drschwabe

Where's the drama in that !?

Have you really given GLM 5.2 an honest go ?

When you use opencode to use Claude and Chatgpt models ie- Fable or GPT 5.6 I assume you are getting billed on pure credits? I was under the assumption that you can only use your Claude/ChatGPT paid plans when using Claude Code or Codex. ie- that you would be paying way more via opencode since you would not be getting the extra limits subsidized by your paid plans.

Same here, and if you really need its key features just add JSDoc and Zod to any given vanilla JS codebase.

Passing around mixed types is not necessarily an anti-pattern either ie- you can do type-checking at runtime and use them as conditions for branching into different control flow patterns - so I don't see the benefit of universal enforcement; not that most TS codebases do anyway (ironically).

I personally don't have a need for TS which introduces an extra build step, extra ritual when defining functions, and no performance benefit over vanilla JS. If I want stricter coding pattern with performance benefits I will use C/C++ for the job instead.

Nice, lucky! The Opencode Go GLM 5.2 quota gets used up so fast. It's an expensive model. And while impressive for being open weight, it seems slower than Opus and GPT. So I typically only use it after exhausting quotas of discounted GPT5.5 or Opus 4.6^ paid plans.

Are you micromanaging your GLM costs? It seems the best bang for buck strategy right now is a Opencode Go subscription to get the subsidized rate and then switch to Openrouter's model above and beyond that + make use of a dual model strategy by having GLM 5.2 do planning and Deepseek V4 Flash for implementation.

Sure but if you're really unhappy with your employer employeeing you for 8 hours a day you can also harness this power on your own personal projects to help break free from the 9-5 grind if you so desire.

This isn't a direct answer, because I agree it is difficult to move away from Github's familiarities amongst developers - so it doesn't necessarily solve the problem for a collaborative codebase involving a handful or more developers (complete with Issues, PRs, etc) but if you are working with a less technical/developer oriented team ie- regardless of org size it's just you or maybe one other teammate who are the only ones involved directly in code/PRs then you can fairly trivially roll your own issue tracker or wiki.

Particularly if your work and the employer/client/org is primarily based on a web project (extra points, you already are managing their auth) then you could simply add a new subdomain or route to your existing web project that serves said self-hosted issue tracker or wiki.

Of course these things can get into the weeds but I do think that given the dramatically reduced turnaround times for a competent dev to spin up and customize in-house/self-hosted solutions for basic things like issues and wikis the strategy is more relevant and prudent than ever.

What type of engineer, who until a year ago - because of AI apparently - is suddenly no longer concerned about code? Personally I'm just as concerned about code because AI has not changed the fact that it still takes a really long time to develop stable, secure software (ie- if you're making software to do ambitious things). Nothing about modern AI tools eliminate the need to get in the zone; using AI to amplify one's engineering skills let's us solve the next problem faster - but in software there are unlimited problems.

Too little too late, open source Windows 7 and give it a new 10 year LTS commitment then we can talk.

Unironically JavaScript is quite good for single file projects (albeit a package.json usually needed)

You can do a huge website entirely in a single file with NodeJS; you can stick re-usable templates in vars and absue multi-line strings (template literals) for all your various content and markup. If you get crafty you can embed clientside code in your 'server.js' too or take it to the next level and use C++ multi-line string literals to wrap all your JS ie- client.js, server.js and package.json in a single .cpp file

When I first read the source for his original QuickJS implementation I was amazed to discover he created the entirety of JavaScript in a single xxx thousand line C file (more or less).

That was a sort of defining moment in my personal coding; a lot of my websites and apps are now single file source wherever possible/practical.

Bravo, been looking forward to this but AnimeJs v3 has just been so solid for so long honestly you did amazing on v3 that v4 is just icing on top; excited to try this out on my next project.

async/await + animation (ie- with AnimeJS) is highly underrated.

And mad props for skipping the now dying trend of refactoring entire codebase to TypeScript :)

That's the idea, you won't have to discover - instead you can just create the game you want.

Refactoring your entire codebase just to use that one ES module that is incompatible with CJS is a big pill to swallow if your codebase is ... big (or you have many).

On the flip side we have other amazing devs who are also active with cutting edge libraries and simply do the small amount of extra work to make their modules available in both ESM and CommonJS to this day:

https://github.com/WebReflection/uhtml

Not sure about realtime voiceover and not in emergency services myself but having an LLM fed the call audio to generate supplmentary summary of what was discussed could certainly be beneficial in situations where the operator needed clarity or there was a mistaken address or important detail that needed double-checking after the call.

Indeed, that is primarily where it excels. Your layered source files are SVG and you can export to SVG (and you can import in SVG obviously). The bitmap selection/exporting is also excellent as such you can have these massive vector canvases (with any number of bitmaps and vector shapes/graphics mixed in) and quickly export any slice or selection you make without having to resize the canvas or copy/paste somewhere - and it will remember the export path when you click on the object or layer again later (aside from a bug with symlinks on Linux); ideal for iterative work/exporting revisions to clients or colleagues.

Inkscape is a daily driver for me and its OSS

It has its quirks and even some annoying bugs but where it excells its IMO way better than what any competing/proprietary design tool can do (vector and bitmap exporting, vector+bitmap combined layering, shape/colors/text layout, PDF editing/creating, vector pen tool, etc). I use it to create UI for games & apps and more generally to build sprawling UX scenarios and concept flowcharts.

In more recent builds its performance has become quite good which was a problem it had before. Granted, lots of room for improvement still particularly wish it had more natural flowcharting capability.

Amazon's repo here is open source at least so perhaps this could be remedied with an edit to their README and a PR - go for it !

Agree with you there should be credit where credit is due - I have been using QuickJS for some time and its awesome. For the cost of about 1MB you can get entire modern JS in your C/C++ binary.

Well, if you're talking SvelteKit then it requires a build step so yes, jQuery is way more lightweight.

jQuery is also pure JS whereas Svelte is Typescript so it may be more difficult to debug/hack if your primarily JS coder.

It's good for quick things and prototyping cause you can always swap out those calls with native later. It's API is generally easier to remember/less typing that most native equivalents. You can also use its API serverside via Cheerio to do parsing & manipulation of html without a dom.

edit: also its way more lightweight than React/Vue/Svelte i don't necessarily disagree you shouldn't reach for jQuery if you have a dynamic page (something like uhtml+preact signals would be good if you have fair bit of rendering logic going on) but I would say you should totally try seeing how far you can get with jQuery instead of Svelte/React/Vue on simple pages.

This is a brilliant idea that seems to lack execution.

Ironically now with AI art tech the fans themselves can now carry on the series as they like (including making animated flicks). Not saying this wouldn't violate C&H's IP protections nor that the resulting product will have the same essence but a creative+technically inclined fan tinkering in spare time now has the power to create an all new 'Calvin & Hobbes book' of their dreams. Even if they want to just do it for fun on their own.

Of course, if you were to put in enough effort to make a complete book (ie- 100+ page Sunday comic collections) that could stand among the rest in the original series probably best to adjust your art style, theme and content enough away from the original to make what you are doing essentially a spiritual successor - enabling you to comfortably share it with the outside world.

Will have to play with those more thx for the headsup; I do find however for scribble outlines I like to often draw my own lines by hand instead of an auto-generated one to emphasize the absolute key areas that would not otherwise be auto-identified. Logo and 2D design for example where you may have very specifc text shape that needs be preserved regardless of contrast or perceivable depth.

Yeah that tutorial is decent its what I used to get going.

Note that all of the images in those comfy tutorials (except for images of the UI itself) can be dragndropped into ComfyUI and you'll get the entire node layout you can use to understand how it works.

Another good resource is civit.ai and specifically look for images that have a comfy UI embedded metadata. I made a feature request that they create a tag for uploaders to flag comfyUI pngs but not sure if they've added that yet. Or caroose Reddit or Discord for people sharing PNGs with comfy embeds.

Trying out different models (also avail from civit) is a good way to get an understanding of how swapping out models affects performance and the results. I've been abusing Absolutereality (v1.81) + More Details LORA because its just so damn fast and the results are great for almost any requirement I throw at it. AI moves so fast but I don't even bother updating the models anymore there is just so much potential in the models we already have; more pay off would be mastering other techniques like the depth map Control Net.

I would say that above all extensive familiarity with an image editor like Photoshop, Gimp, or Krita - will get you the most mileage particularly if you have specific needs beyond just fun and concepting. AI art makes artists better, people who struggle with image editing will struggle to maximize this new tech just as people who struggle with code will have issues maintaining the code Copilot or ChatGPT is spitting out (versus a coder who will refactor and fine tune before integrating to the rest of their application).

Yeah the PNG embedded 'drag and drop to restore your config' is brilliant.

Reminds me of Fireworks which Adobe killed off (after putting out a decent update or two to be fair) which used PNGs for layers and meta ala PSD format.

But its more analogous to a 3D modeller suite like Blender or Maya but with theoretical feature such that you could take a rendered output image and dragndrop it back into the 3D viewport and have it restore all the exact render settings you used instantly back. That would be handy!

ControlNet model specifically the scribble ControlNet (and ComfyUI) was major gamechanger for me.

Was getting good results with just SD and occassional masking but it would take hours and hours to hone in and composite a complex scene with specific requirements & shapes (with most of the work spent currating the best outputs and then blending them into a scene with Gimp/Inkscape).

Masking is unintuitive compared to the scribble which gets similar effect; no need to paint masks (which is disruptive to the natural process of 'drawing' IMO) instead just make a general black and white outline of your scene. Simply dial up/down the conditioning strength to have it more tightly or fuzzily follow that outline.

You can also use Gimp's Threshold or Inkscape Trace Bitmap tool to get a decent black & white outline from an existing bitmap to expedite the scribble procedure.

Consider switching to Linux, I think you will see this trend a lot more.

One theory: TypeScript

Maybe tangential, but I have seen a pattern of projects that go all-in on TypeScript only to become overly complex & lose development steam. I'm not saying that is or will be the case with Codemirror but I do think there is some merit to the idea that JavaScript helps developers code effectively at the high-level whereas TypeScript has a way of taxing that process; detracting from what made the original project come together in the first place.

Of course, there are plenty of projects that have done exceptionally well after going all-in on TypeScript (or from the get-go) so take this with a grain of salt.