Don't be disingenuous. It's a crowd-funding campaign. Not an actual product I can buy now. I'm so sick of clicking on headlines on HN which sound like they're describing some interesting product or service, only to be taken to the virtual begging bowl of a crowd-funding page.
If you don't find our product interesting, well, that's really our problem. I'm here to have valid arguments, potentially to convince you of the contrary, I'm not here to get attacked for no valid/obvious reason.
I'm still open to have a conversation like normal human beings usually have, so if you have questions related to our crowdfunding campaign, please ask.
Vikings exists to further libre software, libre-friendly hardware and improve the existing, locked-up "hosting industry" that often works against their own users and at the very least don't care for their privacy.
Where did I say I wasn't interested in the concept of your [non-existent] product? What I'm objecting to are the constant 'product announcement' headlines on HN which, rather than link to an actual, real product or service people can actually buy, link to a crowd-funding campaign.
Why can't you see that that is click-baity?
So if I post an article entitled "£10 Device Converts Household Waste to Diamonds" and link people to a crowd-funding campaign where I say. "If you give me eleventy-billion pounds I'll make a magic machine that converts household waste into diamonds —and I'll sell it for only£9,99" you don't see anything wrong with that?
Your own crowd-funding campaign blurb says:
"...Vikings plans to change that. We WILL build a state-of-the-art data center that WILL cover all your hosting needs; based on fully libre[1] software..." [my emphasis on "will"]
..and you're asking for people to give you between €520 000 and €800 000 to do so!
So you want the public to buy the data-centre and hardware for you, and you want to run the data-centre on software that other people have written and provided for free for you, while you presumably charge people for hosting their data there and 'laugh all the way to the bank' —and then you have the cheek to act all entitled and hurt when someone calls you out on the fact that your 'product announcement' headline is promoting something which doesn't actually exist.
"An eco-friendly data center that is based on 100% on libre software and runs libre-friendly hardware" - Does that include switches and routers? What about the UPS's?
We will build a state-of-the-art data center that will cover all your hosting needs; based on fully libre[1] software, libre boot firmware and performant libre-friendly hardware. We will always respect your freedoms and your privacy - we don't spy on you.
- based on 100% libre software and libre-friendly hardware with Libreboot, including our front- and back-ends
So if you're limited to Libreboot, the last set of Intel processors you can support came out in, what? 2009? 2010? Cross referencing the list of "supported" systems[0] with some general googling for their release dates[1], 2010 seems to be it, unless their data center is gonna run on a rack of Macbook 2's and Chromebook C201's.
Either way, it's hardly "state of the art" hardware by any stretch. Maybe they'll go with some custom RISC processors?
We've chosen the D16 because it is available now - with free/libre boot firmware, without binary blobs. The systems are very performant, despite their "age" (the CPUs we use were released 4 years ago (6284 SE), a system can have up to 32 CPU cores with 2.7GHz).
So from the performance perspective, systems based on the D16 are still able to compete with other, newer systems. From the "respecting your freedoms"-perspective, the D16 is unsurpassed AFAIK.
These systems are in fact a practical solution for free/libre software idealists and people who don't want to live in an Orwellian society.
That being said, relying only on this hardware would be foolish indeed. Not only because it will become too expensive to run at some point, like any hardware that gets older, but also because x86 seems to be a dead-end because it's becoming more and more locked-up.
We're planning to use openPOWER systems in the near future and move away from x86, to provide the best services for people who care about their freedom.
More will be revealed in campaign updates, we've just started the campaign and there will be campaign updates over the next 39 days that will interest you.
> "Since the Summer of Snowden what we have feared had become true: almost all modern hardware platforms used for hosting contain potential backdoors or exploitable bugs.
Vikings plans to change that."
You won't change that with Libre Software : as long as your mobile / computer / server is connected to the internet it is hackable no matter what. Libre Software is not exempt of exploitable bugs, there's a lot of non libre parts in hardware etc and someone from the team could be subverted to insert a backdoor.
Basically they're bringing wooden shields to the fight when the NSA is shooting enriched uranium shells from Big Bertha !
Their goal is noble but they shouldn't let people believe that their data will be totally secure on these servers.
I think this axiom is pretty common in the Free Software field: Free Software tackles exactly the issues that are important
So when something is deemed an important issue, by the axiom it is solved through free software.
In reality most issues are orthogonal to licensing (using a cloud or not, encryption, ...) and in some classical free software performs arguably even worse (usability)
It raises a lot of questions, like... a ton.
1. Really libre until what? Do you have ANY Libre DataCenterInfrastructureManagement/BuildingManagementSystem that can really be used in production? What about the Chillers/UPS/CRAC/GENsets/... What about the design of the DataCenter ? The access system, badging system, ...?
2. What about networking devices?
3. They provide VPNs/... but what about ISPs/...?
4. If it's they're 1st DC, what about the quality of the designs? The redundancy?
...
> What could go wrong?
Everything. From the beginning with how you choose your chillers (a bad choice could easily double your PowerUsageEffectiveness), to the future with how you operate it.
Nonetheless, I find the initiative awesome and I wish you a huge luck and I really hope they'll succeed! The intention is clearly laudable.
P.S.: I work at Etix Everywhere (www.etixeverywhere.com) a company designing, building and operating modular DataCenters around the globe. We're recruiting a lot so don't hesitate to take a look, and help us to build data centers really everywhere!
24 comments
[ 84.6 ms ] story [ 1690 ms ] threadI'm still open to have a conversation like normal human beings usually have, so if you have questions related to our crowdfunding campaign, please ask.
Vikings exists to further libre software, libre-friendly hardware and improve the existing, locked-up "hosting industry" that often works against their own users and at the very least don't care for their privacy.
Where did I say I wasn't interested in the concept of your [non-existent] product? What I'm objecting to are the constant 'product announcement' headlines on HN which, rather than link to an actual, real product or service people can actually buy, link to a crowd-funding campaign.
Why can't you see that that is click-baity?
So if I post an article entitled "£10 Device Converts Household Waste to Diamonds" and link people to a crowd-funding campaign where I say. "If you give me eleventy-billion pounds I'll make a magic machine that converts household waste into diamonds —and I'll sell it for only£9,99" you don't see anything wrong with that?
Your own crowd-funding campaign blurb says:
"...Vikings plans to change that. We WILL build a state-of-the-art data center that WILL cover all your hosting needs; based on fully libre[1] software..." [my emphasis on "will"]
..and you're asking for people to give you between €520 000 and €800 000 to do so!
So you want the public to buy the data-centre and hardware for you, and you want to run the data-centre on software that other people have written and provided for free for you, while you presumably charge people for hosting their data there and 'laugh all the way to the bank' —and then you have the cheek to act all entitled and hurt when someone calls you out on the fact that your 'product announcement' headline is promoting something which doesn't actually exist.
Jeebus!—talk about 'Generation Snowflake'!
Yeah. I'll hang on to my money thanks.
- based on 100% libre software and libre-friendly hardware with Libreboot, including our front- and back-ends
So if you're limited to Libreboot, the last set of Intel processors you can support came out in, what? 2009? 2010? Cross referencing the list of "supported" systems[0] with some general googling for their release dates[1], 2010 seems to be it, unless their data center is gonna run on a rack of Macbook 2's and Chromebook C201's.
Either way, it's hardly "state of the art" hardware by any stretch. Maybe they'll go with some custom RISC processors?
[0]https://en.wikipedia.org/wiki/Libreboot#Supported_systems
[1]https://www.techpowerup.com/119540/asus-releases-kgpe-d16-so...
They're planning to run a Asus KGPE-D16 AMD motherboard. This is a 2009 motherboard.
From Libreboots website:
> Libreboot has support for fam15h AMD hardware (~2012 gen) and some older Intel platforms (~2006-2009 gen).
> It is extremely unlikely that any post-2008 Intel hardware will ever be supported in libreboot
> It is extremely unlikely that any post-2013 AMD hardware will ever be supported in libreboot
https://libreboot.org/faq/
This project has no legs.
So from the performance perspective, systems based on the D16 are still able to compete with other, newer systems. From the "respecting your freedoms"-perspective, the D16 is unsurpassed AFAIK.
These systems are in fact a practical solution for free/libre software idealists and people who don't want to live in an Orwellian society.
That being said, relying only on this hardware would be foolish indeed. Not only because it will become too expensive to run at some point, like any hardware that gets older, but also because x86 seems to be a dead-end because it's becoming more and more locked-up.
We're planning to use openPOWER systems in the near future and move away from x86, to provide the best services for people who care about their freedom.
More will be revealed in campaign updates, we've just started the campaign and there will be campaign updates over the next 39 days that will interest you.
PS: This project has wings instead.
You won't change that with Libre Software : as long as your mobile / computer / server is connected to the internet it is hackable no matter what. Libre Software is not exempt of exploitable bugs, there's a lot of non libre parts in hardware etc and someone from the team could be subverted to insert a backdoor.
Basically they're bringing wooden shields to the fight when the NSA is shooting enriched uranium shells from Big Bertha !
Their goal is noble but they shouldn't let people believe that their data will be totally secure on these servers.
So when something is deemed an important issue, by the axiom it is solved through free software.
In reality most issues are orthogonal to licensing (using a cloud or not, encryption, ...) and in some classical free software performs arguably even worse (usability)
> What could go wrong? Everything. From the beginning with how you choose your chillers (a bad choice could easily double your PowerUsageEffectiveness), to the future with how you operate it.
Nonetheless, I find the initiative awesome and I wish you a huge luck and I really hope they'll succeed! The intention is clearly laudable.
P.S.: I work at Etix Everywhere (www.etixeverywhere.com) a company designing, building and operating modular DataCenters around the globe. We're recruiting a lot so don't hesitate to take a look, and help us to build data centers really everywhere!
An open source DCIM that's used in production: https://github.com/digitalocean/netbox
Intel having a backdoor to cloud servers in the current climate seems like a situation worth working to amerliorate. Best of luck!