Hacker News Reader

Comment by arjie | Hacker News Reader

Very nice to see these dev tools get an exit. e.g. I love `uv` and friends but did consider that perhaps dev tools are just a bad business and then no one will go into making that kind of stuff. Good exits means more of these tools.

I have only used Astro for toy stuff but it seemed neat. Congrats to the team.

EDIT: To put paid to the sidebar discussion below, yes I meant "for instance, consider `uv`; they might do these nice things and go nowhere but now that companies like Bun and Astro have gotten acquired, it demonstrates a future for others; therefore we will get more things like Astral's `uv` and so on". Hope that clarifies.

View on HN

This is Astro, not Astral. uv is Astral :-)

Edit: OP clarified what they meant, I'm sorry for the misunderstanding on my part!

They know, hence why they used e.g., i.e. exempli gratia

I don't think that's really clear. I think we could both defer to the OP clarifying.

For pedantry's sake: neither i.e. nor e.g. would be correct here. You want cf. ("conferatur") to invite a comparison; e.g. is when an example pertains to an instance. In this case uv would not pertain to the instance, because Astro is not Astral.

cf. would invite a fair bit of confusion on an article about cloudflare

I agree! That's why I think it's probably just a confusion between entities. It doesn't make sense either as example or as a comparison (although IMO it makes more sense as the latter).

(For the OP: I'm sorry if I misinterpreted you.)

It's all good. Hardly matters. It was just becoming too big a discussion for something far too minor. Any frustration I had from being misunderstood (primarily self-directed) was alleviated from satvikpendem guessing correctly what I intended.

"e.g." IS correct because uv is an example or instance of a dev tool.

"e.g." isn't used correctly here. It's intended use is as a connector linking a clause to examples supporting that clause. You can't simply substitute "for example" with "e.g." anywhere in a sentence and expect it to function correctly.

Regardless, these Latin abbreviations best avoided entirely due to the surprising number of readers who don't understand them.

For the perplex:

e.g. is latin for "exempli gratia" = for example i.e. is latin for "id est" = that is

A good way to remember it is to use a backronym:

e.g. - example given

i.e. - in effect

i.e. - in eother (words)

in explanation

As someone who was perplexed, I've only heard perplex used in past tense (I was perplexed) so seeing "For the perplex" just made me confused as to what "perplex" meant and I had to do a further search to decipher this tree of comments haha

Ha, ha, sorry. English is not my native language.

I'm not sure. I wouldn't generally call Astro a "dev tool". It's more of a framework.

It's possible you are right, but it isn't clear from the content of the comment.

Frameworks are a category of development tool. Things that developers utilitise to be productive.

Like coffee?

IMO saying a framework is a dev tool is like saying a cake mix is a cooking tool, because it allows you to be more productive when making a cake. Sure, if you look at it a certain way, it is correct. But that isn't the way the term is usually used.

It confused me as hell as my first thought was "oh great, astral.sh got bought by a large company, now we've eliminated the last obstacle to using uv in enterprise context" only to realize that it's another company with similar name.

I mean good for them, but it would be nice if the same happened to e.g. Astral (cf.).

I think DevTools can be a very good money maker… I wrote two apps that were basically dev tools and they were the biggest of my money makers. I think it’s easier to make money from dev tools that are “apps” than dev tools that are “fundamental technologies” though so it probably heavily depends on the type of dev tools…

It's probably also easier to make cottage-industry money from a single useful tool and some associated services/consulting than it is to turn the whole thing into a big company expected to do the hockey stick curve thing (eg Docker).

It'll be interesting to see where Astral ends up landing on that; afaik they have a small team and have only raised seed money, but who knows.

I don't know if it's easier. I definitely think it's better, but there can be a lure to VC money that puts a company like Docker on that path. There's a world where Docker is a small company earning individually good money but with no hockey stick curve, and I think that's more sustainable and ultimately better for them.

I suppose they might disagree, of course :)

Docker was not gonna make it as a small individual tool as something podman like would get traction in that market.

Maybe, but I think it could have made a small team very wealthy and successful much more than it did a large company.

It was obviously pitched as an ecosystem/platform play, like "the next vmware" or something, but there was never anything close to a real moat there. Running a registry involved a lot of storage and transfer costs plus spam/abuse management, and private registry was always going to be a better fit for being integrated with CI platforms and the like more than a standalone service with its own auth and billing concerns.

i'm interpreting this in the reverse

if dev tools can only be "monetized" by being bought out, it does not feel sustainable on any level

we will see companies attempt to do things like close source these projects, go subscription based, or just straight up drop support

there is no incentives for cloudflare to make astro better, or even keep it around

same goes with bun, svelte, and i'm sure countless others

Who's to say there's no incentive. Anthropic using Bun internally is plenty incentive to make it better even if for their own use-case. I think it is a bit of a doomer perspective to think anything being bought out means the end of the line for that project. Sure, some things might change depending on the interests of the new owners but that's not to say it'll automatically become bad. Microsoft bought Github and Mojang, they're both doing better than ever for example

GitHub is without question worse for users now.

GitHub continues to become less and less usable (slower performance) while the number of features proliferate. On the one hand, being part of larger company _can_ relieve the pressure to stay profitable. But on the other hand, acquisitions tend to get cannibalized by the companies that acquire them. Profitability tends to become less and less aligned with usability.

Realistically to maintain a modern web framework to even a minimal standard you need a few people working fulltime on it, and more than a few if you want to take it places. There needs to be some kind of long term sustainable vehicle for funding those developers, either a corporate sponsor or a foundation.

So all those frameworks have to end up somewhere, and I’d rather it be somewhere else than vercel, as they already own way too much of the web frontend space.

that very much depends on your definition of "modern web framework".

> there is no incentives for cloudflare to make astro better, or even keep it around

There is - to counter NextJs. NextJs is a pain to host outside Vercel and makes Cloudflare lose customers. In theory that's why Gatsby was bought out.

> same goes with bun, svelte, and i'm sure countless others

Svelte was also taken over by Vercel. To control frontend hosting. Bun isn't in the same bucket. There isn't such competition going on.

Then we'll finally be getting the world we deserve.

Decades of VC cash has trained developers to never pay for anything that powers their entire career like dev tools. The assumption is you can always squeeze some rich business customer that employs the dev.

If AI kills hiring of software engineers, then there's less devs inside businesses to sell to. So we can either pay for the products we use directly, or not have them at all.

It will take a decade to shift penny pinching behavioral habits of devs but slowly over time the market will correct itself. Chefs have always had to buy their own knives. This is good imo.

> Very nice to see these dev tools get an exit. e.g. I love `uv` and friends but did consider that perhaps dev tools are just a bad business

I can't even begin to comprehend what kind of world view you need to have, to think that being bought out by some megacorp with an at-best 50/50 chance of continuing existing products is an accurate measure of a "good business", much less that its the only measure.

> Very nice to see these dev tools get an exit...

You'll see a lot more in the next 12 months ;)

yeah the thing about dev tools is that devs love writing them and your market is just devs

Dev tools that support an ecosystem of things. A dev tool by itself isn’t that valuable unless it’s a backbone of an ecosystem. Astro is. React is. TUI.js is not.

A brand - like charm bracelet - with dev tools is another avenue. Why build one when you can build dozens?

In the end it’s about providing value. Not novelty but value. Make your new overlords better or faster or make everyone better or faster. Provide a value that can be measured.

Reply on news.ycombinator.com