently.
— James King (@elroyjetson) December 28, 2018Huh, I'd never actually heard of Newsboat until now. For a bunch of years, building a feed reader was my go-to learning project for new web dev frameworks
— Les Orchard (@lmorchard) December 28, 2018It is mostly a terminal based app, but it handles the crapy stuff like RSS and Atom parsing. Stores everything in an SQLite3 DB. You can still use it as a learning project, but not have to deal with the yuck part.
— James King (@elroyjetson) December 28, 2018
Uonai (Twitter)It is. One of my side projects at the moment is building an RSS feed reader that has social capabilities. So you can add friends, curate feed lists, and see what other people in your network are reading and recommending.
— Uonai (@uonaiii) December 28, 2018
Neil Mather (doubleloop)I’ve been playing around with Metabase to view a few stats about my website. frequency and spikes
It was fun to look at the frequency of my posts over time – you see quite a prominent spike around March and April 2017, and then there’s a slowish decline in frequency until around August/Septemb...
I haven’t really delved into microsub yet, but I suspect it’s going to have an even more profound effect on my reading and posting habits.
Damian Yerrick (Twitter)I tried to recommend #IndieWeb technologies to those affected by a recent Tumblr policy change. But many I talked to claim that recommendation is the whole draw of silos, and IndieWeb's lack of a recommendation engine is a deal breaker.
"None currently."https://t.co/k9c7ehsoZb — Damian Yerrick (@PinoBatch) December 4, 2018
Of course going the direction of old school blogs and following those who comment on your own site has historically been a quick way to build a network. I’m also reminded of Colin Walker’s directory which creates a blogroll of sorts by making a list of websites that have webmentioned his own. Webrings are also an interesting possibility for topic-related community building.
Since Tumblr is unlikely to shut down immediately, those effected could easily add their personal websites to their bios to help transition their followerships to feed readers or other methods for following and reading.
Of course the important thing in the near term is to spend a moment downloading and backing up one’s content just in case.
Dries Buytaert (Twitter)“It touches me that some people have been reading my blog for over a decade.” https://dri.es/forty I've been thinking about that since. It would be nice to know who is subscribed to your blog so you can build a relationship. #indieweb @davewiner
— Dries Buytaert (@Dries) November 22, 2018
With relation to your desire to know who’s subscribed and potentially reading your posts, I think there are a number of ways forward, and even better, ways that are within easy immediate reach using Drupal as well as many other CMSes using some simple web standards.
I suspect you’ve been following Kristof De Jaeger’s work with the Drupal IndieWeb module which is now a release candidate. It will allow you to send and receive Webmentions (a W3C recommendation) which are simple notifications much the way they work on Twitter, Facebook, etc. I’ve written a bit about how they could be leveraged to accomplish several things in Webmentions: Enabling Better Communication on the Internet.
Not mentioned in that article for brevity is the ability to send notifications via Webmention when one makes follow or subscription posts.
As an example, I’ve created a follow post for you for which my site would have sent a Webmention. Unfortunately at the time, your site didn’t support receiving it, so you would have missed out on it unless you support older legacy specs like pingback, trackback, or refback.
I also created a larger related Following page of people and sites I’m subscribed to which also lists you, so you would have received another notification from it if you supported Webmention.
I’m unaware of anyone actually displaying these notifications on their website (yet!), though I’ve got some infrastructure on my own site to create a “Followed by” page which will store and show these follows or subscriptions. At present, they’re simply stored in my back end.
Heh – that’s funny – @lizzjoy and I were just saying we need to find a way of knowing who is reading the Drupal community and association blogrolls…
— Rachel Lawson 🇪🇺 (@rachel_norfolk) November 22, 2018
As for Rachel’s request, this too is also possible with “read” webmentions. I maintain a specific linkblog feed (RSS) with all of the online material I read. All of those posts send notifications to the linked sites. While it’s not widely supported by other platforms yet, there are a few which do, so that online publications can better delineate and display the difference between likes, bookmarks, reads, etc. There’s at least one online newspaper among 800+WordPress websites which support this functionality. I suspect that with swentel’s Drupal module and some code for supporting the proper microformats, this is a quick reality in the Drupal space as well. Because the functionality is built on basic web standards, it’s possible for any CMS to support them. All that’s left is to ramp up adoption.
Dave Winer in his reply to you linked to a post about showing likes on his site (presumably using the Twitter API) where he laments:
I know the Like icon doesn’t show up in your feed reader (maybe that can change)
Interestingly, swentel’s module also supports Microsub, so that reader clients will allow one to like (bookmark, or reply to) posts directly within readers which will then send Micropub requests to one’s website to post them as well as to potentially send Webmention notifications. These pieces help to close the circle of posting, reading, and easily interacting on the open web the way closed silos like Facebook, Twitter, et al. allow.
Toward this end I thought I’d double down on my use of Inoreader in my daily web consumption workflows. I wanted to make it easier to use my feed reader to post all these types of posts to my website which will still handle the notifications. In some sense, instead of relying on a feed reader supporting Micropub, I’ll use other (older) methods for making the relevant posts. As I see it, there are two potential possibilities using Inoreader:
(1) using a service like IFTTT (free) or Zapier (paid) to take the post intents and send them to my WordPress site, or
(2) using the custom posting interface in Inoreader in conjunction with post editor URL schemes with the Post Kinds plugin to create the posts. Using WordPress’ built-in Post This bookmarklet schemes could also be used to make these posts, but Post Kinds plugin offers a lot more metadata and flexibility.
Below is a brief outline of some of the IFTTT recipes I’ve used to take data from posts I interact with in Inoreader and post them to my own website.
IFTTT has an explicit like functionality with a one click like button. There is an IFTTT recipe which allows taking this datum and adding it directly as a WordPress post with lots of rich data. The “then that” portion of IFTTT using WordPress allows some reasonable functionality for porting over data.
IFTTT also has explicit favorite functionality using a one click starred article button. There is an IFTTT recipe which allows adding this directly as a WordPress post.
Since the “starred” article isn’t defined specifically in Inoreader as a “favorite”, one could alternately use it to create “read” or “bookmark” posts on their WordPress websites. I’m tempted to try this for read posts as I probably wouldn’t often use it to create favorite posts on my own website. Ultimately one at least wants an easy-to-remember 1 to 1 mapping of pieces of functionality in Inoreader to their own website, so whatever I decide I’ll likely stick to it.
While there is no specific functionality for creating bookmarks in Inoreader (though starred articles could be used this way as previously mentioned), there is a “saved webpage” functionality that could be used here in addition to an IFTTT recipe to port over the data to WordPress.
While Inoreader has a common feed reader read/unread functionality, it is often not used tacitly and this is a means of reducing friction within the application. Not really wanting to muddle the meaning of the “starred” article to do it, I’ve opted to adding an explicit “read” tag on posts I’ve read.
IFTTT does have a “New tagged article” recipe that will allow me to take articles in Inoreader with my “read” tag and post them to my website. It’s pretty simple and easy.
For dealing with replies, there is an odd quirk within Inoreader. Confoundingly the feed reader has two similar, yet still very different commenting functionalities. One is explicitly named “comment”, but sadly there isn’t a related IFTTT trigger nor an RSS feed to take advantage of the data one puts into the comment functionality. Fortunately there is a separate “broadcast” functionality. There is an IFTTT recipe for “new broadcasted article” that will allow one to take the reply/comment and post it to one’s WordPress website.
Like many of the above there is a specific IFTTT recipe that will allow one to add subscriptions directly to WordPress as posts, so that any new subscriptions (or follows) within the Inoreader interface can create follow posts! I doubt many people may use this recipe, but it’s awesome that it exists. Currently anything added to my blogrolls (aka Following Page) gets ported over to Inoreader via OPML subscription, so I’m curious if them being added that way will create these follow posts? And if so, is there a good date/time stamp for these? I still have to do some experimenting to see exactly how this is going to work.
In addition to the IFTTT recipe functionality described above, one could also use IFTTT RSS functionality to pipe RSS feeds which Inoreader provides (especially via tags) into a WordPress website. I don’t personally use this sort of set up, but thought I’d at least mention it in passing so that anyone who might like to create other post types to their website could.
If using a third-party service like IFTTT isn’t your cup of tea, Inoreader also allows custom sharing options. (There are also many pre-built ones for Facebook, Twitter, etc. and they’re also re-orderable as well.) I thus used WordPress’ post editor URL schemes to send the data I’d like to have from the original post to my own website. Inoreader actually has suggestions in their UI for how to effectuate this generically on WordPress. While this is nice, I’m a major user of the Post Kinds Plugin which allows me a lot more flexibility to post likes, bookmarks, favorites, reads, replies, etc. with the appropriate microformats and much richer metadata. Post Kinds has some additional URL structures which I’ve used in addition to the standard WordPress ones to take advantage of this. This has allowed me to create custom buttons for reads, bookmarks, replies, likes, and listens. With social sharing functionality in Inoreader enabled, each article in Inoreader has a sharing functionality in the bottom right corner that has a configuration option which brings up the following interface:
Once made, these custom button icons appear at the bottom of every post in Inoreader, so, for example, if I want to reply to a post I’ve just read, I can click on the reply button which will open a new browser window for a new post on my website. The Post Kinds plugin on my site automatically pulls in the URL of the original post, parses that page and–where available–pulls in the title, synopsis, post date/time, the author, author URL, author photo, and a featured photo as well as automatically setting the specific post kind and post format. A lot of this data helps to create a useful reply context on my website. I can then type in my reply to the post and add any other categories, tags, or data I’d like in my admin interface. Finally I publish the post which sends notifications to the original post I read (via Webmention).
With either of the above set ups, there are a few quick and easy clicks to create my posts and I’m done. Could it be simpler? Yes, but it likely won’t be much more until I’ve got a fully functional Microsub server and reader up and working.
Of course, I also love Inoreader and its huge variety of features and great usability. While I’m patiently awaiting having my own WordPress Microsub server, I certainly wouldn’t mind it if Inoreader decided to add some IndieWeb functionality itself. Then perhaps I wouldn’t need to make the switch in the near future.
What would this look like? It could include the ability to allow me to log into Inoreader using my own website using IndieAuth protocol. It could also add Micropub functionality to allow me to post all these things directly and explicitly to my website in an easier manner. And finally, if they really wanted to go even further, they could make themselves a Microsub server that enables me to use any one of several Microsub clients to read content and post to my own website. And of course the benefit to Inoreader is that if they support these open internet specifications, then their application not only works with WordPress sites with the few appropriate plugins, but Inoreader will also work with a huge variety of other content management systems that support these specs as well.
Whether or not Inoreader supports these protocols, there is a coming wave of new social feed readers that will begin to close many of these functional gaps that made RSS difficult. I know things will slowly, but eventually get better, simpler, and easier to use. Soon posting to one’s website and doing two way communication on the internet via truly social readers will be a reality, and one that’s likely to make it far easier to eschew the toxicity and problems of social sites like Facebook and Twitter.
Jon Udell (Jon Udell)Yesterday Luann was reading a colleague’s blog and noticed a bug. When she clicked the Subscribe link, the browser loaded a page of what looked like computer code. She asked, quite reasonably: “What’s wrong? Who do I report this to?” That page of code is an RSS feed. It works the same way as...
It’s not just that the silos can shut down their feeds. It’s that we allowed ourselves to get herded into them in the first place. ❧
December 02, 2018 at 09:16PM
“Who do I report this to?”
Everyone. ❧
A brilliant ending!
December 02, 2018 at 09:17PM
Where’s my Net dashboard? ❧
Interestingly, I came to this post in my feed reader while randomly looking for something I could use as an example in something I was writing about feed readers!!!
December 02, 2018 at 09:18PM
Where’s my next dashboard? I imagine a next-gen reader that brings me the open web and my social circles in a way that helps me attend to and manage all the flow. There are apps for that, a nice example being FlowReader, which has been around since 2013. I try these things hopefully but so far none has stuck. ❧
I’m currently hoping that the next wave of social readers based on Microsub and which also support Micropub will be a major part of the answer.
December 02, 2018 at 09:20PM
I mention it because I was specifically intrigued by a small piece of excellent user interface and social graph data that Reading.am unearths for me. I’m including a quick screen capture to better illustrate the point. While the UI allows me to click yes/no (i.e. did I like it or not) or even share it to other networks, the thing I found most interesting was that it lists the other people using the service who have read the article as well. In this case it told me that my friend Jeremy Cherfas had read the article.1
In addition to having the immediate feedback that he’d read it, which is useful and thrilling in itself, it gives me the chance to search to see if he’s written any thoughts about it himself, and it also gives me the chance to tag him in a post about my own thoughts to start a direct conversation around a topic which I now know we’re both interested in at least reading about.2
The tougher follow up is: how could we create a decentralized method of doing this sort of workflow in a more IndieWeb way? It would be nice if my read posts on my site (and those of others) could be overlain on websites via a bookmarklet or other means as a social layer to create engaged discussion. Better would have been the ability to quickly surface his commentary, if any, on the piece as well–functionality which I think Reading.am also does, though I rarely ever see it. In some sense I would have come across Jeremy’s read post in his feed later this weekend, but it doesn’t provide the immediacy that this method did. I’ll also admit that I prefer having found out about his reading it only after I’d read it myself, but having his and others’ recommendations on a piece (by their explicit read posts) is a useful and worthwhile piece of data, particularly for pieces I might have otherwise passed over.
In some sense, some of this functionality isn’t too different from that provided by Hypothes.is, though that is hidden away within another browser extension layer and requires not only direct examination, but scanning for those whose identities I might recognize because Hypothes.is doesn’t have a specific following/follower social model to make my friends and colleagues a part of my social graph in that instance. The nice part of Hypothes.is’ browser extension is that it does add a small visual indicator to show that others have in fact read/annotated a particular site using the service.
I’ve also previously documented on the IndieWeb wiki how WordPress.com (and WordPress.org with JetPack functionality) facepiles likes on content (typically underneath the content itself). This method doesn’t take things as far as the Reading.am case because it only shows a small fraction of the data, is much less useful, and is far less likely to unearth those in your social graph to make it useful to you, the reader.
I seem to recall that Facebook has some similar functionality that is dependent upon how (and if) the publisher embeds Facebook into their site. I don’t think I’ve seen this sort of interface built into another service this way and certainly not front and center the way that Reading.am does it.
The closest thing I can think of to this type of functionality in the analog world was in my childhood when library card slips in books had the names of prior patrons on them when you signed your own name when checking out a book, though this also had the large world problem that WordPress likes have in that one typically wouldn’t have know many of the names of prior patrons necessarily. I suspect that the Robert Bork privacy incident along with the evolution of library databases and bar codes have caused this older system to disappear.
This general idea might make an interesting topic to explore at an upcoming IndieWebCamp if not before. The question is: how to add in the social graph aspect of reading to uncover this data? I’m also curious how it might or might not be worked into a feed reader or into microsub related technologies as well. Microsub clients or related browser extensions might make a great place to add this functionality as they would have the data about whom you’re already following (aka your social graph data) as well as access to their read/like/favorite posts. I know that some users have reported consuming feeds of friends’ reads, likes, favorites, and bookmarks as potential recommendations of things they might be interested in reading as well, so perhaps this would be an additional extension of that as well?
[1] I’ve certainly seen this functionality before, but most often the other readers are people I don’t know or know that well because the service isn’t huge and I’m not using it to follow a large number of other people.
[2] I knew he was generally interested already as I happen to be following this particular site at his prior recommendation, but the idea still illustrates the broader point.
Richard MacManus wrote about the state of feed readers as he saw it in his AltPlatform.org post titled “The state of feed readers”. He mentioned a couple things in his Feedly wishlist that prompted me to think more about what I’d like to see added to Feedly.
Apparently there were a bunch of us thinking and writing about feed readers and the open web a year ago last June. Several week’s prior to Richard’s article, I’d written a piece for Richard’s now defunct AltPlatform entitled Feed reader revolution (now archived on my site), which laid out some pieces similar to Paul’s take here, though it tied in some more of what was then the state of the art in IndieWeb tech.
Around that time I began tinkering with other feed readers including Inoreader, which I’ve been using for it’s ability to auto-update my RSS feeds using OPML subscriptions from the OPML files I maintain on my own website. Currently I’m more interested in what the Microsub specification is starting to surface in the feed reader space.
I’m not sure if he’s played around with it since, but, like Paul, I was using some of the Press This bookmarklet functionality in conjunction with David Shanske’s Post Kinds plugin for WordPress to make posting snippets of things to my website easier.
Feedly has a Pro (aka paid) functionality to allow one to share content using custom URLs.
While one can use the Share to WordPress URL functionality, I’d recommend the Custom Sharing feature. Using the Post Kinds plugin, one can use the following example URL to quickly share things from their Feedly account to their personal website:
Matt Baerhttps://example.com/wp-admin/post-new.php?kindurl=URL&kind=bookmarkOne should change the URL to reflect their own site, and one can also change the word “bookmark” to the appropriate desired kind including “like”, “favorite”, “read”, or any of the others they may have enabled within the Post Kinds plugin.
I personally don’t use this method as it only allows one custom sharing URL (and thus allows only one post kind), and instead (again) prefer Inoreader which allows one to configure custom sharing similarly to Feedly, but doesn’t limit the number of kinds and the feature is available in their free tier as well.
In addition to some of what I’ve written about the Post Kinds plugin before, I’ve also detailed how to dovetail it with sharing from my Android phone quickly in the past.
Also like Paul, I was greatly interested in quickly creating highlights and annotations on web content and posting them to my own website. Here I’m using a modified version of the Post Kinds plugin to accomplish this having created highlight posts and annotation posts for my site. Next I’m utilizing the ability to prepend http://via.hypothes.is to URLs on my mobile phone to call up the ability to use my Hypothesis account to easily and quickly create highlights and annotations. I then use some details from the outline linked below to capture that data via RSS using IFTTT.com.
Naturally, the process could be streamlined a lot from a UI perspective, but I think it provides some fairly nice results without a huge amount of work.
An Outline for Using Hypothesis for Owning your Annotations and Highlights
I will mention that I’ve seen bugs in trying to annotate easily on Chrome’s mobile application, but haven’t had any issues in using Firefox’s mobile browser.
Jan Cavan Boulas (Twitter)@ChrisAldrich I didn't catch your last name at WordCamp! But I actually have been following your blog for a while now. That article you wrote about the Facebook algorithm is hilarious
— Jan Cavan Boulas (@jancavan) June 21, 2018
My friend Jack Jamieson is in the midst of building a WordPress-specific Microsub server implementation which he’s indicated still needs more work, but which he’s self-dogfooding on his own website as a feed reader currently.
If it’s of interest, you or your colleagues at Automattic might want to take a look at it in terms of potentially adding a related Microsub reader interface as the other half of his Microsub server. Given your prior work on the beautiful WordPress.com feed reader, this may be relatively easy work which you could very quickly leverage to provide the WordPress ecosystem with an incredibly powerful feed reader interface through which users can interact directly with other sites using the W3C’s Micropub and Webmention specifications for which there are already pre-existing plugins within the repository.[1][2]
For some reference I’ll include some helpful links below which might help you and others get a jump start if you wish:
While I understand most of the high level moving pieces, some of the technical specifics are beyond my coding abilities. Should you need help or assistance in cobbling together the front end, I’m positive that Jack, Aaron Parecki, David Shanske, and others in the IndieWeb chat (perhaps the #Dev or #WordPress channels–there are also bridges for using IRC, Slack, or Matrix if you prefer them) would be more than happy to lend a hand to get another implementation of a Microsub reader interface off the ground. I suspect your experience and design background could also help to shape the Microsub spec as well as potentially add things to it which others haven’t yet considered from a usability perspective.
In the erstwhile, I hope all is well with you. Warmest regards!
Indiepaper is a read-later service built for the open web. Save your articles, and never again get locked out by a proprietary service like Instapaper.