nodejs.org icon indicating copy to clipboard operation
nodejs.org copied to clipboard

Add blog post publishing guidelines

Open mcollina opened this issue 6 months ago • 18 comments

After discussing with @rginn, I'm adding the updated guidelines for publishing blog posts on the Node.js website.

mcollina avatar Jun 12 '25 14:06 mcollina

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name Status Preview Updated (UTC)
nodejs-org ✅ Ready (Inspect) Visit Preview Jun 12, 2025 2:01pm

vercel[bot] avatar Jun 12 '25 14:06 vercel[bot]

cc @nodejs/tsc

mcollina avatar Jun 12 '25 14:06 mcollina

Codecov Report

:white_check_mark: All modified and coverable lines are covered by tests. :white_check_mark: Project coverage is 73.40%. Comparing base (c375408) to head (b395a81). :warning: Report is 184 commits behind head on main. :white_check_mark: All tests successful. No failed tests found.

Additional details and impacted files
@@            Coverage Diff             @@
##             main    #7860      +/-   ##
==========================================
- Coverage   75.44%   73.40%   -2.04%     
==========================================
  Files         101      173      +72     
  Lines        8305    14949    +6644     
  Branches      218      468     +250     
==========================================
+ Hits         6266    10974    +4708     
- Misses       2037     3973    +1936     
  Partials        2        2              

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

codecov[bot] avatar Jun 12 '25 14:06 codecov[bot]

Why not putting tsc as code owner ?

AugustinMauroy avatar Jun 12 '25 15:06 AugustinMauroy

Why not putting tsc as code owner ?

That's an implementation detail, which is governed by governance. This PR updates the governance. Surely, lets' do that as well.

mcollina avatar Jun 12 '25 15:06 mcollina

I wonder if we should label the blog posts somehow differently between "official" content and community content, and give project contributors sufficient trust to manage the community content. Putting an approval chain on every blog post gives me the "you need to get approval from the management for everything you do" megacorp vibes and seems out of place for a community-driven OSS project.

joyeecheung avatar Jun 14 '25 08:06 joyeecheung

What kind of community content are we publishing on this blog? The vast majority posts are releases, and the rest are various kind of official announcements that represents the project in one form or another.

mcollina avatar Jun 14 '25 12:06 mcollina

What kind of community content are we publishing on this blog?

For example https://nodejs.org/en/blog/announcements/making-nodejs-downloads-reliable

The vast majority posts are releases, and the rest are various kind of official announcements that represents the project in one form or another.

I think that's because we are mostly only doing the bare minimum for content curation, which is not to say this is the best content we could put out. Personally I want to see the Node.js blog to be more like https://v8.dev/ - I participated in the writing of some of the content there despite not being a V8 team member/Google employee and I appreciated that we only needed technical review from V8 team members that reviewed the same technical work the posts were talking about. I think that's the kind of community content we should try to nurture, even if they are dwindling now, and treating everything on the blog as if they can only be official announcements means that the content would be biased towards official announcements. That's why I suggest the approval process should be based on the category of the content, unless we want to fully abandon the official blog as a channel to put out technically insightful content.

joyeecheung avatar Jun 16 '25 11:06 joyeecheung

I don't think we need these guidelines explicitly stated, they seem to be implied. I feel like if we set @nodejs/TSC @nodejs/marketing as the CODEOWNERS of non-release blogs, it implies all of this information.

If we do insist on the inclusion of these guidelines, maybe they can go in the new "Adding Pages" document

avivkeller avatar Jun 20 '25 22:06 avivkeller

For example https://nodejs.org/en/blog/announcements/making-nodejs-downloads-reliable

I'm not sure how this counts as "community content". It was written by a member of the website team discussing the changes that were made to improve release downloads. That would put it in the project content category.

jasnell avatar Jun 21 '25 01:06 jasnell

I'm not sure how this counts as "community content". It was written by a member of the website team discussing the changes that were made to improve release downloads. That would put it in the project content category.

I think we might just have different definition of the community - here I mean "people who are not employed by the foundation or chartered to take on responsibilities, free of charge". I think even in this case, we should grant people sufficient trust. For example, I would say a review from the TSC or the marketing team for this article doesn't look super helpful unless any of us was specifically involved in the work described. Even as a member of the TSC I would trust the website team on the reviews of content about the development of the website more than myself.

What I am seeing from the current guideline is a lack of trust. I think there is a better way to go about this that would be closer to how PRs work in the organization in general...

joyeecheung avatar Jun 23 '25 20:06 joyeecheung

Explaining my suggestions somewhere else in case it gets folded:

  1. The marketing team has a bus factor of 2. AFAIK they are both based on U.S. west coast and seem to be in proximity to each other quite often. This is not a very optimal set up IMO.
  2. There is no fast track in case incorrect information is missed during reviews and gets published. And again it's not great if this is bottlenecked on two people based on the U.S. west coast.
  3. We should recognize neither the TSC nor the marketing team necessarily have the correct expertise to review all content. I think in that case we should trust the teams that have the correct expertise on this. That also altogether lower the bus factor even further.

If there is anything not compliant with the legal and branding requirements, 48 hours on business days + 7 days should be sufficient to catch problems, if that's already considered adequate for Node.js core reviews done by people who are not bound by employment to review PRs.

joyeecheung avatar Jun 23 '25 21:06 joyeecheung

The marketing team has a bus factor of 2. AFAIK they are both based on U.S. west coast and seem to be in proximity to each other quite often. This is not a very optimal set up IMO.

We've never applied this same rubric to any other teams that I'm aware of. We shouldn't single out the marketing team just for this.

jasnell avatar Jun 24 '25 02:06 jasnell

We've never applied this same rubric to any other teams that I'm aware of.

I do remember this was brought up in the build team before (which was where I learned the word "bus factor") in the context of sharing accesses to various resources and there is a vague consensus to increase the bus factor whenever we can (as in, I never saw anyone against increasing the bus factor in the limited build team discussions I was in).

joyeecheung avatar Jun 24 '25 16:06 joyeecheung

Did the PR become stale? Is there anything needed here to unblock?

ovflowd avatar Aug 02 '25 16:08 ovflowd

Bump @mcollina

avivkeller avatar Aug 12 '25 23:08 avivkeller

Sorry, I’ll get back to this at some point.

mcollina avatar Aug 30 '25 22:08 mcollina

@mcollina is this still in the works?

avivkeller avatar Nov 11 '25 16:11 avivkeller