Fork me on GitHub

Those 2 last things are awesome, Dustin. I'm preparing to make a pact with myself to go through the tutorial and give all kinds of feedback while doing so. I'm not sure when I can start (I'm finishing up my previous pact which was an everyday thing for the entire year, and I'm catching up a bit of lateness before dec 31 comes.) As soon as I confirm I'm good on this pact, I'll start that pact.


Hi! Note: I could even record myself doing the tutorial if it's some kind of thing you might like watching.


Well, let me think


As you can see, our #1 problem is that the tutorial in the past has been daunting, and I am having trouble seeing it like a beginner


So any ideas you have to solve this, i am highly interested in


I'm almost a good beginner - I spent the last 9 years reading and watching many/most of what can be read about Clojure, without using it much, though. Ok, I'll be sure to ping you when I know the time comes - and I'm sorry it's not, say, tonight. 😉


When is a good day to email you to follow up on this? BTW, I am also happy to sit on video chat with you (and we can record the converation, that would be valuable to me)


As for HF, I plan to use it both in my personal life and at work in some capacity. What capacity exactly is left as an exercise for now.


lol, that's a good question and the answer would depend on what kind of conversation we have - I must protect my current engagement (I'm 77% through and there are only 21 days left to this year!) so if you don't mind, let me think a few days/weeks to see the trend before I answer you - I'm sorry, again (hey, I'm canadian, so I must say a few sorries here and then, as I've heard we're seen to do!?)


Thanks 🙂 Although it almost hurts to make you wait, especially since the tutorial is newly revised. I'll also think about maybe taking a half day off from work to go through this independently.


There is more work to be done this week. I want the tutorial to take like 30 mins of actual "work", after a 1-2 page high level overview of the data model. But don't sweat it. Appreciate you reaching out


Ok, and we wouldn't lose on the details of the rest of the tutorial? I'm the kind who wants the everything, so I hope I don't lose in the waiting. (If my english is a bit weird, it's because it's not my main language.)


"Trail-based tutorials" might be a great approach. Bonus points for keeping track of our progress through each trail. 🙂


Also, Dustin, I love reading your blog posts - if you ever attach a feed to one of them, please tell us so that we may subscribe! Something like this might come in handy:


Which blog posts are you talking about? The ones on the hyperfiddle blog (quarterly updates), or the older technical ones?


And are rss readers still a thing? We can integrate atom into hyperfiddle (from userland!), but i didnt think anybody got content that way anymore - all social media now


both HF blog and your technical blog (if you're going to continue publishing there) I don't know how many people still use RSS like I do but I still follow 60 active sources. That's those sources where I want to know about everything they post (like you). I use social media (twitter, reddit, clojurians) to stay up to date on the bleeding edge. When I find a good source with more detailed content, I subscribe to their blog so that if I skip social media for any number of months, I don't lose on their developments.


@khardenstine What if we use a core async "bus" to broadcast every change to every grid cell. Changes would be oeav where reasonable, or can be entity diffs. Each cell can do a very cheap filter on the channel for the entities it is interested in, and do an efficient update. 10000 equality checks is nothing A hole in this model is non-entity query results are not efficient, but not clear that they need to be (only pulledtrees are deeply nested and all pulledtrees can represent diffs in oeav)


It is async and streamy, it would never block the cursor, it is deterministic