Road to 1.0.0

@ryanseys

What remaining features would we like to offer before we make the jump?

API complete for datastore and storage (@ryanseys)
To be able to tell users that they'll love it (@ryanseys)

@silvolu

@silvolu @ryanseys @rakyll please go through and tag/create issues with the 1.0.0 milestone.

Sorry for missing the big party -- I'm still half offline, half awake :)

Is there any more context about 1.0 versioning? Does it indicate stability and/or API freeze?

Haha, no problem. We had a discussion in the hangout about 1.0.0, and have some differing opinions. I thought we should huddle up so we and the public knows when they can start safely using our project.

Semver lets us be as reckless as we want with our API < 1.0.0, however I propose now that we have more eyes on the project and users trickling in, we should offer more reliability. After some of our bigger issues we tackled (instantiation duality [love this phrase] and dataset.key revamp), I believe we are ready for 1.0.0 (maybe 1.0.0-beta?) and to start incrementing our versions as we add features / fixes accordingly.

If we have more things we want to fit in before 1.0.0, just tag 'em so we're all on the same page.

Does it indicate stability and/or API freeze?

Sorry I didn't catch this specific question the first time. 1.0.0 indicates we won't break the API until 2.0.0. We can add endless features and bug fixes in, but as soon as we have to break backwards compatibility, we must bump 2.0.0. And likewise, we wouldn't want to release 2.0.0 just because we added a hot new feature.

I'm against 1.0.0 right now because I know 2.0.0 will be right around the corner and to me, 1.0.0 suggests another celebration & blog post 🍰 . If we're just breaking stuff, even small minor things in small minor ways like .key() and bumping versions because "semver" tells us to, that's bad. The kinks still need to be massaged out, and we also should have feature completeness even in a less-than-optimal way, e.g. ACL for storage, bucket.file(), support for all current api calls for storage and datastore in some way.

What do I want for 1.0.0? I want to be able to tell users that I think they'll love it and that I think they'll find it suits their needs better than google-api-nodejs-client, and that they shouldn't need to update to 2.0.0 for at least 4 months. I kinda just pulled 4 months out of nowhere, but the idea is that we shouldn't have a major version already coming around the mountain.

I updated the comment at the top with a couple check-boxes to make it more clear.

Can you tag the issues as well, so we have a clear milestone? It won't only help us, but any curious users who want to know when they can expect stability.

As part of (2), do you plan to have a list of projects using gcloud-node? In case, count me in.

Definitely, what's your project?

It will be released end of this month / start of Jan. Anyway, it's a web audio editor, next iteration of http://hya.io.
If you are going to provide a list, I can send the project as soon as I release.

That's awesome! Let us know if we can help with anything. We did have a
list, but I think it's in a closed issue now. Will have to dig it up and
tag you over when I find it.