I Will Never Be Excited to Share

There is a sentence I will not write. You know the one. It starts with "Excited to share" and ends with a link to something I made, and it is sitting in the compose box every time I open LinkedIn, waiting for me to give in. It annoys me considerably more than it deserves to.

For a while I assumed that annoyance was snobbery, and some of it is. But I don't think the phrase says much about whoever types it. A feed gives you a reason to sound pleased, keep it short, and get to the point before somebody scrolls past. Do that a few hundred times and it becomes how you write. The box won.

The usual argument for publishing on your own site is about rights, or privacy, or not being deleted by a company having a bad quarter. I believe all of that, and most of this site is one version or another of it. What I notice first is that I feel less pressure to sound a particular way here. I can spend a paragraph admitting that my objection to a phrase is partly snobbery. There is room for it. A text editor and a stylesheet I wrote.

I still have to put something in the compose box if I want people there to find it. I would just like to have finished writing before I open it.

Where the article lives

The IndieWeb crowd has had a name for this for years: POSSE, publish on your own site, syndicate elsewhere. The original lives at an address you control. You put copies or links on the platforms where people might read them.

Cory Doctorow's Pluralistic is a working example, with full-text versions available through feeds and other platforms. Reading him is a useful reminder that an article can exist in several places without leaving any doubt about where to find the original.

Try finding a LinkedIn post from three days ago. Not one you remember clearly, just one you know exists. I find this much harder than it should be. Once I have the URL, I can cite it. Whether it stays available is somebody else's decision.

Here, I choose the address and maintain the page. Keeping it available in ten years will require me to keep paying for the domain and looking after the files, but those are jobs I can do. I cannot maintain a LinkedIn URL.

Anything I leave only inside somebody else's product depends on their willingness to keep it there. They can change what readers see, what I can export, or what I have to pay for. They may have perfectly ordinary business reasons. I still have to live with the result.

I am not leaving LinkedIn. I am making sure it holds the second copy.

What publishing here involves

I write the article in Markdown. Then I write the page by hand, because there is no build step on this site and I have decided I like it that way: an HTML file linking two stylesheets, the metadata in the head, one line added to the index, one item added to feed.xml. Then git push, and a workflow resets the checkout into the webroot on the server. Nginx serves the files. I have written up the deployment and server setup separately.

None of this was built in order to do POSSE. I built it because I wanted to run my own server, and the publishing workflow is a side effect. Without that, the compose box would have been much harder to argue with. I am not going to tell anyone to stand up a server so their posts have better URLs. I already had one.

There is still some fuss every time I publish. The feed needs updating. The page needs checking. A typo may need fixing in more than one place. I like having so few moving parts, although that does mean I move them myself.

The bit that goes in the feed

Every article here gets a social card rendered from a template in the repository, using the same fonts as the page. A headless browser makes the image. When the preview appears in LinkedIn, a piece of this site's design comes with the link.

I find that quietly satisfying in a way I cannot entirely justify. I spent time on the typography. I like seeing it survive the trip.

That is also why I put the link in the post body. There is plenty of advice about hiding it in the first comment to avoid a supposed reach penalty. I cannot tell whether putting it in the post costs me readers. I like the preview card, and I want the link where people can find it. That is enough for me to keep putting it there.

I use a short blurb instead of pasting the whole article. Full-text syndication makes sense if you want people to read wherever they happen to be. For this site, I want to send them to the page I made. Two or three lines written for the post, then the link.

That means asking someone to leave the feed. Some will not bother. It also means writing a second thing that needs its own care, and this is where the voice I was trying to avoid comes back. The article is finished; now I have to describe it without turning into someone announcing an exciting milestone.

I don't know what this costs me in reach. Writing the extra post is work I can account for.

Reading myself later

The other reason has nothing to do with readers arriving from LinkedIn.

In The Memex Method, Doctorow describes years of blogging becoming a searchable store of things he has thought about. He also makes a point that is easy to lose in a discussion of archives: writing for other people makes him explain things more carefully than he would in private notes. The publication is part of the thinking.

I am months into this rather than decades. The part I have already noticed is smaller. This article started as a file in a folder of article ideas, cross-referenced with a note about the small web and another about hardened static sites. Following those notes led me back to something I had written down and forgotten completely. I found it because it was in a folder I own, in a format I can grep.

That would have worked without publishing anything. The folder was useful on its own. But an article gives me a more complete account to come back to: what I was trying to do, what I decided, and enough explanation that I might still understand it after I have forgotten the afternoon I spent on it.

I would like to be able to read myself in ten years. It is a strange reason to run a web server, and it is the one I actually have.