I am not sure, deno seems really bleak as I had read the HN post about deno actively shrinking its datacenters and fresh activity.
This seems an attempt to show to the world that its not that, but I would say that it's still not working.
Deno deploy, their product is better on supabase or even bunny cdn which provides more locations than deno itself.
Bun also exists. And did we forget about node itself?
Cloudflare workers is a beast if we can work with non standards/ I personally feel a little bit that some code meant for node won't work on cf as much as it would on deno
hey Andy from Deno here. thanks for your comment. we reduced our Deno Deploy regions not because Deno is in decline. (in fact we have seen about 2x adoption since release of Deno 2.) it's just that we noticed more users using Deno Deploy for hosting applications vs. edge functions, which was our original vision for Deploy. in many scenarios, application performance is improved with fewer, more highly trafficked regions vs. many spread out but more idle regions. we will be covering that in more detail in a dedicated blog post next week.
Wow! Didn't expect anyone from Deno team to respond. I just thought I was shouting stuff in the void for my fellow HN people at most reading it.
Thanks for the clarification.
I have some questions though:-
Aside from the fact that deno allows npm compatibility, how is deno deploy any better than cloudflare hosting which I currently use, since cloudflare workers also have some compatibility layer I suppose and cloudflare workers can and do host applications and edge functions both.
I can understand if things like nextjs which don't run as smoothly on cf workers, if at all (I haven't tried them but I do know that its easier to run nextjs on vercel than any other competitor), but nextjs can run easily on deno, so that might be a really big niche tbh
But as a sveltekit user, here's my opinion
I have deployed so many sveltekit websites to cloudflare workers, 100k limit never disappoints me or limits me, cf workers also has a kv which is good for simple databases. I am seriously considering to always use cloudflare workers since currently I am just a student and it gives me 100k requests per month limit and after that its still really really cheap.
I had compared deno's tier and cf tier's sometime ago and cf was the winner there, I don't know what's the situation now but I am willing to hear.
Cf workers wrangler developer-experience is genuinely decent in my opinion, not as easy as deno deploy I suppose but still its worth it given all the previous points.
Deno is really nice compared to node, I in fact was the guy who watched fireship video and then the deno video itself and then I legit went to my brother's room knocking on his door (my brother is also a coder, in fact he knows his stuff whereas I am just this 16 year old student) and I legit wanted him to try out deno. But I am having mixed opinions now and I don't mean any defamation of deno team who are really nice and kind for what I've seen.
Have a really nice day Andy. Hope deno achieves success.
There are plenty of moribund open source projects. But looking at this month's 2.3 release, Deno's future strikes me as the opposite of bleak.
As for Deno Deploy, sure —¹ it may fail. But regarding the Deno folks scaling their footprint in response to changes like real-world usage patterns, adoption timelines, our rapidly-contracting economy, or whatever, that's a far better response than ostriching in the longer run.
¹ This em-dash was lovingly hand-crafted by a Mac user. Please don't em-shame.
I meant deno's future as of how the company would survive with competitors/ the ecosystem growing and node adopting. Because the company is basically the only one "really" working on the product, if the company can't find a decent source of revenue, (I am not sure "what" is the product that deno sells that others can't / aren't selling already ie. bunny cdn / supabase hosting deno subhosting)
I would love deno to win, maybe be even faster than bun and maybe help with permissions a little bit more as I played around with deno scripts and it quickly devolved into a permission nightmare from my personal experience, maybe I am inexperienced, but I didn't had the issues in bun.
Same, my typographically-savvy friend. This "dead giveaway" too shall pass — first it was certain words, today it's em-dashes, and tomorrow we'll be forced to introduce grammatical errors.
I might add it in the original comment attributing it to you as well. Thanks again.
Edit: wait I can't edit my original comment, I think it might be because its already commented on by people, so yeah this is the only way for people to find out about it.
Yes, I think that is why I had written "have we forgotten about node" because most people seem to forget node given deno/bun (including "me"), so Of course, node won't have deno or bun occupy its market share and it would also add those features.
Its a net positive for everybody except the people who are working in alternative implementations unless they are being sponsored or they are doing it for fun.
As a company, I am sorry, but I just see no ways of pure revenue once node improves
This seems an attempt to show to the world that its not that, but I would say that it's still not working.
Deno deploy, their product is better on supabase or even bunny cdn which provides more locations than deno itself.
Bun also exists. And did we forget about node itself?
Cloudflare workers is a beast if we can work with non standards/ I personally feel a little bit that some code meant for node won't work on cf as much as it would on deno