Topic: [ANN]Arweave (Techstars 2018): A new internet in a very scalable blockchain - page 10. (Read 11984 times)

This project is very interesting. Something new. Everybody is tired of all this so called "new altcoins" and new "assets" created by scam artists. I wish you good luck ! 

Hi FredDag,

Thanks for your questions and interest in Archain!

First, the cost of storing a document will be proportionate to the size of that document -- for both 'verified' web pages and 'unverified' files. We do not envisage a daily block size limit.

Support for webpages that load their main content through Javascript is still under review at the moment. These pages (after their contents is loaded) could always be archived in their 'unverified' form, but may not be verifiable (with all their content loaded) when the network first launches. This is something we will look at in the first few major updates to the Archain web-scaper post-launch. One of the methods we are exploring is similar to the way Google partially evaluates some Javascript in order to make dynamic sites indexable.

Videos are a difficult area for Archain, owing to their size. While it will always be possible to upload unverified videos to the network, this will likely prove prohibitively expensive for most situations (given the permanent nature of Archain storage). Instead, however, we aim to support replacing videos in webpages with hashes of videos, in a similar way to the method described for images and other multimedia objects in the whitepaper. This will allow users to 're-unite' videos that have been stored off the chain, with the web pages that have been stored on it. In this way, as long as the video is stored somewhere that it is accessible later, it can be proved to have belonged to the archived web page, at the time of archiving, after the fact (for only the cost of storing the hash itself).

We have not considered a 'Content Approver' role, as this could potentially lead to systemic censorship. Data getting onto the weave, however, is dependent on being accepted by the miners. The system has been designed in such a way that if the community decided to, they could build their own 'content approval' nodes, which miners could choose to consult before accepting transactions. These could simply be typical Archain Developer Toolkit apps (discussed on the website and in the whitepaper).

Finally, we have an experimental anti-virus system (built atop SamAV, utilising ClamAV malware definitions) that can be optionally used to scan incoming transactions for dubious contents. Please note that this feature is 'experimental' at this stage and may not make it into the first release (hence its absence from the whitepaper).

I hope this is helpful. Please don't hesitate to let us know if you have any further comments or questions.

Thanks,
Archain Team

Great to see someone finally tackling this concept; It's something I have been thinking about for a while.

Having read through the whitepaper, I have a few questions...

• How will the submitted file size affect the cost to submit the URL ?
• Do you have any plans to limit the daily  block data size ?
• What controls do you plan to prevent viruses & malware being added to the library ?
• How do you plan to handle Javascript webpages which load content after the HTML has loaded ?
• How would submitting videos be handled ? YouTube webpages (for example) might be updating frequently due to likes, views etc. So a plain video submit may be preferable.
• Have you considered including an incentivised Content Approver role ? This could be used to check unverified submissions for bad content (porn,hate etc) before being flagged for inclusion by miners.

Hi mgao6767,

Excellent! Python support is high on our priorities list.

Thanks,
Archain Team

I use Python a lot, so Python!

Hi mgao6767!

Thanks for your support of Archain.

The core of the Archain system is written in Erlang (for the scalability reasons you mention), but we plan to build an 'Archain App Bridge' in the future, allowing the use of other languages.

At the moment we are considering adding support for Java and Python, but we are open to suggestions. Which languages were you particuarly interested in writing Archain apps in?

Thanks,
Sam
Archain CEO

I really love the idea of Archain. But why do you guys choose Erlang as the programming language? For scalability only? Honestly it isn't really friendly.... Beyond this, I'm gonna support Archain anyway.

The Archain pre-sale now supports direct Bitcoin payments, for those that were enquiring.

Thanks,
Team Archain

Thanks for your interest in Archain bounties!

Signature bounties will be available soon. We are just finalising the structure and delivery of our bounty programme and will PM you personally when it has launched.

Sad. No signature bounty campaign yet. It really does help awareness among crypto enthusiasts. I think you should start one right now.

This project have a good specification, I see the development of this thread a few days ahead, of course with a lot of information, I'm interested, hopefully run as planned, good luck

when it comes, please PM me sir. thank you

We'd be very grateful if you did this for us when the time comes.

Good question. The simple answer is 'enough'. The longer answer is that the incentives to store blocks increase dramatically as the number of copies of the blocks decreases. At the start, we expect that most miners will hold most blocks, then over time the proportion of the weave held by each miner will slowly decrease.

While its hard to predict what the behaviour of the network will be in the long term, the algorithm has been designed such that the miners are always selfishly encouraged to provide a good coverage of all the blocks.



Thank you, this is an excellent suggestion. We will have to talk to our legal team about this first, but we will look into doing it.

Thanks,
Sam + Will
CEO and CTO at Archain

Thank you so much for replying. I was guessing correct that the blocks will be saved over the nodes randomly with some redundancy. How many copies for redundancy do you expect?

reserved translation to Vietnamese
my previous work: https://bitcoingarden.org/forum/index.php?topic=17332.0
https://bitcoingarden.org/forum/index.php?topic=17454.0
https://bitcoingarden.org/forum/index.php?topic=17235.0
https://bitcoingarden.org/forum/index.php?topic=17309.0

Can I suggest creating a ERC20 token that can be traded in the interim, only for those that would like to do so, and that will be 1:1 to your token upon genesis launch? 

Hi,

Thank you for the feedback. We do try and explain how our blockweave implements true scaling in the whitepaper, but if it wasn't clear, then perhaps we can do better. To quickly summarise: the blockweave incentives even distribution of the data because a randomly selected old block is required in the mining of each new block. In this way, mining can only efficiently be achieved by storing


a) the whole blockweave (this is what we expect people to do early on)

or, if storing the whole blockweave is not feasible

b) storing the bits of the blockweave that are least stored by everybody else.

In this way, as the blockweave becomes too large to store in one place, the best strategy for each miner individually results in an even spread of the stored data across the group.



Hi,

You will recieve your public/private key immediately, and will be able to use your wallet on the testnet as soon as it is live.

Obviously, tokens wont be available for full use until the genesis block is released.


We will open source everything related to archain.

Very interesting about Blockweave. Is it similar to XRB, IOTA or Byteball's tangle data structure? The whitepaper doesn't explain much on how it achieves consensus.

After the ICO on Sept 18 when will holder receive their tokens?  I hope they will not be locked until genesis Q1 or Q2 2018?