@cservers said: The main reasoning behind the offer for 99 years is actually twofold: it's both to support our current expansion, and to have a deal we can honestly commit and fulfill throughout the times (making it a "lifetime"), regardless of how the wind changes and the world evolves, we are saying clearly: we will fulfill it.
Why do you specify the specific type of chip, one that is already several years past it's EOL.
I doubt these chips will be economically viable to run compared to "current gen" chips in another 10-20 years time, and I don't you'll even have a large enough of chips to even keep running that processor beyond 30-40 years even if the running cost wasn't the issue.
What happens when it's simply impractical to even run these CPUs any more?
Obviously, you know and I know that you don't really want to target this specific chip after all. It's going to be easier and cheaper to move everyone to a new node. Are you planning on upgrading the specs of this new node as time goes on? Otherwise, in 10 years time this deal will be probably useless for almost any tasks of that era, and in 20 years time probably there will be literally no point in keeping it.
Would your plan be to always emulate at similar spec level to the original, or maintain a moving target like every 10 years we'll update to a 10 year-old tech?
Given that the projections between investing the money yourself and buying a $10 per year server (adjusted to today's value in real terms) from the savings would leave cash behind over 30 years, it doesn't seem to make any sense for anyone to consider a server lease this long if there isn't a plan to increase the specs every N years.
@cservers said: The main reasoning behind the offer for 99 years is actually twofold: it's both to support our current expansion, and to have a deal we can honestly commit and fulfill throughout the times (making it a "lifetime"), regardless of how the wind changes and the world evolves, we are saying clearly: we will fulfill it.
This is an honorable goal, but considering the lifespan of the average human, it isn't something you can guarantee.
But I do like to think that 100 years from now people will be treating vintage servers the same way we treat original Model Ts in our time.
This is the sysadmin speaking, my name is Tiago. Sorry in advance for the large post, some relevant details there.
This service (as described, but omitted by OP in the first post) are part of a pre-sale, with deployment announced since early December as scheduled until the 14th of January 2025.
I must address several aspects here:
This is still pre-deployment (actually our first fully owned server, non-upstreamed to Hetzner and 100% pure metal, it's an HPE server);
The main reasoning behind the offer for 99 years is actually twofold: it's both to support our current expansion, and to have a deal we can honestly commit and fulfill throughout the times (making it a "lifetime"), regardless of how the wind changes and the world evolves, we are saying clearly: we will fulfill it.
Obviously a "lifetime" VPS is no joke of a deal, we take these very seriously, and therefore it could never be priced low, but the money will go both to current and future use of the VPS and to help fund new servers, new services and everything else associated, as current purchases already are and will still be for quite some time, which is normal and an usual business practice basically everywhere.
This is not that new of a company anymore, in fact we are already year and a half in since foundation, although only started trading some months later. We are not here to go anywhere, we are here to stay, for as many years as they come, and so am I in this business. And sustainability is key for C-Servers; in fact, this company is self-sustainable since the 2nd month after it started trading, and we have no debt. It's all as organic as possible. This is also why we are cautious when putting prices, deployment and everything involved. It must be sustainable and a question of responsibility to our customers.
Let me also detail one final thing regarding the unusual location and the deployment: this will be changed, in fact it will actually be improved. To Central Europe, no less. I am handling this and in talks with multiple datacenters as we speak (so, no basements) for a long-term relationship, and I'll personally fly myself, and the server, to the destination datacenter; if I can I won't send no post or mail of a thing this heavy and this important. Flights exist for a reason, and I don't care paying more.
To break it in short: we had a contact in Portugal that was offering us a low-enough colocation price that we actually were looking forward to, hence our location being communicated and decided as Portugal since December, and the deployment interval was defined after these talks occurred; but after Christmas and the New Year things have made quite a U turn (pun intended), as final pricing doubled to higher-than-mainland Europe levels, and I found it unreasonable to deploy in Portugal for something that is (as we all know) objectively second-best in terms of connectivity to mainland Europe. And if I can have the real deal at an even better price, why would I settle for less, right?
As we speak, I am negotiating with datacenters in France, Germany and the UK, and working very hard to fulfill the deadline (14 January 2025), which is on a Tuesday.
I can also advance the following information: if I find it not possible to reasonably deploy the servers in-person until that day, everyone will be formally notified via e-mail with the best advance I can possibly give, and the deployment day will have an exceptional adjustment of a couple days at most (to e.g. January 17th or January 21st at most). If I for some reason fail that second deployment date, everyone on the pre-sale gets a refund. Direct.
As you see, no, no deadpooling, no nothing. Only someone that works hard for their customers.
Any further questions, I'm available.
Tiago
My primary reason for ordering the VPS was its location in Portugal, so this news is quite disappointing. There are many low-cost VPS options available, but not specifically from Portugal.
@cservers said: The main reasoning behind the offer for 99 years is actually twofold: it's both to support our current expansion, and to have a deal we can honestly commit and fulfill throughout the times (making it a "lifetime"), regardless of how the wind changes and the world evolves, we are saying clearly: we will fulfill it.
Why do you specify the specific type of chip, one that is already several years past it's EOL.
I doubt these chips will be economically viable to run compared to "current gen" chips in another 10-20 years time, and I don't you'll even have a large enough of chips to even keep running that processor beyond 30-40 years even if the running cost wasn't the issue.
What happens when it's simply impractical to even run these CPUs any more?
Obviously, you know and I know that you don't really want to target this specific chip after all. It's going to be easier and cheaper to move everyone to a new node. Are you planning on upgrading the specs of this new node as time goes on? Otherwise, in 10 years time this deal will be probably useless for almost any tasks of that era, and in 20 years time probably there will be literally no point in keeping it.
Would your plan be to always emulate at similar spec level to the original, or maintain a moving target like every 10 years we'll update to a 10 year-old tech?
Given that the projections between investing the money yourself and buying a $10 per year server (adjusted to today's value in real terms) from the savings would leave cash behind over 30 years, it doesn't seem to make any sense for anyone to consider a server lease this long if there isn't a plan to increase the specs every N years.
When the time comes for that chip to be sunset, it will happen as it has happenned already for the previous NAT customers on the Zeta2 server (MultiNAT servers that we offered between August and November): the CPU is directly upgraded between changes. That's what happenned to them - we moved the NAT customers from a Core i7-7700T to a Core i7-8700.
Another example: we had a customer ordering multiple servers (some 8 servers) on the Outlet offer, which predicted a Core i7-8700. In order to evenly split load accross nodes (and to go around a temporary IPv6 shortage), we went ahead and deployed around half of the VPS purchases he had done between both servers, let around 3 or 4 of them marked as "In Progress", and deduplicated this exact plan on the other server we have (Zeta 3), which has a much more powerful CPU (Ryzen 9 3900), at the exact same cost. (This was around Christmas, so you can imagine the extra work.)
We added very recently a full-backup to all customers (C-Servers GoBack) as well, and that was entirely for free as well, with optional paid plans (GoBack Plus) for the most self-conscious customers.
When we introduce DDR5, which we plan to do quite soon for example, if we close the Hetzner-upstreamed servers, we'll migrate the customers to our own servers and they win the better CPU and the better DDR5 compared to their original plan, for example. And when the Zeta1 server no longer delivers and the CPU starts not to be reasonable enough for current usage (possibly in 2-4 years?), we'll migrate these customers as well to something better and with a better CPU. And so on, and so on.
It's really a question of keeping paying customers and keeping them happy as well, as technology evolves. We do prefer to change customers between servers and update the dedicated server's main CPUs/RAM whenever convenient or necessary, at the same cost, than disrupting everything else. Technology rates have been decreasing the cost and increasing the performance, it's not hard for us to do it. It also helps to establish an userbase and keep it as well, which as you may understand, we want.
It doesn't mean we always keep the CPUs 10 years delayed, but it always means, whenever we change, it's an improvement.
It would be too much work to attempt to emulate the CPUs, we would need a platform within the platform, the added performance loss and all the hassle just doesn't justify it. It's simpler to adjust the VPS to the server it's into and let the share occur from the pool. For us, the reasoning behind the purchase must stay (e.g. if it's something with little RAM or lots of RAM, if it's something with less vCores or more vCores), and that's what we attempt to keep, adjusted to each reality at any given moment.
Any customer can request a written declaration on what we just said and we'll gladly give it. After all, a purchase is always a binding contract for both parties.
@bitshub said:
My primary reason for ordering the VPS was its location in Portugal, so this news is quite disappointing. There are many low-cost VPS options available, but not specifically from Portugal.
We understand, no problem at all. It's one of the exception use-cases. Send us a ticket and we'll refund, you naturally have no responsibility on any location change, it's our responsibility.
It is expected for some customers to request one, and that's okay.
@Neoon said:
Put your NAT into locations that nobody else has and it will sell like hotcakes.
But no, again, France, Netherlands and fucking Germany, meh.
@Neoon said:
Put your NAT into locations that nobody else has and it will sell like hotcakes.
But no, again, France, Netherlands and fucking Germany, meh.
I'm quite puzzled about the unknown image files of the Shapes that appear in the top left and bottom right corners on the homepage of https://c-servers.co.uk/.
@admax said:
I'm quite puzzled about the unknown image files of the Shapes that appear in the top left and bottom right corners on the homepage of https://c-servers.co.uk/.
Offering a 99-year VPS is an absurd notion and if these guys sincerely think they can fulfill the obligation, they're deluded. Even ignoring specs and the questionable utility of a VPS in 99 years, there's also an intense amount of regulatory uncertainty on such a long timescale.
I think the most any company can reasonably offer is a 1 year commitment. You need to have a sensible time limit on any service you're providing since you can't be certain of the costs of providing the service in the distant future. The only way to make it sort of work is to have future customers subsidize the 99-year customers in a manner similar to a ponzi scheme.
@whoami said:
Offering a 99-year VPS is an absurd notion and if these guys sincerely think they can fulfill the obligation, they're deluded. Even ignoring specs and the questionable utility of a VPS in 99 years, there's also an intense amount of regulatory uncertainty on such a long timescale.
I think the most any company can reasonably offer is a 1 year commitment. You need to have a sensible time limit on any service you're providing since you can't be certain of the costs of providing the service in the distant future. The only way to make it sort of work is to have future customers subsidize the 99-year customers in a manner similar to a ponzi scheme.
Actually, there doesn't need to be a ponzi scheme. If the company invests the money in an S&P ETF, they would earn enough to cover the annual $10 fee for 99 years. They could likely afford to keep providing the service ad infinitum.
The real problem (and hence my question) is that the service is for a specific spec. They said as much that they'll completely ignore the promised spec and just upgrade it as and when to ("just trust me bro") something better. However, if you've entered into the agreement as described on their website they're under no obligation to do so at all.
The much smarter choice for any rational person is to just invest the money themselves and buy it annually. That way they have complete flexibility to upgrade to something else or cancel as and when the service no longer meets your needs.
Comments
Why do you specify the specific type of chip, one that is already several years past it's EOL.
I doubt these chips will be economically viable to run compared to "current gen" chips in another 10-20 years time, and I don't you'll even have a large enough of chips to even keep running that processor beyond 30-40 years even if the running cost wasn't the issue.
What happens when it's simply impractical to even run these CPUs any more?
Obviously, you know and I know that you don't really want to target this specific chip after all. It's going to be easier and cheaper to move everyone to a new node. Are you planning on upgrading the specs of this new node as time goes on? Otherwise, in 10 years time this deal will be probably useless for almost any tasks of that era, and in 20 years time probably there will be literally no point in keeping it.
Would your plan be to always emulate at similar spec level to the original, or maintain a moving target like every 10 years we'll update to a 10 year-old tech?
Given that the projections between investing the money yourself and buying a $10 per year server (adjusted to today's value in real terms) from the savings would leave cash behind over 30 years, it doesn't seem to make any sense for anyone to consider a server lease this long if there isn't a plan to increase the specs every N years.
This is an honorable goal, but considering the lifespan of the average human, it isn't something you can guarantee.
But I do like to think that 100 years from now people will be treating vintage servers the same way we treat original Model Ts in our time.
no sad no dedicated ipv4
I bench YABS 24/7/365 unless it's a leap year.
My primary reason for ordering the VPS was its location in Portugal, so this news is quite disappointing. There are many low-cost VPS options available, but not specifically from Portugal.
When the time comes for that chip to be sunset, it will happen as it has happenned already for the previous NAT customers on the Zeta2 server (MultiNAT servers that we offered between August and November): the CPU is directly upgraded between changes. That's what happenned to them - we moved the NAT customers from a Core i7-7700T to a Core i7-8700.
Another example: we had a customer ordering multiple servers (some 8 servers) on the Outlet offer, which predicted a Core i7-8700. In order to evenly split load accross nodes (and to go around a temporary IPv6 shortage), we went ahead and deployed around half of the VPS purchases he had done between both servers, let around 3 or 4 of them marked as "In Progress", and deduplicated this exact plan on the other server we have (Zeta 3), which has a much more powerful CPU (Ryzen 9 3900), at the exact same cost. (This was around Christmas, so you can imagine the extra work.)
We added very recently a full-backup to all customers (C-Servers GoBack) as well, and that was entirely for free as well, with optional paid plans (GoBack Plus) for the most self-conscious customers.
When we introduce DDR5, which we plan to do quite soon for example, if we close the Hetzner-upstreamed servers, we'll migrate the customers to our own servers and they win the better CPU and the better DDR5 compared to their original plan, for example. And when the Zeta1 server no longer delivers and the CPU starts not to be reasonable enough for current usage (possibly in 2-4 years?), we'll migrate these customers as well to something better and with a better CPU. And so on, and so on.
It's really a question of keeping paying customers and keeping them happy as well, as technology evolves. We do prefer to change customers between servers and update the dedicated server's main CPUs/RAM whenever convenient or necessary, at the same cost, than disrupting everything else. Technology rates have been decreasing the cost and increasing the performance, it's not hard for us to do it. It also helps to establish an userbase and keep it as well, which as you may understand, we want.
It doesn't mean we always keep the CPUs 10 years delayed, but it always means, whenever we change, it's an improvement.
It would be too much work to attempt to emulate the CPUs, we would need a platform within the platform, the added performance loss and all the hassle just doesn't justify it. It's simpler to adjust the VPS to the server it's into and let the share occur from the pool. For us, the reasoning behind the purchase must stay (e.g. if it's something with little RAM or lots of RAM, if it's something with less vCores or more vCores), and that's what we attempt to keep, adjusted to each reality at any given moment.
Any customer can request a written declaration on what we just said and we'll gladly give it. After all, a purchase is always a binding contract for both parties.
We understand, no problem at all. It's one of the exception use-cases. Send us a ticket and we'll refund, you naturally have no responsibility on any location change, it's our responsibility.
It is expected for some customers to request one, and that's okay.
In 100 years 60 GB of NVMe space might turn out to become something like what a floppy disk is nowadays.
Stop the planet! I wish to get off!
Put your NAT into locations that nobody else has and it will sell like hotcakes.
But no, again, France, Netherlands and fucking Germany, meh.
Free NAT KVM | Free NAT LXC | Bobr
ITS WEDNESDAY MY DUDES
100%
Noted.
It turns out the Portugal VPS location is already gone! Then there's no need to buy it.
Feeling a bit tired, let's call it a day.
Thirty four cents per month averaged over ninety nine years...
99years is a very adventurous assumption and optimism
I bench YABS 24/7/365 unless it's a leap year.
If it’s still around in three or four years, that would already be the greatest luck.
Feeling a bit tired, let's call it a day.
I'm quite puzzled about the unknown image files of the
Shapesthat appear in the top left and bottom right corners on the homepage of https://c-servers.co.uk/.Feeling a bit tired, let's call it a day.
got a email, it will be delayed
Fill with Tart sister
Offering a 99-year VPS is an absurd notion and if these guys sincerely think they can fulfill the obligation, they're deluded. Even ignoring specs and the questionable utility of a VPS in 99 years, there's also an intense amount of regulatory uncertainty on such a long timescale.
I think the most any company can reasonably offer is a 1 year commitment. You need to have a sensible time limit on any service you're providing since you can't be certain of the costs of providing the service in the distant future. The only way to make it sort of work is to have future customers subsidize the 99-year customers in a manner similar to a ponzi scheme.
Actually, there doesn't need to be a ponzi scheme. If the company invests the money in an S&P ETF, they would earn enough to cover the annual $10 fee for 99 years. They could likely afford to keep providing the service ad infinitum.
The real problem (and hence my question) is that the service is for a specific spec. They said as much that they'll completely ignore the promised spec and just upgrade it as and when to ("just trust me bro") something better. However, if you've entered into the agreement as described on their website they're under no obligation to do so at all.
The much smarter choice for any rational person is to just invest the money themselves and buy it annually. That way they have complete flexibility to upgrade to something else or cancel as and when the service no longer meets your needs.
No assumptions here. It is completely unsustainable.
do you plan to live that long ? or u want that ur business still runs after you pass away ?
Can you support pre-authorization of $399, but debit after 99 years?
Have the honor of being the crybaby who pays $20 for a 128MB VPS at VirMach in 2023.
We know when we'll die.
We plan to buy a 999-year server during our last year of lifetime, so that our legacy lives on for a long time.
No hostname left!
2660v4 is legacy but not that legacy you imagine it to be.
I bench YABS 24/7/365 unless it's a leap year.
Interesting points were done here. We are evaluating this question.