Draft of 0.4.0 blog post #30

kyledrake · 2016-02-16T02:20:24Z

This announces 0.4.0. It may need some more specific examples than provided if there's new related functionality, but the main thing is that it makes sure people know this is a breaking protocol change and that they need to upgrade.

whyrusleeping · 2016-02-16T03:58:39Z

cc @RichardLitt

whyrusleeping · 2016-02-16T04:15:30Z

on top of this we have:

utp: ipfs can now use this udp based protocol for inter-node communications
ipfs now uses gx for vendoring
libp2p has been extracted from the core ipfs codebase
ipns has been sped up a bit

dignifiedquire · 2016-02-17T14:22:54Z

src/10-ipfs-0-4-0-released/index.md

+
+[IPFS (go) 0.4.0](https://github.com/ipfs/go-ipfs/releases/tag/v0.4.0) has been released! Among the many changes are a revamped implementation of the IPFS communication protocols, increased performance, improvements to IPNS, many bugfixes, and a couple new features to make IPFS even more powerful.
+
+This is a **breaking change** release, with a major refactor and upgrade to the IPFS networking protocol. Because of the refactor, **all IPFS (go) daemons earlier than 0.4.0 will not be able to communicate with the newest version**. It is strongly recommended that everyone running an IPFS node upgrade to the latest version as soon as possible, as these nodes will, after a certain time, no longer be able to communicate with the majority of the network until they are upgraded.


upgrades to the latest version

Both are grammatical.

I would add a subheading before this: ### On Breaking Changes. Just style, though.

dignifiedquire · 2016-02-17T14:25:34Z

Files api is missing as well as I can tell

RichardLitt · 2016-02-17T17:18:02Z

src/10-ipfs-0-4-0-released/index.md

+
+So by supporting as many muxers as we can, we get to choose the best multiplexors for the job. It also makes it much easier to implement the IPFS protocols in a new language. And of course, if a better multiplexor standard comes along, it will be easier to upgrade IPFS to support it in the future.
+
+In addition to the multiplexor changes, the protocol revamp has also improved efficiency and performance in a few important ways, including the elimination of a double wrapping of the length prefixer, and the removal of some unneccessary round trips between nodes.


s/multiplexor/multiplexer

RichardLitt · 2016-02-17T17:18:53Z

LGTM. A few things to add, some comments, but overall, this is pretty awesome.

RichardLitt · 2016-02-17T19:04:54Z

src/10-ipfs-0-4-0-released/index.md

+
+[IPFS 0.4.0](https://github.com/ipfs/go-ipfs/releases/tag/v0.4.0) has been released! Among the many changes are a revamped implementation of the IPFS communication protocols, increased performance, improvements to IPNS, many bugfixes, and a couple new features to make IPFS even more powerful.
+
+### On Breaking Changes


I guess breaking and changes shouldn't be capitalised for consistency. Sorry.

jbenet · 2016-02-22T12:16:43Z

Notes:

great to cover the breaking change and reasoning! (btw this post builds on https://ipfs.io/blog/9-v04x-migration/ and will be linked to from it later)
the majority of this post should be about the 0.4.0 improvements. the breaking change is minor compared to everything else. I'd expect 90% other stuff, 10% breaking change.
There is a ton to cover and not a single person will be able to do this. We should enumerate all the things that have changed (@whyrusleeping make a CHANGELOG entry?) and from there craft an outline of what things to highlight, broken up into sections.
then we can assign sections to different people to ensure this is still manageable.
yes i expect all the content to be very big. this is a huge release for us.
if desired, for big things (like the files api, or the new pinning implementation) we can write own blog posts and link to them from this one.

jbenet · 2016-02-22T12:16:58Z

thanks @kyledrake for pushing on this! 👍 👍

jbenet · 2016-02-22T18:42:03Z

Here's a great example of a great update of this magnitude:

https://about.gitlab.com/2016/02/22/gitlab-8-5-released/

RichardLitt · 2016-03-08T18:15:02Z

Next steps to edit this: basically, we need to know everything that has changed to meaningfully change the copy. This depends on @whyrusleeping's list of things in 0.4.0 that are different.

whyrusleeping · 2016-03-08T21:41:31Z

src/10-ipfs-0-4-0-released/index.md

+
+## Why we're changing the protocol
+
+The most important change was to allow IPFS implementations to use pluggable stream multiplexers, such as [yamux](https://github.com/hashicorp/yamux), [spdystream](https://github.com/docker/spdystream), or [muxado](https://github.com/inconshreveable/muxado). Instead of locking IPFS permanently into a single multiplexer that won't work for every language or situation, this change allows the implementations to implement the multiplexers of their choosing.


"... and then nodes will negotiate which multiplexer to use between them during the initial connection handshake"

whyrusleeping · 2016-03-14T23:52:42Z

any update here? cc @RichardLitt

whyrusleeping · 2016-03-21T16:43:59Z

closing in favor of #37

draft of 0.4.0 blog post

980dd09

dignifiedquire mentioned this pull request Feb 17, 2016

Sprint: February 16 (Tuesday) ipfs/team-mgmt#91

Closed

dignifiedquire reviewed Feb 17, 2016
View reviewed changes

RichardLitt reviewed Feb 17, 2016
View reviewed changes

grammatical fixes, IPFS (go) -> just IPFS

abd3a1f

RichardLitt reviewed Feb 17, 2016
View reviewed changes

RichardLitt mentioned this pull request Mar 8, 2016

Make a Changelog for 0.4.0 whyrusleeping/todo#8

Closed

whyrusleeping reviewed Mar 8, 2016
View reviewed changes

whyrusleeping mentioned this pull request Mar 17, 2016

post: 0.4.0 release #37

Merged

whyrusleeping closed this Mar 21, 2016

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Draft of 0.4.0 blog post #30

Draft of 0.4.0 blog post #30

kyledrake commented Feb 16, 2016

whyrusleeping commented Feb 16, 2016

whyrusleeping commented Feb 16, 2016

dignifiedquire Feb 17, 2016

RichardLitt Feb 17, 2016

RichardLitt Feb 17, 2016

dignifiedquire commented Feb 17, 2016

RichardLitt Feb 17, 2016

RichardLitt commented Feb 17, 2016

RichardLitt Feb 17, 2016

jbenet commented Feb 22, 2016

jbenet commented Feb 22, 2016

jbenet commented Feb 22, 2016

RichardLitt commented Mar 8, 2016

whyrusleeping Mar 8, 2016

whyrusleeping commented Mar 14, 2016

whyrusleeping commented Mar 21, 2016


		[IPFS (go) 0.4.0](https://github.com/ipfs/go-ipfs/releases/tag/v0.4.0) has been released! Among the many changes are a revamped implementation of the IPFS communication protocols, increased performance, improvements to IPNS, many bugfixes, and a couple new features to make IPFS even more powerful.

		This is a breaking change release, with a major refactor and upgrade to the IPFS networking protocol. Because of the refactor, all IPFS (go) daemons earlier than 0.4.0 will not be able to communicate with the newest version. It is strongly recommended that everyone running an IPFS node upgrade to the latest version as soon as possible, as these nodes will, after a certain time, no longer be able to communicate with the majority of the network until they are upgraded.


		So by supporting as many muxers as we can, we get to choose the best multiplexors for the job. It also makes it much easier to implement the IPFS protocols in a new language. And of course, if a better multiplexor standard comes along, it will be easier to upgrade IPFS to support it in the future.

		In addition to the multiplexor changes, the protocol revamp has also improved efficiency and performance in a few important ways, including the elimination of a double wrapping of the length prefixer, and the removal of some unneccessary round trips between nodes.


		[IPFS 0.4.0](https://github.com/ipfs/go-ipfs/releases/tag/v0.4.0) has been released! Among the many changes are a revamped implementation of the IPFS communication protocols, increased performance, improvements to IPNS, many bugfixes, and a couple new features to make IPFS even more powerful.

		### On Breaking Changes


		## Why we're changing the protocol

		The most important change was to allow IPFS implementations to use pluggable stream multiplexers, such as [yamux](https://github.com/hashicorp/yamux), [spdystream](https://github.com/docker/spdystream), or [muxado](https://github.com/inconshreveable/muxado). Instead of locking IPFS permanently into a single multiplexer that won't work for every language or situation, this change allows the implementations to implement the multiplexers of their choosing.

Draft of 0.4.0 blog post #30

Draft of 0.4.0 blog post #30

Conversation

kyledrake commented Feb 16, 2016

whyrusleeping commented Feb 16, 2016

whyrusleeping commented Feb 16, 2016

dignifiedquire Feb 17, 2016

Choose a reason for hiding this comment

RichardLitt Feb 17, 2016

Choose a reason for hiding this comment

RichardLitt Feb 17, 2016

Choose a reason for hiding this comment

dignifiedquire commented Feb 17, 2016

RichardLitt Feb 17, 2016

Choose a reason for hiding this comment

RichardLitt commented Feb 17, 2016

RichardLitt Feb 17, 2016

Choose a reason for hiding this comment

jbenet commented Feb 22, 2016

jbenet commented Feb 22, 2016

jbenet commented Feb 22, 2016

RichardLitt commented Mar 8, 2016

whyrusleeping Mar 8, 2016

Choose a reason for hiding this comment

whyrusleeping commented Mar 14, 2016

whyrusleeping commented Mar 21, 2016