for - quote - silos - internet limitations - location addressed server architecture limitations - silos - cannot link data from each silo - Juan Benet - IPFS

for - quote / insight - We shouldn't be doing the web over this simple file distribution system that works over TCP - Juan Benet - IPFS

for - IPFS - etymology - Inter Planetary - designing to avoid large network delay differences over long distances - Juan Benet

for - IPFS Search for - Indyweb - Indranet - Conceptapedia

https://subconscious.substack.com/p/noosphere-a-protocol-for-thought

imho, the main problem with IPFS is that it does DHT over TCP, which is crazy-inefficient, compared to bittorrent, which does DHT over UDP, which "just works"

one reason for DHT over TCP is to make this work in a web browser, which does not support UDP. so instead of teaching web browsers to talk UDP, IPFS took the simple route of "lets take bittorrent and run DHT over TCP"

IPFS is obsolete, many goals of IPFS can be achieved with bittorrent v2

generally, treating web browsers as the main target platform (and thus inheriting their limitations) is a bad idea, equally stupid as the "lets run everything in javascript/WASM" idea

Content addressing is when a piece of content has its own permanent address, a URI. Many copies may exist of the content, hosted by many in the network, all copies have the same address. Whoever is best situated to serve you a copy, does so. It makes the servers interchangeable. My blogposts have a canonical fixed address, but it's tied to a specific domain and only found on 1 server (except when using a CDN).

IPFS starts from content addressing.

Content addressing, assuming the intention 'protocol for thought' here, does match with atomic notes type of pkm systems. All my notes have unique names that could as human readable names map to CIDs. CIDs do change when the content changes, so there's a mismatch with the concept of 'permanent notes' that are permanent in name/location yet have slowly evolving content.

Noosphere was presented at the Render conf on tools for thought. There's overlap with Boris' work on IPFS, judging by his tweets he's involved in this effort too.

This Noosphere Explainer explains the tech used, not the 'massive-multiplayer knowledge graph' it is posited to be, how that would come about with this tech, or what that is meant to be for.

The public Indyweb is eminently designed as a public space for holding deep, continuous, asynchronous conversations with provenance. That is, if the partcipant consents to public conversation, ideas can be publicly tracked. Whoever reads your public ideas can be traced.and this paper trail is immutably stored, allowing anyone to see the evolution of ideas in real time.

In theory, this does away with the need for patents and copyrights, as all ideas are traceable to the contributors and each contribution is also known. This allows for the system to embed crowdsourced microfunding, supporting the best (upvoted) ideas to surface.

Participants in the public Indyweb ecosystem are called Indyviduals and each has their own private data hub called an Indyhub. Since Indyweb is interpersonal computing, each person is the center of their indyweb universe. Through the discoverability built into the Indyweb, anything of immediate salience is surfaced to your private hub. No applications can use your data unless you give exact permission on which data to use and how it shall be used. Each user sets the condition for their data usage. Instead of a user's data stored in silos of servers all over the web as is current practice, any data you generate, in conversation, media or data files is immediately accessible on your own Indyhub.

Indyweb supports symmathesy, the exchange of ideas based on an appropriate epistemological model that reflects how human INTERbeings learn as a dynamic interplay between individual and collective learning. Furthermore, all data that participants choose to share is immutably stored on content addressable web3 storage forever. It is not concentrated on any server but the data is stored on the entire IPFS network:

"IPFS works through content adddressibility. It is a peer-to-peer (p2p) storage network. Content is accessible through peers located anywhere in the world, that might relay information, store it, or do both. IPFS knows how to find what you ask for using its content address rather than its location.

There are three fundamental principles to understanding IPFS:

Unique identification via content addressing Content linking via directed acyclic graphs (DAGs) Content discovery via distributed hash tables (DHTs)" (Source: https://docs.ipfs.io/concepts/how-ipfs-works/)

The privacy, scalability, discoverability, public immutability and provenance of the public Indyweb makes it ideal for supporting hyperconversations that emerge tomorrows collectively emergent solutions. It is based on the principles of thought augmentation developed by computer industry pioneers such as Doug Englebart and Ted Nelson who many decades earlier in their prescience foresaw the need for computing tools to augment thought and provide the ability to form Network Improvement Communities (NIC) to solve a new generation of complex human challenges.

Making a Memento

To create an archived version of the page that could be played back properly, I used the Internet Archive’s “Save” feature by going to this URL in my web browser:

http://web.archive.org/save/http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt

…which created this snapshot:

http://web.archive.org/web/20150709104019/http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt

From here, we can use wget to look at what gets played back:

$ wget --server-response http://web.archive.org/web/20150709104019/http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt

…giving:

  HTTP/1.0 200 OK
  Server: Tengine/2.1.0
  Date: Thu, 09 Jul 2015 10:41:38 GMT
  Content-Type: text/plain;charset=utf-8
  Content-Length: 13
  Set-Cookie: wayback_server=19; Domain=archive.org; Path=/; Expires=Sat, 08-Aug-15 10:41:38 GMT;
  Memento-Datetime: Thu, 09 Jul 2015 10:40:19 GMT
  Link: <http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt>; rel="original", <http://web.archive.org/web/timemap/link/http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt>; rel="timemap"; type="application/link-format", <http://web.archive.org/web/http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt>; rel="timegate", <http://web.archive.org/web/20150709104019/http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt>; rel="first last memento"; datetime="Thu, 09 Jul 2015 10:40:19 GMT"
  X-Archive-Orig-x-cache-hits: 0
  X-Archive-Orig-x-served-by: cache-sjc3122-SJC
  X-Archive-Orig-cache-control: max-age=600
  X-Archive-Orig-content-type: text/plain; charset=utf-8
  X-Archive-Orig-server: GitHub.com
  X-Archive-Orig-age: 0
  X-Archive-Orig-x-timer: S1436438419.302921,VS0,VE141
  X-Archive-Orig-access-control-allow-origin: *
  X-Archive-Orig-last-modified: Wed, 08 Jul 2015 22:33:03 GMT
  X-Archive-Orig-expires: Thu, 09 Jul 2015 10:50:19 GMT
  X-Archive-Orig-accept-ranges: bytes
  X-Archive-Orig-vary: Accept-Encoding
  X-Archive-Orig-connection: close
  X-Archive-Orig-date: Thu, 09 Jul 2015 10:40:19 GMT
  X-Archive-Orig-via: 1.1 varnish
  X-Archive-Orig-content-length: 13
  X-Archive-Orig-x-cache: MISS
  X-Archive-Wayback-Perf: {"IndexLoad":359,"IndexQueryTotal":359,"RobotsFetchTotal":1,"RobotsRedis":1,"RobotsTotal":1,"Total":371,"WArcResource":10}
  X-Archive-Playback: 1
  X-Page-Cache: MISS

Extracting a WARC record

Once we’ve identified the offset and length of a particular record (in this case, an offset of 1260 bytes and a length of 1085 bytes), we can snip out an individual record like this:

$ tail -c +1261 hello-world.warc | head -c 1085

Making the CDX

To generate a content index (CDX) file, we have at least two options. There’s JWATTools:

$ jwattools cdx hello-world.warc

…(which created cdx.unsorted.out), or the cdx-indexer from OpenWayback:

$ cdx-indexer hello-world.warc > hello-world.warc.cdx

…(which created hello-world.warc.cdx).

Making the WARC

To create a WARC, we used wget:

$ wget --warc-file hello-world http://iipc.github.io/warc-specifications/primers/web-archive-formats/hello-world.txt

…which created the compressed hello-world.warc.gz file. These special block-compressed files are often used directly, but in this primer, we uncompress it so we can see what’s going on:

$ gunzip hello-world.warc.gz

…leaving us with hello-world.warc.

IPFS + Social Media as a simple implementation.

Good overview on IPFS and Filecoin services.

object是遵循Merkle DAG protofbuf 数据格式的block

实现了IPFS Core API，以及一些应用函数。

移植相对简单，将数据拷过去就差不多了。不过要注意如果原备份还运行的话Pear ID一样会有问题的。

感觉象是断网状态似的，不影响其它节点，只影响本地节点，有点不解。

IPFS CID 总是使用IPLD codec

因为支持未来算法，所以不会过时。

支持任何强加密算法。

这一句运行有问题！undefined！

使用js-ipfs-http-client更小，并且控制一个已经运行的节点。

建议使用独立节点，然后使用Http api。

IPFS主要的实现有 go-ipfs 和 js-ipfs。官方二进制包使用go实现。

Web Hosting with IPFS and Nix

ipfs.io documentation

IPNS

This is interesting, since the opening video for https://hypothes.is/ mentions the early web also - in this case, for its annotation features that were removed.

It seems to me that hypothes.is is even more powerful used on IPFS content identified by hash since that underlying content cannot change.

Thanks to both services I'm doing exactly this right now!

I think this is exactly what I've wanted - and what a lot of people have wanted - for a long time. It's certainly not the first time I've seen someone call for using hashes for referring to files, but the design and implementation behind this look like they do a lot of things right.

Same with this one. Dropping the /ipfs/ fixed it.

Also, this gives me an error from ipfs add:

$ hash=`echo "I <3 IPFS -$(whoami)" | ipfs add -q`
Error: Argument 'path' is required

Use 'ipfs add --help' for information about this command

Inspiring call to action by Kyle Drake

This is neat, as I understand it's like kio slaves in KDE or GIO, FUSE, etc.

Might be useful for exposing some of the tagged distributed storage systems as browsable filesystems with JIT access.