V8 version for 7.0.0 release

[targos]

I'm opening this to discuss about the V8 version that we would like to integrate in the v7.0.0 release and how it will affect the timing of the given release.

The LTS plan states that the current v6 release will transition to LTS when v7 is cut and this should happen in October.

According to this web page, next Chromium releases are estimated as:

• 53: Sep 6th, 2016
• 54: Oct 18th, 2016

It means that if we want to include V8 5.4 in the v7 version, we will have to wait until the second half of October. I think we should do that if possible and maybe start working very early on the integration of this version by skipping 5.3.
IMO this doesn't have to block the release of v6 LTS on October 1. We could have an overlap of current and LTS for a couple weeks.

Ref: #7824 (comment)
cc @ofrobots @srl295 @nodejs/lts @nodejs/v8

[fhinkel]

/cc @natorion

[natorion]

Considering that the LTS release is going to live for some time I would try to maximize the time of "official support". This means go with 5.4. Waiting for V8 5.4 has a few other benefits too:

• More mature Interpreter
• More mature async/await implementation including debugging (no clue though if this is going to be shipped or staged though)

[targos]

v7.0.0 is not going to be an LTS release. It will live for about 8 months.
The next LTS will be based on v6.x, so it will ship with V8 5.0 or 5.1 (hopefully).

[bricss]

It would be nice to see V8 5.4 in the node v7.0.0 🐋

[Fishrock123]

¯_(ツ)_/¯ as @targos says it'll only be around just over a half a year.

[Fishrock123]

V6 LTS doesn't need to release until mid-way through october somewhere though.

[mhdawson]

I lean towards delaying V7 to wait for 54, while at the same time not having that determine when v6 becomes LTS. The net would either be no new stable for a few weeks which seems acceptable.

[ronkorving]

What are the benefits of 54 compared to 53?

[bricss]

@ronkorving it's all about performance and stability improvements with ES7 features support.

[jasnell]

We will likely be cutting the v7 branch off master in early september so we can begin making sure things are stable for the release. If we can land a pre-stable version of 54 there it would allow us to do additional testing to make sure things are going to work. It would be a bit odd, however, since we don't typically land v8's until they reach stable. I just don't like the idea of doing a series of v7 betas or release candidates with an older v8, then update v8 to 54 at the last minute before the actual v7.0.0 release.