Dave Winer says that he largely makes software for himself these days.
Ideal Node.js hosting service
Here’s the URL of a GIT repo, containing a Node app. Run it with forever.
I want to be able to do a tail -f on the log so I can see what’s going on.
And to view its file system, perhaps through a browser-based JS app.
I think that’s about it. I give you a URL, you run it.I know some services get close, but I don’t want close. I want just this. A Node virtual machine.
Source: A Node virtual machine
A preview of Dave Winer’s new blogging system : 1999.io
Worth a listen : Podcast: What’s going on behind the scenes at Scripting News
Dave Winer on why he’s using Twitter for identity?
OK. At this point I’m now officially confused by the Dave Winer road-map / strategy.
I’m sure it’s evolving and exploratory. But Dave’s productivity means that I’m no longer keeping up with how this is meant to go down.
So we had an outliner. (Fargo). And then an open-sourced version of the editor (Yay! Concord). Dave got the religion of “Unhosted” apps. running entirely in the browser and using various back-end storage. (Cool). Although he has a node server for that too.
And now there’s a MyWord which looked like a way to format long-form essays to make an open competitor to Medium. And Little Pork Chop to cut long stories down to Tweet-size chunks to post on Twitter.
But MyWord is also meant to post to Twitter. (I think)
And now MyWord also seems to be blogging software. (Following the path of Medium?) But when I asked about how the outline editor would connect to it, Dave said he wasn’t thinking about it. Instead MyWord is gaining an ordinary in-browser text editor. But it’s also getting a “front page” which shows an ordered list of posts … pulled from an RSS feed.
Dave’s own blog seems to be outline based. There’s a liveblog version for “narrating his work” that basically IS in an outline. But MyWord seems to be going off in a different direction.
So what’s actually going on here? Are these all pieces of a single jigsaw puzzle that can be put together to make a comprehensive whole (a version of Radio Userland as a swarm of Unhosted apps)? Or is Dave running a bunch of experiments in parallel, pursuing an outliner-first strategy on the one hand, and chasing Medium and Twitter on the other?
So, Dave has a nice template for showing essay-length writing.
And naturally, the way things are evolving in 2015, it’s a “single-page web-app” which can pull in the actual content from another URL.
Of course, I’d like to write essays using an outliner. And I’m using OWL which is really just the Concord editor that saves files in OPML. But myword.io has a new custom JSON format for its input (that also seems to allow Markdown).
So here’s the question : what’s the story for converting an outline to an essay? Is myword going to evolve to import OPML directly? Is the myword JSON format going to become a standard? How should OPML turn into myword documents?
In one sense it’s easy. You can always just flatten OPML and dump it to a raw text file. Perhaps with some indentation based on outline depth. (Although that might fight the Markdown.)
Or are there plans to represent any structure in the essay (sections, section headings etc.) with structure in the OPML? Is there a project to work on this? Or thoughts on a standard that I could start to support?
Obviously I have a couple of my own solutions in this area : GeekWeaver and bootdown. But if there are any emerging standards / conventions planned, then I’m up for supporting them. Both GeekWeaver and BootDown are useful, but they’re both rather old-skool command-line Python scripts which don’t necessarily fit into the workflow of many people in 2015.
Update : Dave replied :
It is what it is.
I have no plans at this time to do anything with it.
I might swing back around to this, or not.
MyWord is MIT Licensed, so if you really want it, you can do it.
OK. So right now, I’ll probably just do a flat export of text written in OPML and a quick JSON wrapper. And then think further. As this ties into questions about GeekWeaver, the future of some of my Python code base (as opposed to moving more into the browser)
Dave also points to his code :
Here’s the code I added to Fargo to generate the myword.io JSON.
As you can see, there ain’t much there! 🙂
Still. Very useful.
I’m almost ready to start playing with it. Just have to figure out where / how to host the result. Will almost certainly be a WordPress blog.
Will I use it for Smart Disorganized though? One issue with radio3 currently is that it only seems to be connectable to one WordPress. What if I have several?
Dave says it well.
A tool to slice long stories into “tweetstorms”? A meme creator that posts directly to Facebook? A tool to prioritise tweets from important friends? What’s going on here?
Dave doesn’t trust major platform vendors an inch. But he’s obviously decided that at this point, you can’t beat ’em. So you’d better join / embrace (and extend) them.
I guess also some of the problems he had with the Fargo server last year are pushing him towards clients to other people’s servers. So faced with a world that insists on using Twitter and Facebook as its transport protocols, the question he’s answering is how to make the authoring tools he wants, sit on top of those protocols.
Update : Dave has a short podcast about them.