I don't remember. But give me some time -- kind of rings a bell and it sounds like something Dan would be interested in.
BTW this is a Markdown browser too.
People assume Markdown has to look a certain way for certain applications.
But this is evidence that Bluesky or Twitter or micro.blog could be Markdown browsers.
We have far too limited view of what a browser is. It's our habit of trusting big companies that gives us the limited idea that a browser has to look like browsers have "always" worked.
There might be other even better ways to read.
BTW, I have terabytes of backups, and lots of .root files to look for.
But I'm not sure what to look for.
Esp want the manilasites.root from the hosting sites, to see what's in there.
We are getting close to a workable Frontier.
Picked a file at random, docserver.root -- it's an object database that contains all the writing and code to build the docserver.userland.com site.
Took a few trials, uncovered bugs in the new kernel, it's not as simple a site as you might imagine. More here.
thanks for posting this here. it's still pretty new. i diverted for a bit to get Frontier running on new OSes, and then it's back to this network. we will pick it up with FeedLand, first creating an API for it, and then integrating its feed reading prowess and design patterned after the social networks like twitter, masto and bluesky. but as with rss.chat -- powered only by open formats and protocols in wide use.
We’re starting to focus on scriptable apps, esp web services like GitHub, Mastodon, Slack.
We also have SQLite baked in, and I was wondering while watching the news is we could have a verb that creates a new database, one one that sends a query one of those databases.
I'm deep in Frontier territory. No podcasts, can't publish them right now. At this moment I can't even update my blog! ;-)
I built a lot in Frontier over the years, and moving that whole body of code and writing to a new app I wrote with Claude in a few months is a pretty ambitious thing to do.
But it is working to some extent. I am typing this from the new system.
I have a new Chromebook, I keep it on the kitchen table so I can get online quickly from downstairs. I've been able to log on to every service I need except X. It insists that to log in I have to use their app, but if they have a Chromebook app, the Google store doesn't know about it. Why they can't just let me enter the email address and password? They could send a code to my phone. What else. I think this has something to do with having encrypted direct messages.
These are the big features that remain to be done in Atlantis, the new version of Frontier in development with myself and Claude. The one that's bold is what we're working on now. I want to get it into shipping shape. And a couple of the things on the list came from Claude and I don't know what they mean.
I got rid of all the claude verbiage. everyone knows what the menu editor is, and all the rest of it.
the menu editor is working now.
we're moving toward getting a shipping disk put together, ready to be installed on a user's machine. that's my next step, installing the virgin frontier.root and then bringing in nodeEditor.root as a guest database, and shake that out.
my aim is to have the desktop app running everywhere Electron runs, i want to get this thing going, tired of being the only user.
it is starting to feel reliable, but not all the way there yet, the outliner is the core, and it's very much debugged, having been part of my JS outliners since 2013.
Thinking out loud about blogging from this space.
It's pretty close to being able to do what I do in Electric Drummer in an outline, just need a few more options, they would fit in the context menu on each post as you edit. Lots of room there.
The blog rendering would run off the RSS feed. I probably would write something new and very simple, almost like an RSS feed browser.
The reason I would start with a pilot project here is because there are other people here, already. And if I said something provocative that you knew other people would see, would you be drawn to that? I trust the people I've invited here not to be spammers.
We might be ready for that. We put a lot of effort into the comment system here. We plan to use the new features to create a very new kind of web.
No one tries out new ideas in the blogging world. That could change.
Obama should have seated Garland if the Senate wouldn’t have a vote to advise and consent. assume they consent.
And the governor of Kentucky should appoint a successor to McConnell unless someone can prove he’s actually alive.
Creative computer people, consider consolidating all the different OSes and programming languages. Give humans a fair fight in competition with the AI's who have no trouble deeply understanding every syntax human designers have come up with, where it totally disadvantages us, who have much more limited memory.
Claude Code is like Victor Wembanyama, assuming you follow the NBA and are interested in first hand experience with CC as a programming partner for the last 8 months or so.
I read Bill Gates' memo, and there's one thing almost no one has mentioned about AI systems, is how dependent we're getting on them. What if one day they stop working. What's the backup plan?
When we started up with PCs and then the web, it took a long time for us to build mission-critical systems around them. But this is happening very quickly.
That was fast! ;-)
I'll link to this on Scripting News shortly.
Thanks for doing this work Scott.
BTW we all know there isn't an RSS 2.0 feed, it isn't a podcast, but rather than getting all preachy, which no one likes, mention a "podcast feed" as a real thing that's expected. Always add "feed" when you say podcast.
The feed gives you and podcast producers the thing that makes podcasting so different -- freedom. For listeners it means if you don't like something the software does, you can switch without losing anything. And for producers there's no one who can stop you from speaking. There is always a way around a blockage. But only if there's an RSS feed.
Frontier work going much better today we’ve got all the built in verbs in the product reviewed tested and I’m not finding any problems as I create software for deploying this stuff.
The first thing we had to get working was being able to move objects from the 2011 frontier to the 2026 frontier and we got there today. The next thing that’s coming is being able to do updates for users with new parts and changes to existing ODB elements.
So once that’s working, we can start fixing bugs and looking for problems and things didn’t get ported correctly. We have a new gate. Every change must be vetted by reviewing the 2011 code, and doing exactly what we did then.
John Kerry asked who wants to be the last casualty in Vietnam, a lost cause like the war we are bogged down in Iran for no good reason, like Vietnam.
What happens when humans run out of things they want to do with AIs. When everything has been done and it takes ten tokens to do what used to cost ten million. What will we do with all the data centers? Will they be empty shells like the old shopping malls that were so popular before the web and Covid?
Added a chapter to the story.
https://github.com/scripting/helloFrontier/issues/1
Turns out Claude didn’t understand how Frontier calls scripts. It’s much simpler than what it concocted.