I'm less than convinced. Especially because now, there are models where the new FLoC style associations will now leak interests and allow for certain leaching activity and weird models of competition against niche markets.
I'd rather see television-like models where advertising is targeted/related to the content itself, or general to a given audience over individual targeting at all.
It's sometimes easy to fall into a singular mindset with how Databases should work.
In some environments, the transaction/consistency are king... for others, you relax certain guarantees for potentially expired data, or slower consistency models for performance and scale.
It really will depend on the environment and target audience(s) for your applications. Not to mention how unbalanced reads vs. writes are at scale and this will really shape decisions.
This *could* be useful if you're creating a leaner application where responsiveness on mobile or otherwise very limited devices as a priority.
Outside of the above context, I'm not sure it offers that much above current bundlers, especially if you are limiting support to modern (self-updating) browsers.
Currently, the killer feature for React is the component libraries. While Hyper and related alternatives offer some competitive advantages in some limited ways. When it comes to developing applications, being able to find readily available component libraries and even one-off components is invaluable. Material-UI is a personal favorite, and some of the Bootstrap based libraries are very nice as well. Yes, you can ad-hoc many of these things.
Another advantage, by nature of JSX being a JS abstraction instead of a DSL is that you get intelligent interactions with your IDE, more so with TypeScript. hyperlit won't autocomplete your available component properties for example.
Yes, Hyperapp has a lot to offer, and if you're able to do ground up development, or enhancing existing, or mostly-static/server rendered applications, it can be great. That said, it won't replace with React (or even Angular and Vue) offer for application development with teams of developers.
Would probably lean towards material-design or more common OS styling... thinking in terms of making it easier to fit in a larger component model.
Another issue, is positioning is always below, so no screen/scroll context. Would also test on mobile as it may not work in a portrait mode phone well.
Edit: Found the source - https://github.com/CodeLikeAGirl29/lindsey-portfolio
It's kind of OT, but leaving it so I can comment. On github, should really prune some of your dangling repositories, or unify them into branches on one playground repository... there look to be a few dozen "profile" repositories, so it was king of hard to determine which this was. There's no JS at all involved here. You can make a "codelikeagirl29" repository, and that will be your default for your gh profile, you can also point a custom domain at it (regarding email mentioned below).
---
As to the profile site itself...
First, it's a bit jarring in a desktop browser... Your profile image takes up the full screen before even getting to anything meaningful. Would reduce this down to a smaller section, with maybe a banner/header background image at the top, especially for desktop use. Think similar header/banner to what you would see on social media. Also, consider replacing your personal image with an avatar or caricature, there are times and companies where having actual photos can work either for or against you, so it's your choice.
The single page navigation is fine, but would suggest pinning the navigation to the top once your scroll position would otherwise shift it off the page.
Finally, should have each of the "projects" navigate or expand with more content. Again, it's jarring with the shear size of these images vs. the content on a desktop browser.
Should also consider some of the consistency in implementation, the borders, padding and spacing are inconsistent at best. There are options to get started with from Bootstrap to Bulma or rolling your own, but consistent CSS and markup are key here. There also appear to be some broken html elements.
Would suggest a subtle background image as well. Beyond this, would suggest registering your own domain name, if you use domains.google.com, there's free email forwarding, so you can use whatever@yourname.dev or similar, and forward to your icloud email. On the domain, would go for short over descriptive, something someone is comfortable tapping into a phone keypad. While I'm critical of the desktop view above, a lot of people do look at things on mobile devices, you need to account for both.
Most of the judgement above is on styling and layout only because that is what is presented. I don't bring this up to discourage you. I would suggest taking the time to re-visit it all, put your github and linked in links as social media links in the top-right of the header so it's more prominent, especially if this is for demonstration.
There's a bit more work to go before I would use this in a job search... but it's an okay start.
It's always been one of those things that are powerful if you're using JS for data input validation and the coersion can be useful in that way.
If you're that concerned about it in practice, I would suggest using TypeScript.
Agreed. This just feels like bringing in a library to do something JavaScript already does.
My first instinct was that this would be a rough equvalent to a `Money` or `Decimal` data type/conversion to/from string for use with Larger or arbitrary values stored as a Uint8Array or similar for serialization/deserialization. Or even BigInt with N positions reserved for "decimal" offset or something.