In the browser, you can console.group commands. Always wished this could be managed more easily with logs, in a way that you can expand on the logs that you want to dive into

https://developer.mozilla.org/en-US/docs/Web/API/Console/group

@Shawn McKay cool, I didn't know that. Grouping log messages and getting some hierarchy into them would be really cool!

That looks great Felix Kohlgrüber, really looking forward to how the prototype evolves!

Thanks Daniel Garcia! I don't know where these experiments will lead, but I've got lots of ideas. Might take some time though as it's only a side project.

It's really interesting how much potential for improvement terminals have once you start thinking about it. If my experiments just showed some people that terminals aren't "done" and inspired them to think about innovations, I'd be happy with the result. That being said, I really want to make this a proper project I can use daily though.

Let's see where this goes 😉

Nice!

Do you plan to keep working on this? If so, what sort of feedback (if any) would be helpful?

Here's a mature project that supports this: https://github.com/patriciogonzalezvivo/glslEditor

To try out glslEditor online, see https://thebookofshaders.com/

Chris Maughan it would be cool to have these features in Zep (which is C++/ImGUI, not JS, in case you don't have the context).

Yeah Doug Moen Thats entirely the plan 😉 I wrote an 'axis viewer' a while back (well, ported one from elsewhere), for that very reason. I'm thinking about spending a couple weeks on Zep squashing some bugs, cleaning things up and trying to add better embedded widgets (I already do have embedded sliders as a prototype)

My interest in using Zep has just increased. Thanks Chris.

I'm curious why people like this particular interaction. IMO it's hardly discoverable, it lacks reference values to know how far to drag and feels like poor UX.

What others think about it?

To avoid beating up on Dominik, it might be worth discussing scrubbing numbers in general (as opposed to this specific project) in a separate thread in #thinking-together. I think that'd be a great discussion to have. Daniel, would you be up for starting that discussion there?

Oh wow, this thing blew up a bit, interesting conversation though.

Anyways, Ivan Reese I might spin it into an SVG learning tool.

SVG path syntax for instance is extremely powerful but quite cryptic and it helps greatly to have immediate feedback.

Furthermore I just wanted to bring more attention to this idea in the context of online code editors, and it has garnered some steam on twitter.

Doug Moen That GLSL editor is very cool, thanks for the link, liking the context aware widgets.

As someone who regularly edits large (eg: 1k to 100k) SVGs by hand, I absolutely see the appeal. If you end up making that, be sure to share it here so we can follow your progress :)

Ivan Reese Awesome, great to hear there could be a real use in that, will keep you all posted for sure.

Reminds me of a more flexible version of https://storybook.js.org/

Destroying the editor using the editor reminded me of this:

I HAVE NO TOOLS BECAUSE I’VE DESTROYED MY TOOLS WITH MY TOOLS.

(https://www.usenix.org/system/files/1311_05-08_mickens.pdf)

Nice work!

Just a small note about GT: it’s not for Pharo, just based on Pharo 🙂.

Great! 🙂

I'd recommend putting waypoint into the far left navigation panel (where search currently exists). It's an option as of one of the last few releases.

You could even configure .vscode/settings.json to hide the default search and use your extension.

I think WayPoint could be tweaked to become a more powerful "replace", which I could see using

this looks amazing! are there any gifs or videos of it in action?

Yes. There are few here - https://waypoint.netlify.app/

I actually saw this on twitter somewhere. I haven't figured out how to fully utilise it yet but I've found it most useful when trying to navigate code bases that I haven't entirely written myself. What would you like to add in the future? I'm excited to see where it goes!

I often find myself digging through open tabs to go back to a particular function I was reading previously so I can connect the dots.

So I want to explore If I have a timeline visualized with all the files (and symbols within those files) I have been in the past, I can jump back with a click.

Very early sketch of how something like that would look like. I have no idea how useful this would be but I’m curious to try it out.

Yes, I have that same problem :amiga-tick:

It could also be good for refactoring. As hinted at above, when renaming code or moving to seperate files, a simple find & replace doesn't always do the trick. I imagine it could be quite useful at saying this module has now been renamed and moved to this location, please magically update all dependant code.

This is a subject close to my heart, and I've had thoughts about working on a similar searching tool (either general or more specifically c++) within Visual Studio Code. I really like Visual Studio Code, but for me it lacks the killer feature which makes Visual Studio (VC) better in my eyes. In VC you can hit 'CTRL+,' (comma) and get a search box. But the search is completely general, and fuzzy matched. VC will search symbols, file names, function names, class names, etc. etc. Effectively this somewhat 'hidden' feature lets you jump to anything you need while programming - instantly. It is perfect because:

• Fingers never leave the keyboard, it takes < 1 second to jump
• Fuzzy matching means I can just type fragments
• Searches everything relevant.The built-in search in Visual Studio Code unfortunately searches for either symbols or files depending on which command you invoke. There is no combined search. And in fact, I have not seen the same thing as VC in any other editor.

I'm not a JS programmer, but I will certainly play with your project with interest!

Shoutout to MercuryOS for lots of inspiration here!

If you are interested in our underlining architecture to pull this off please ping me — It’s Rust and ECS 😋 and we are hiring. We also stream on Twitch from time to time at https://www.twitch.tv/storyscript

Checkout a list of Skills we thought of https://airtable.com/shrsiPAT6DcKbm6Vm

Did I miss something — is there an actual demo of this tool that I was supposed to see? Or should we give feedback just based on your description of the tool from the video?

Just based on the description.

To be clear, we build the workflow engine that powers this — That is demonstrated in our Twitter feed and other FoC casts. This original post was to share ideas, very early prototype wireframes of a implementation concept. No demo necessary as it should be obvious enough, and if it’s not please let me know because we missed the mark.

Coming full circle. Back when Siri was a DARPA project called CALO, serving as an umbrella for many interests, our group at IHMC was interested in roughly the same. Let's see, a link... ah. https://www.microsoft.com/en-us/research/video/plow-a-collaborative-task-learning-agent/

I would argue our approach is different. But that's a debate we would need to have in voice. Storyscript is a progressive NLP (where NLP is secondary) development OS (broad use of the term). But as they say, sometimes the best ideas are recycled at a better time with better packaging

I was going to draw a comparison to OpenDoc. Without a specific prototype, or even a mockup, to look at, the best we can do is take the broad strokes picture you've painted and relate it to the more concrete thing we're already familiar with. Makes it hard to offer substantive feedback.

Absolutely different. A real product is very different than than some somewhat similar research that never genuinely went anywhere.

🤔 my questions were very clear: what small pieces of software would you make and how would you extend a site (in other words of course). Is that hard to offer feedback on? I've provided examples in the Airtable to help the narrative along. 🤷‍♂️ What am I missing here?

From my experience; it's often I find myself wishing I had more functionality at my finger tips from websites. This concept, one could argue, is the entire "MicroSaaS" movement....

(I've never heard of the "MicroSaaS" movement /tugs collar/)

When I find myself wishing I had more functionality, it's not really the sort of thing that would be exposed as an API — either it's because I'm doing something counter to the wishes of the people to made the website, or because the thing I want isn't laser-focussed simple like most of the the examples in your Airtable.

It's usually frustration about the functioning of some website:

• Bypass this paywall
• Render this JS lazy table view in full as static HTML
• Pre-load this infinite-scrolling list (eg: my twitter timeline)
• This JS-powered button isn't working because my internet connection is slow — fix it
• Redo the layout of this page so that it doesn't have all these empty spaces where my adblocker prevented ads from loading

Or frustration about the limitations of browser history search:

• I want to sew a mask — show me all the tweets I've seen in the past 6 months, then find all the ones that talk about or link to studies of the efficiency of different mask designs at blocking transmission of COVID
• I need to win an argument against my idiot father-in-law — show me every graph I've seen in the past 3 years tracking the effects of climate change
• I feel like I've seen this idea before, but I can't remember where — show me all related articles and papers that I've previously read or seen links to (even if I didn't click those links)

Or is just the desire for my own websites to be more useful without me having to do the hard work:

• Change the structure of my personal site so that all my pages are organized in a single chronological list, rather than on separate lists by year
• Let me know when my website gets more than 5x the average traffic (without me having to build an API to expose traffic, because that's the hard part)
• Let me know what sorts of searches people are doing where they should find my website, but aren't

Looking at your "Desktop" website in the Airtable gave me this thought:

• Normalize and compress (ratio: 4, threshold: -20db) all the audio of the Two Minute Week videos, so I don't have to adjust my volume as much. Update all videos in place, so nobody else has to adjust their volume either.

^ That might seem like an unreasonably big ask, but here it is rephrased in terms of pseudo-APIs for things that exist on my computer:

• When I view a video embedded in or linked from the #two-minute-week Slack channel, tell https://rogueamoeba.com/soundsource/ to add an AUDynamicsProcessor audio unit to the System output stream. When the video ends, remove the audio unit.

The examples mostly make it look like you want to build an NLP IFTTT, but then there's some stuff very outside of that ("make a figma component by voice"). This leaves me unsure where the feasibility boundaries are located for your system.

the examples on the cover image of the video made me think of zapier, if it works in desktop apps it would be interesting for people doing repetitive work across apps, for example my brother uses excel, whatsapp, google maps and a web app that calculates routes across points and has to copy paste and format data across the apps if he could automate those it would free him up of a lot of repetitive work

it could be the unix pipe of desktop apps, the problem is that desktop apps don't expose data as uniformly as the web or cli apps...

Thank you to all that watched the video. I make no apologies for sharing ideas pre-product as the best time to share ideas are now. I think if we all shared ideas more before engineering and having touch/honest conversations we would get further along.

I honestly think the feedback is missing the point here. I’m hearing “this remind me of Zapier” while the entire UX is different (which is most important but often overlooked). The subtitles of this project are also overlooked/ The fact that we can capture the users intent at point of thought — I’ve never seen before, especially to this caliber. Zapier products alike and the entire software industry require people to “hold-on” to the thoughts and translate them into a medium that is not natural (unlike voice or text).

As for the confusion about bounty. We have examples of simple “invert this image” and “create some Figma shapes” — These are very different… are they? They both have input and output. Image goes into box, inverted image comes out. Voice/text goes into box, Figma shapes come out. Both of these have been done before, both are products, both are solutions, both are hard to discover and hard to use. You may disagree with how hard, but take a moment to reflect on that… think about what you have to do before you can leverage those technologies, think about how far you are from that software, think about the people that are not as brave as you to install several extensions, use the terminal, login to Zapier and diagram a simple, yet for some reason complex ask of the computer.

I hear complacency here. Simple tasks take people a long time because we believe that is how things are; but we can change that. After thinking in this domain for a while; I’ve changed from complacency to annoyance of how computers work FOR me but now WITH me.

I’ll leave saying this: It’s not what you do, it’s how you do it.

I see where your approach is simpler and more expressive than zapier, but still reminds me of zapier/ifttt "recipes", your customers will also try to understand it in terms of things they know, and it will make things simpler for you to do the same when explaining it to them "it's like X, but Y" works better than starting from nothing.

I make no apologies for sharing ideas pre-product as the best time to share ideas are now. [...] I honestly think the feedback is missing the point here.

The sort of feedback we can give you is limited by the clarity of your description. Jack Rusher spoke to that really well. It feels like you have a really clear vision, but your description of it doesn't give us much to respond to. This is why people like to share mockups and prototypes — they are more of a surface for our impressions to stand upon. People who share prototypes get better feedback. Instead, what we're doing here is more like sharing what we all think a cloud looks like.

I hear complacency here.

Please don't belittle us. We're trying to respond to you in good faith, despite the dear lack of clarity in what you're asking us for.

the main barrier to adoption for your product is not people overlooking the subtleties, it's the same reason why I don't use zapier/ifttt either

what would users type in google when they search for your product? if it's unlike anything else, what do users search for? how do they know they need it?

maybe starting with "what if you could program the same way you talk to alexa? (but it gives you the right answers more often 😛)"

I don’t use Zapier/IFTTT either because it feels like more “work” than it’s worth. It’s not coming into play at the right moment of thought. If those tools were more in the moment of thought I would be a power user.

you could then start with what alexa/siri recipes are more popular, even going a step further, search what are the most common tasks on mechanical turk or real virual assistants (usually they are used to automate some sort of web scrapping)

Alexa/Siri is amazing because it’s at your finger tips and they will only get smarter and smarter. Why is software not at your finger tips?

The barrier to entry is a steep mountain. This prototype attempted to express what it could be like to access software (regardless of functionality) at your finger tips. It’s right there, waiting to be leveraged, without an expectation of switching context into yet-another-tool that fails to meet me where I’m at.

one of the unofficial internal mottos of our product is "what if creating dashboards and visualizations was so easy you would create one to answer a question and then throw it away". In your case it could be the same, what if automating tasks was so easy you would do it even for one result

basically be the answer to this: https://xkcd.com/1319/ and this https://xkcd.com/1205/

I must ask: was that obvious from my video and initial question? I’m having a hard time empathizing with the feedback from this thread because it’s so quick to “this is like that”… clearly I dislike when people do that — “this is like that” IMO is belittling in the same way I said “I hear complacency” (CC Ivan Reese)

Fundamentally, I see an industry of products that get in the way more then they “work with me”. It’s a symptom of their lack of context and poor interoperability. These charts that Mariano Guerra express the same stress I feel.

I believe I’ve expressed a novel approach to (a) meet users where they are (b) discover software/tools/features/skills as easy as Alexa (c) automate without changing context (d) show, mostly vis Storyscript core, how interoperability looks like

I understand the frustration one gets when others don't understand what you try to express, but isn't that part of the feedback? if it doesn't become clear to the FoC community, how probable is that it will be better received by the programming community at large or worse, by non programmers?

There was a conversation a while ago here about "how much you automate" and most of the responses were "not that much", and that comes from programmers that joined this community, it probably gets worst "out there"

I understand the frustration one gets when others don’t understand what you try to express, but isn’t that part of the feedback?I’m sorry, but I disagree here. I think we are too quick to compare products without taking a couple minutes to think more deeply. I’ve shared this product to other communities and people who 1:1 and get tremendous feedback beyond “well this looks like Zapier” when it’s nothing like it. What’s frustrating is the shallow feedback, sorry to be blunt. At Amazon, people present one page (no demo) ideas that get scrutinized deeply. I understand FoC is not Amazon…. at times I wish it behaved more like it however. Where even the smallest ideas got more attention than blink feedback that aims to compare more than ask questions.

This is also rampant in HN replies… people just “this is like that” constantly. Sad.

Sad.

We've https://futureofcoding.slack.com/archives/CCL5VVBAN/p1597990388025100?thread_ts=1597966470.010300&cid=CCL5VVBAN with substantial https://futureofcoding.slack.com/archives/CCL5VVBAN/p1597990756025300?thread_ts=1597966470.010300&cid=CCL5VVBAN that https://futureofcoding.slack.com/archives/CCL5VVBAN/p1597996491026000?thread_ts=1597966470.010300&cid=CCL5VVBAN to meet you halfway, AND meta-level feedback that what you're asking us to do is avoidably difficult because of your vagueness, articulated concretely and amicably.

You've responded with you all don't get it, other people do this better than you. That won't lead to better feedback, but it will lead to people resenting you for being flippant about our time and attention.

There are lots of different types of feedback, and I think the way you've presented this makes it hard to give deep feedback. Case study:

• Your highlight line was interesting enough, and makes it sounds like you've already made a tool. So I clicked the video.
• It's far too long for the interest the headline generated, so I skipped around. Looks neat, but I didn't really glean anything by skipping around, beyond some of the obvious: more nlp-like zapier.
• Final thoughts: "neat, I guess, I wonder what the comments say"

You should try a different form factor than a video. Videos are great if you're already hooked. They're terrible for the first introduction to a product, unless it's more like a trailer/teaser sort of thing.

If you wrote a short insightful blog post, I'd be more likely to skim over that and offer ideas. I think you've also perhaps applied the "fake it til you make it" strategy in the wrong place. This is a community that values clear and distilled thoughts.

I’m not trying to sell you anything or trick people, not the slightest. And I don’t agree that I’ve been saying people are wrong; I’m expressing my concern of shallow feedback. I want to be wrong — it’s your job to point it out, that’s the point of feedback, to collaborate on ideas, ask questions, and inspire a diversity of thought. I don’t believe I got any of that. Instead we have debated politics and feelings. Clearly we are not going to have a conversation about the substance of this idea (regardless of how the format of presentation is) instead I’m going to retire from this thread as it’s getting far to personal. If anyone wants to give critical feedback, ask thoughtful questions, and inquire about our approach I would be delighted to continue a conversation.

Steve Peak as someone who’s been down this road before, please do take the meta comments you’re getting to heart. You will only cause yourself more and more pain, and push yourself away from the communities that can be your greatest champion if you continue to engage this way.

Now let’s address the primary frustration:

people just “this is like that” constantlyThis is something all of us just have to accept. It’s how people engage with anything new no matter what it is. We are pattern matchers by nature and so we will always try to draw analogy to what we know. If that takes you somewhere you don’t want to be then it’s time to take a hard look at the message or reframe yourself such that you can lean into that analogy. Attempting to be inherently novel is going to cause people to dismiss you much more quickly than just doubling down on people’s understanding and helping them move past it. At the end of the day, that’s your job as the chief of evangelist of whatever project you’re pushing forward. Your goal isn’t to get people to fall in line and make them see things the way you do. It’s to help them engage with your vision meaningfully on their own terms and find the common ground that will ultimately turn them into champions. You accomplish that not through frustrated replies and throwing stones, bu t by genuinely listening and working “with” them rather than trying to convince them to work “for” you.

Thank you Chris Granger — I’ll be honest that I do feel driven away and turned off based on this and other conversations. This likely will result in me leaving the community which I consider myself a thought-leader valuable member of. (Correction: apparently thought-leader is not appropriate and has a sentiment I was unaware of)

I honestly thought I did a fine job at describing our approach and pointing out the novel behaviors. But why does novel behavior “cause people to dismiss” If that is the case then I’m 💯 in the wrong community. I thought we were the “Future of Coding” not “Today, in Coding” — I think novel ideas should be the norm here, the expectation, the rule… not the exception.

I’ve since had deliciously heated debates about this topic with other communities that has resulted in really good feedback that changes our perspective and improved our approach. We should ask ourselves: why did that not happen here?

The response here is roughly what I would expect in a well meaning audience who’s genuinely trying to be helpful, but honestly doesn’t find what you’re presenting compelling. I don’t think there’s anything wrong with that. The message fell flat - the fact that it didn’t elsewhere isn’t a statement that this group is somehow doing something wrong. It just means you didn’t ignite this group’s imagination. There’s something to learn from that. You could say “well the FoC group just isn’t the audience” and maybe that’s true, but I somewhat doubt that’s the case. As was said earlier, many of these folks value something that’s further along because ideas are cheap and many of us have seen how hard it is to make that beautiful slideshow into something that actually works. I honestly think this would’ve gone very differently if instead of calling this a “prototype” you had just started a thread ruminating about a concept and then explored the space together with people rather than trying to force things in a specific direction.

You’re basically trying to do market research, which if done directly feels out of place here and won’t likely be well received. But there are ways of engaging purposefully with the community where they are getting as much out of it as you are. You’d still technically be doing market research, but in a way that’s beneficial to everyone.

If I had to pick one thing of “what we/I are doing here” — It’s inspire and find like minds to join our team IE recruiting. Just being honest. I hope however that the secondary effect of I/us participating in this community is that we spark imagination, inspire others, and can also provide valuable feedback (which I have countless times before, I can provide references if necessary).

It’s not market research. I’m not seeking approval either. I’ll get that both of those with users when we launch — or we will fail. Either way. I was hoping this community would provide insightful feedback and ideas to help each other find success (again what I have done with others, I can provide references if necessary).

I was hoping this community would provide insightful feedback and ideas to help each other find success

Steve, I empathize with the frustration you're feeling. After putting many sleepless nights of work and thought into Eve, we had conversations in the Eve community that went in entirely unexpected directions too. That said, communities like FoC and Eve are folks who banded together because of a shared belief in something better. They've got every incentive to help each other, and if they're failing to do so, that likely means there's a serious issue preventing them from understanding how. It's particularly dangerous to assume the problem is that they don't care enough or aren't putting the time in -- these folks are dedicating their free time to try to help because they do want Storyscript to succeed. Suggesting otherwise is a very quick way to lose friends and advocates.

It might be a good idea to try and revisit some of the feedback and see if you can find a way to tailor the message to help us engage with the idea more like you intended. As is, it seems pretty clear that there is a market segment full of the best evangelizers money can't buy who aren't connecting with it yet.

I’ve been asked to leave the the community. So I’ll no longer be commenting on this thread. I appreciate those that have helped us along our journey and thank others for their time and insights. I’ll post more in a meta channel about my departure feelings. Thank you

Understanding that this thread is over and Steve Peak has departed, I nonetheless feel it might be useful to add this as a teachable moment for those who might see it.

I went to university to study physics when I was still just a kid, several years younger the usual age, and when it came time to teach incoming freshman I was only 17 or 18 years old. It went very badly. For various path dependent reasons, I found most of the "hard parts" of maths and physics easy and natural, whereas they very much did not. My -- extremely embarrassing in retrospect -- perception was that they were being willfully thick just to vex me, and I was an absolute asshole to them for "not getting it" and "not trying hard enough" and so on.

Later, I read https://creatorsvancouver.com/wp-content/uploads/2016/06/rocky-road-from-actions-to-intentions.pdf conducted by Elizabeth Newton regarding "Tappers and Listeners" that's since gotten repeated a bit in the popular science press, It really helped me to understand what was going on there and improve my own ability to communicate. The TL;DR of the study provided in Heath and Heath's Made to Stick conveys the point well enough:

Tappers received a list of twenty-five well-known songs, such as “Happy Birthday to You” and “The Star Spangled Banner.” Each tapper was asked to pick a song and tap out the rhythm to a listener (by knocking on a table). The listener’s job was to guess the song, based on the rhythm being tapped.>

The listener’s job in this game is quite difficult. Over the course of Newton’s experiment, 120 songs were tapped out. Listeners guessed only 2.5 percent of the songs: 3 out of 120.>

But here’s what made the result worthy of a dissertation in psychology. Before the listeners guessed the name of the song, Newton asked the tappers to predict the odds that the listeners would guess correctly. They predicted that the odds were 50 percent. > The tappers got their message across 1 time in 40, but they thought they were getting their message across 1 time in 2. > Why?>

When a tapper taps, she is hearing the song in her head. Go ahead and try it for yourself — tap out “The Star-Spangled Banner.” It’s impossible to avoid hearing the tune in your head. Meanwhile, the listeners can’t hear that tune — all they can hear is a bunch of disconnected taps, like a kind of bizarre Morse Code.>

In the experiment, tappers are flabbergasted at how hard the listeners seem to be working to pick up the tune. Isn’t the song obvious? The tappers’ expressions, when a listener guesses “Happy Birthday to You” for “The Star-Spangled Banner,” are priceless: How could you be so stupid?>

It’s hard to be a tapper. The problem is that tappers have been given knowledge (the song title) that makes it impossible for them to imagine what it’s like to lack that knowledge. When they’re tapping, they can’t imagine what it’s like for the listeners to hear isolated taps rather than a song. This is the Curse of Knowledge. Once we know something, we find it hard to imagine what it was like not to know it. Our knowledge has “cursed” us. And it becomes difficult for us to share our knowledge with others, because we can’t readily re-create our listeners’ state of mind.When we try to communicate something to a (virtual) room full of intelligent people operating in good faith and find that they cannot understand what we're saying, we have a choice: either accept that we haven't said our piece in a way that successfully translates the knowledge in our head -- which is necessarily much larger than the message we have transmitted -- or we can blame the listeners for our own failure. One way leads to effective communication and positive relationships, the other one to being the same kind of asshole I was in the cautionary tale with which I started this message.

Just finished watching - I especially enjoyed the clip about “reinventing the wheel”, the survey slides about peer work in this space, and this “show don’t tell” approach to getting people to understand the beauty of elegance/natural simplicity in the world from experiencing that in a simulation. A project that I was reminded of after viewing was this “Pokédex for polyhedra”, which lets viewers walk through the space of shapes by applying a small set of transformation operations (see bottom of page after clicking any shape). https://polyhedra.tessera.li/ . Will look through some of your prior videos, look forward to seeing more as this project progresses :)

if you get a web version of some of the demos going, you may be able to attract some open source contributors to port to vscode (the screenshot editor plugin looks like the community plugin for Vega Lite, which I believe was made by a community member and not the Vega core team)

Bret Victor has a dope "Geometric Algebra" sticker on his laptop: https://twitter.com/worrydream/status/1079189354340265984

Super exciting!

I was just thinking about https://futureofcoding.slack.com/archives/C5U3SEW6A/p1595628356371600 a few weeks ago. I wasn't aware of Apparatus, or your work! very cool! thank you 🙂

Ray Imber Your relating AwM to a "node-and-wire" interface is thought-provoking. My first reaction was to draw a distinction – snapping in AwM is a direct action on the artifact being created, while snapping in node-and-wire interfaces is only a direct action on the abstract structure of the program. But it's a really interesting point of comparison, and I think there's a lot of space to blur the lines between them.

For instance – I haven't imagined that AwM would entirely eliminate the "spreadsheet" part of Apparatus. Snaps can't express all possible relationships between parts of a picture, especially when the parts are physically far apart. But where should these spreadsheet cells and links between them live? The node-and-wires model suggests that they could be present in the diagram's canvas, rather than a separate plane (like in Apparatus). Lots of interesting directions to go there!

The node-and-wires model suggests that they could be present in the diagram's canvas, rather than a separate planeA problem with "node-and-wire" interfaces is that they can get very dense, reducing understanding.

Your phrase "separate plane" reminds me of map overlays as a possible solution.

You could represent much of that data as different views on the canvas, that can be switched between or overlaid on top of each other. GIS can be quite beautiful. 😛

Your PL is this? https://hilltop-lang.org/ just checking! 😄

Yup. Needs an update