What is the state of the art in graphical (nodes and edges) programming? Anything general purpose? Or is this limited to a few domains?

Hi guys! I have an idea, I believe it must exists but can’t figure out the exact keywords. First, let me enumerate the keywords which I think are related: • Transactions • OO • Eventstreams • “Transactions in OO” or “Transactions with Eventstreams” • Async & Concurrent computation • Erlang / Akka • STM (too low level?) I want to create an object graph (or eventstream graph?) where objects do not have direct reference to each other, they only communicate via messages. Objects are nodes and there is an edge from o1 to o2 if o2 is subscribed to o1 (o1 dispatches events which o2 listens to). Objects supposed to be “Processes” similar to erlang, objects are some kind of an “Async Computation”. Here comes the “new” stuff: There is also a root “StateManager” node which is subscribed to messages contain data about “UPDATING EXISTING IDENTITIES” like “SetGlobalScore <amount>“, any “Async Computation” (object) which directly depends on changing data must subscribe to this root node. If any data gets altered it does not happen immediately but via transactions. The computation which comes up with the new piece of data and wants to share it sends a message to the root “StateManager” node which handles the update. It must check if the data is in use somewhere down in the dependency graph (which I think might be a DAG with the exception of root node): if it is used, the computation might be interrupted (which is more like an exception in my mind) and the whole state might be rolled back to proper state or “handled properly”. If it is not used the state just can be updated: the root emits the message and it propagates down. This is just the core idea and quite vague, but the general goal is to add some safety to concurrent programming. Ah, and I’d like to do that everywhere, not just on server side but on client as well! 🙂

orleans seems to be close to it: https://dotnet.github.io/orleans/

Another thing: I am curious how do you resonate with “Erlang + BEAM + OTP vs Java + JVM + Akka” stacks? I have almost 0 knowledge on it, read some basic articles on the comparison, but I’d really like to see what are your thoughts on that in this channel. Any theoretical or practical pennies are welcome 🙂

Low-code/no code/red-lang/spreadsheets etc - I have been thinking about no-code apps etc, the problem of how to make programming more accessible. When it comes to making programming more accessible, one challenge that I see is combining the UI, data store and calculations. It is a challenge to combine these in a way that is accessible to the masses. An example of the challenge is no-code applications. There are many no-code applications that make it easy to edit the UI and data store. They offer an easy to use solution with very little learning. But don't make calculations available. A similar problem with most low-code applications, the calculations part of low code applications is typically not accessible in the way that a spreadsheet is, for example. ... Or there is red-lang, it does a good job of linking UI and calculations, but the data store is, as far as I can tell, a regular external database. ... Or take a spreadsheet, it is good at data storage and calculations, these are both accessible with very little learning, but the UI is a problem, you can't create a regular UI with a spreadsheet, you need to use the grid. Or look at scratch (drag and drop coding) it makes calculations and game style UI available, you can create complex games with it, I was surprised how complex, and it is easy to learn, but does not do data storage or data i/o style UI. .... The hard thing looks to be making UI, data store and calculations work as one. So the calculations can do validation and control writes to the data store. So the calculations are able to produce a responsive UI that guides the user. So the data from the data store is visible and editable in the UI. ... it looks to be a hard problem to make the UI, data store and calculations work together. It can be a challenge to do this in a good programming language, and very hard to create something that is easy for people to learn.

@Kartik Agaram is on Hn frontpage again :)

Who thinks this is a good opportunity to evolve how programming works by bringing multiple cultures ( industry/end user) into a discussion about how social (online) meida could look like?

This is on Hacker News right now. Really impressed by it, and I think it's of interest to this audience. https://www.anishathalye.com/2019/12/12/constraint-based-graphic-design/

Anyone seen Roam Research? It's a newcomer in the notetaking Notion/Evernote/etc space. It's been getting rave reviews on nerdy parts of Twitter. The focus on alignment with thought-process seems FoC-like. Not really Future of Coding specific, but it appears to be an adjacent topic here (based on past conversation) https://roamresearch.com/

Hey guys! Working on a node editor here and was curious to know if there are any similar projects out there that use auto-completion ? The idea is that you should get suggestions on what might be the best nodes that your current selection needs, regardless of whether you have those nodes in your graph or not. I can think of this vastly simplifying the loop concept for instance, where you'd suggest the user a loop node whenever they try to create a setup that requires one. Another utility here would be suggestion highlights of possible connections that you can draw between a currently selected socket and all the ones in your graph. Would love to know if there are projects out there that use a similar approach, just to make sure I'm not reinventing a wheel but improving one.

Here's a well-written analysis of the state of Open Source software. Probably of interest to many in this group. http://marktarver.com/thecathedralandthebizarre.html

hello! I’m a fan of mind maps and I’m looking to find out more about related ways of visualising information. can anyone recommend any?

do you know about plan 9? what do you think about it? does it add anything to this group's objectives?

Hey <!everyone>, big news! There's a leadership transition in the works here. Read all about it: https://www.patreon.com/posts/32445075. Comments and questions welcome :)

hi, new experiment! 🎉 Future of Coding Weekly Newsletter, subscribe here: https://tinyletter.com/marianoguerra/ want to collaborate? check instructions here: https://github.com/marianoguerra/future-of-coding-weekly collaborate to the first edition here: https://github.com/marianoguerra/future-of-coding-weekly/issues/1 feedback/promotion more than welcome!

PODCAST I'm working on new episodes! The first will come out in January. I'm hoping to keep a monthly cadence thereafter. The biggest constraint will be time — it currently takes me over 20 hours to make each episode — so I might adjust plans or process as we go. That said, I'm going to do everything in my power (as a music-person who owns a lot of microphones and fancy editing software, haha) to push the production quality of the podcast up to "professional" levels. (If you someday hear theme music, don't bristle — it's going to happen.) I'm also interested in trying out other formats, in addition to the "interview" and "research recap" episode styles Steve patterned. If you have suggestions for programming, production, formatting, ideal episode length, etc — I'm all ears. I've got a slate of guests lined up that I'm really excited to share with you, but I'd also like to keep the podcast very closely associated with our community. If you would like to come on the podcast to talk about your FoC project, shoot me a DM and I'll add your name to the list. Like many of you, I came to this community in the first place because of the podcast, and I'm honoured that Steve asked me to take up the torch. I'll do my best to keep the spirit intact while improving the things I can improve and despite losing Steve's unique perspective. Yes, that means more talk of visual programming than before. Yes, I will make a sincere effort to give static typing its due 🙂

COMMUNITY Just this morning, Jonathan Edwards wrote something that's stuck in my mind like a burr:

Slack is a terrible place to have conversations that last more than a day

This resonates with me, and I know a lot of you feel similarly. We've all been in other communities hosted on other platforms, and we constantly feel the comparative weaknesses of Slack. At the same time, moving the community will lose good people, and I'm not sure we can survive that sort of a seismic shift. We could also introduce a second platform as a complement to Slack, but that doubles the effort needed to keep up. And besides, this community has remote colonies on other platforms already — Lobsters, Lambda the Ultimate, Twitter, HN, Lines, XRA's Discord, and countless others. What we have right here is what it is because of Slack. The aggressiveness with which Slack forces us to lose touch with past discussions and miss out on things happening in threads has the Twitter-like consequence of restricting how much time you can spend reading. For some that's a blessing, not a curse. If we moved to a platform with better support for long-form posts and rich threads, that would surely make it easier to read everything everyone writes, but I don't know how many of us could afford to spend that sort of time. All said, I'm very sympathetic to the position of Jonathan (et al.), and want to keep the search for alternatives alive. I like the idea of starting up a shadow community on another platform just to get a feel for it, so please keep posting suggestions for that in #meta. If enough of us are serious about moving, and we manage to agree on a spot, and we try it out and it feels good, I'd be happy to make it an official home of the community. I'm excited that Mariano Guerra is starting a community newsletter — I really hope that takes off, as it'll probably help any eventual change to the platform our community calls home. If you have any similar project you'd like to try out on behalf of the community, I have a stamp of Official Blessing that Steve left on my desk and I'm eager to use it. Of course, if you have concerns about anything that you'd rather not discuss openly, please send me a DM. I don't have any plans to explicitly change anything about how the community functions, but I am very open to new ideas. I love the energy and activity we have here. I love the split between academic research and industry practice. I love that folks are organizing meetups and helping each other find jobs. I love that many of us are working away on our own FoC projects that will take many years to come to fruition, and we'll get to share that difficult process together. Thanks everyone for making this place what it is, and I hope that we can keep it going and growing for years to come.

So VS Code is going to be more than a text editor... if I were going to make a custom editor, living in VS Code would be attractive. Looks like one must write it in TS/JS. "We continued to work on the custom editor proposal this iteration. (https://github.com/microsoft/vscode/issues/77131) As a reminder, custom editors are webview-based views that can be used in place of VS Code's normal text editor for specific file types. As of VS Code 1.41, Custom editors can now integrate with VS Code's Undo/Redo and Save functionality, which enables many interesting use cases including using them as visual and WYSIWYG editors. You can find the current API proposal in 

vscode.proposed.d.ts

(https://github.com/microsoft/vscode/blob/master/src/vs/vscode.proposed.d.ts) and we have also put together some very simple extension samples that demonstrate using custom editors for text and for binary files. (https://github.com/mjbvz/vscode-experimental-webview-editor-extension) If you are interested in custom editors, please share your feedback on the current proposal and let us know if you would be able to implement it."

Just met up with @irvin on https://hubs.mozilla.com/ to talk about visual programming languages. The VR was good for being delivered over the web. The sense of presence was great. The world was full of weird.

Reminder! tomorrow the first issue of Future of Coding Weekly Newsletter goes out!, subscribe here: https://tinyletter.com/marianoguerra/ want to collaborate? check instructions here: https://github.com/marianoguerra/future-of-coding-weekly collaborate to the first edition here: https://github.com/marianoguerra/future-of-coding-weekly/issues/1

First newsletter is out! it should arrive to your mailbox soon, you can check it here if you are impatient or tinyletter doesn't work 😛 https://tinyletter.com/marianoguerra/letters/future-of-coding-newsletter-2019-12-week-4

feedback more than welcome, I want to maximize its utility, let me know if it display correctly, if the content makes sense etc.

It’s important because it’s a way to conceptualize why neural networks are in a way better than classical ML algorithms. Because this lack of leaks means that anyone can play around with them without breaking the whole thing and being thrown one level down. Want to change the shape ? Sure. Want to change the activation function ? Sure. Want to add a state function to certain elements ? Go ahead. Want to add random connections between various elements ? Don’t see why not… etc.

I truly think that they are among the first of a "new" type of mathematical abstractions. They allow people that don't have the dozen+ years background of learning applied mathematics, to do applied mathematics.

In a social gathering, someone (a music/visual artist) asked me the following:

When you see a working application, can you tell how the code is written and organized?

What do you think?

I'm intrigued by the idea of multiple syntaxes that can map to the same AST, such as used in Forest: https://github.com/forest-lang/forest-compiler#FAQ I'm curious if anyone has seen something like this where the syntax structure remains the same, but keywords are translated to different natural languages. As in Forest, a layer between the code repo and editor would parse the code in its original language and translate the AST into the user's language, and do the reverse on the commit to the repo. One would have to take a bit of care to ensure that there are not namespace collisions between keywords and other names, but otherwise it seems possible to do. I'm also curious if anyone has taught an English programming language to non-English students, and how much the language mismatch was a burden?

If someone has an "elevator pitch" for their project I can add it as "project of the week", comment here or DM me to set it up 🙂

I am quite new to this Slack group, and trying to find out if anyone has a document or overview which lists all the projects here, a brief description and status of the project, the people involved, as that would make it very easy to navigate the group? Or is this something I should create a shared google doc and we can all update it in the group?

What

ideal

tools would you like to have in every project you work on? My list: • Graph db - I’m fed up with Relational dbs and Mongo like key value stores. (OpenCypher as query language would be really nice 😄 ) - offline-first support • Same language both for server and client (like typescript) • Static typing (I’d prefer Dependent Types like Idris, but the ecosystem is just not that mature) • Actor model, visualisation: like XState • Over The Air update support (for mobile apps) • Visual (Snapshot) Testing (like percy.io) • support for Property based testing & random tests