I feel like one of the problems (this might solve) is that the servers can start up instantly since they are firecrackers so I can assume that a service shuts down automatically after some time and then they auto start when a server points to a particular resource aka sleeping (similar to how railway's sleeping mode works)
That being said One of the ideas I like about this is the idea that you can seperate 1 server into multiple chunks for fast loadouts using firecracker
I have built something for my own internal use case on top of firecracker + automatic ssh and I will probably share it in the future but the key idea I like about shellbox is that I can see this one service on which I can build a product, have a genuine use case and they mentioned open sourcing (I hope they do, mine implementation of firecracker-ssh uses bottlefire + golang/ssh library) and you can always cut out the middle man (they are more transparent about it imo than I've seen people)
Honestly, they can add multiple more higher level servers as well and have them be slowdown at 0.005$ as well, I feel like this will be the key unlock of shellbox.dev and I am hopeful that they might work on it.
There are tools for almost everybody now in this space which didn't exist a month ago. Exe.dev for people (currently its burning VC money but once it gets stable) would be for people who have like 25-40 vms and have a constant amount of VM's and shellbox is gonna feel more like for those where you can one day probably have a simple abstraction and you can host software rather easily and use it and then dump the server to either completely remove it or just be when you might need it
Another issue is that if shellbox.dev is like the story of icarus, if they fly too high, I think they might win the war but they will lose the battle and if they fly too low or the competitors and it's major differences just become price, they both might compete to the bottom and the problem is that in times of emergency combined with the presence of Murphy's law, when disaster hits (they are basically arbitraging unused hardware space in the idea that long term enough users might come that things would be much rather full, they need a buffer because the servers might have unused compute for which they will have to pay
So if someone flies completely tries to lowball the price, what would end up happening is that in case of diasaster, they might actually fold/lose money and sustainability practises would be thrown out of the way (Just ask anyone in lowendtalk what happened to veloxmedia recently)
Honestly I am pretty frugal so I would still argue with your point though. I feel like they can still slash the prices in half of suspended costs because I think sprites.dev (their competitor) costs nothing/negligibly virtually when they are not running.
Shellbox does have this one benefit in comparison to sprites.dev in that they are more "No signup, no config, no complexity" and their small niche scale feature might still make sense.
Sprites.dev uses fly.io (a deeply awesome company) as well and it seems to have the most momentum right now as well
I think that these companies/projects will change their pricing and multitude of things will happen. I am really excited about this space.
Someone should probably create a comparison between exe.dev, sprites.dev and now shellbox.dev comparing each and every
Apart from the payment part, this could be used entirely from a machine without a GUI. You can do the same with others using Terraform or aws-cli but it requires setup first.
Maybe this and other future extended features could be configured via some host-accessible mounted conf.d? Otherwise if I forget to use that command on every login, I might just forget, logout, and go on thinking my server is still running.
Weird fun fact (as an Argentinian who went to school in England for a few years): in English-speaking countries, America is not a continent in the same way as in Spanish. In English they have two continents: South America and North America.
So the word "American" in English does not mean the same as "Americano" in Spanish.
There's really no natural word in English to refer to someone from "El continente Americano", because no such continent exists in English. That's why they use the word "American" to refer to someone from USA exclusively.
That sounded fascinating as a rather large difference in world view stemming only from using different languages.
It turns out that there are various models for the number of continents, and that is (phew) known in Spanish, too. See the Wikipedia page [1] (link to Spanish version) for instance. This is for European Spanish though, but I couldn't find a version of the page in es-AR.
I think "the Americas" means the continent(s), and America (to some extent) can mean either but it would feel more like something used as a gotcha at a pub quiz.
You're definitely right about there not being a word for someone from that continent though.
"American" to refer to USA exclusively does make sense either way because USA shares the continent with at least two other countries no matter how you slice it.
Frankly, the model with the single America continent doesn’t make any sense, because south and north Americas are so different in both geographical and cultural/historical sense.
This is all written in python and the AsyncSSH package. Firecracker for VMs with memory mapped files for ram. Paddle for billing. Caddy as a reverse proxy for certificates.
It works on top of very large bare metal instances.
I'm thinking maybe open sourcing but it will take some more work on the code to make it publishable w/o embarrassing myself :)
I am interested in which bare metal instances from which provider are you using if I may ask since I had a similar idea (as mentioned before) and I wanted to deploy it on hetzner but I was always worried that hetzner's policy might be too harsh for it even though they are one of the cheapest options out there
Which server provider did you end up using?
Thanks once again for your in depth response, these are the things I come to hackernews for! cheers and looking to ya response
I wrote a very slow, good citizen DHT scraper the other day to do some research and the second I spun it up on a Hetzner box I got an email from them that summarized as "For real, bro? Computer says no."
They actually look into the stuff you put on the computer? That's even worse than I thought. I did get a strong feeling of distrust towards their customers when I tried to sign up but because of that I never went ahead with the deal. Dodged a bullet, clearly.
Ps not doing anything illegal but I just don't like having copies of my ID everywhere. Too much data getting leaked these days. With Scaleway you just pay with your card and that's it.
Scaleway is on a bit more expensive side of things
Scaleway does have every primitive you might need but I don't know how good scaleway's support system is.
I have tried Upcloud for free and they are more expensive but they offer unlimited bandwidth (and the cap after 24 TB of free bandwidth/month is at 100 mbps)
Personally it depends on the solution, I have evaluated tons of servers and Scaleway's 0.10cents per month or 3 years per month for 1 gig 1 gb server is the cheapest for the smallest amount of resources needed
That being said, I personally prefer OVH as well. I have heard that their support can be shit but they are much more lenient than hetzner but their support is heard at times to be rough (someone recommended here to go talk on twitter to them if you want concerns solved as well and they have discord too so I guess)
But even after all of this, Hetzner's especially auction box are the king of undisputed prices and I am telling you after creating scripts scraping lowendtalks and many other things. There are some other options but none provide the safety of well founded company like hetzner (in terms of uptime etc.) and hetzner's pretty cool
They can be the very best (for some purposes they already are) but for the purposes of hosting other people's stuff or building own server on, they aren't so good because I think hetzner follows a strict policy.
Personally I am on netcup, I somehow looped it in such a way to get server of 8 gigs 4 core cpu 500 gb for 8$ / 3 months. I don't really use that server (too much or even barely, I have to give my brother access to that server so that he can run supabase on it because my servers literally empty aside from some basic stuff), I think at this point, these companies actively lose money for sure if I actually use what I own a decent amount of degree.
VPS/Cloud business is really fascinating ngl. I feel like though for some average use cases for which people used to buy 3$ servers for sprites.dev/exe.dev and now shellbox.dev are gonna be even cheaper (I thought about shellbox.dev and being honest, if I was capable to get hetzner auctions I would've done the same and there must be a middleman and they are more open about things as well and a lot of decisions are same of what I would've built)
Yup So I am probably throwing my full weight behind shellbox.dev as someone who wanted to build a cloud sometime ago. I got the idea of building cloud because I wanted to build something like this and none existed so I went into the rabbit hole but now its built. Now I can go the layer up that I wanted to in the first place now that this primitive has been built (an open source cloud solution where people can pay x$ to get access to a primitive like this)
I think I ended up building my own firecracker based (bottlefire) + ssh/golang based in the end for personal use and I will have it open source in the future (its really simple) and I hope that shellbox.dev can host it as well.
Honestly, we are seeing a lot of products in this space so lets hope that the best one wins! (I always appreciate competition)
> Scaleway is on a bit more expensive side of things
I don't agree. They're a lot cheaper than Amazon. And they don't charge all these hidden fees like ingress bandwidth for their compute services. Their glacier is also a ton cheaper than Amazon's (it does charge transfer fees but that's part of the model). I have 3 VPSes (stardust) for 9 euros a month, it's pretty ideal for me. They're in 3 countries too (you can only get 1 stardust per datacenter)
> Scaleway does have every primitive you might need but I don't know how good scaleway's support system is.
I have not had issues with them for years. But a few years back you could just contact them on slack and they would respond quickly. None of that hassle with creating tickets and CS agents deflecting because they're lazy or don't know. One time there was a recurring random issue and they just plainly told me "we know this particular service keeps having random issues, we'll discontinue it soon, much better to use this one instead".
This was really impressive. I work at work with the big tech names and they would never admit random instability issues like that. And they would never admit that one of their services is less than perfect. Probably because they're afraid someone will post it online and it'll cause a big fuss and requests for refunds. However I don't care about their reasoning. But these guys just plain told me what was up. It gives me a lot of trust, which I don't get from a party like Microsoft or AWS. They outsource their support to companies that outsource it again themselves and the resulting agents have no more knowledge than the kbase you can look up yourself.
So yeah I'm impressed. Not sure if their support is still like this because it's been rock-steady for years now.
> Personally it depends on the solution, I have evaluated tons of servers and Scaleway's 0.10cents per month or 3 years per month for 1 gig 1 gb server is the cheapest for the smallest amount of resources needed
No you can get IPv4 but you pay for it. This is the main cost of my stardust instances, about 3 euro per month including VAT. I don't use much IPv6 myself so I need it and it's not bad. That 3 bucks per VPS is nothing and it performs great. I really get a lot of use out of them.
> That being said, I personally prefer OVH as well. I have heard that their support can be shit but they are much more lenient than hetzner but their support is heard at times to be rough (someone recommended here to go talk on twitter to them if you want concerns solved as well and they have discord too so I guess)
OVH are really cowboys.. The fire in that datacenter exposed a lot of practices that really really should never have happened in a sane company.
And yes Hetzner is the king of power for the buck (if you need 24/7), that is true, especially baremetal if you don't mind your server being a few generations old.
PS: There's also the free server thing from Oracle of course. If you don't mind doing business with them. I do absolutely mind so that was not an option for me.
But I use my VPSes for constantly running stuff so one that goes on standby is simply not an option.
Hm I had thought the same! Interesting thanks for responding once again but what are your thoughts on the fact that someone can abuse the situation and your account might get banned and hetzner has a pretty strict policy in that
When I wished to create something as such, this was the most major thing I was worried about. I am curious what your thoughts are on it and how are you managing it (the fact that anyone might abuse in your service which could then impact you and hetzner relations and they might block/restrict you)
I have heard that hetzner requires you to respond in hours or similar. Like I am interested, did you talk to hetzner people (they are usually very kind and I love that about them) or not, because I remember asking some question to that in similar vein but I had gotten the answer that I am still responsible for what happens downstreams and that worried me
Do you do something similar to the modifications codesandbox has done to firecracker, regarding mmap ram? (They have multiple blogposts about it on their blog)
Sprites pricing is based on usage, not reserved capacity, so depending on what you're doing I think it can actually be cheaper than Shellbox. You'll have to stay below 1GB of memory and have the CPU be mostly idle, which I'm not sure common workloads will.
Nope, unless they changed this recently. It's an ssh-like way to connect and get a console/terminal, but it's not ssh, and there is no transfer capability
reply