VulcanJS / Vulcan-Starter

VulcanJS starter repo. Use as a base for your own VulcanJS projects.
MIT License
127 stars 88 forks source link

current vision to vulcanJS #169

Closed TomFreudenberg closed 4 years ago

TomFreudenberg commented 4 years ago

Hi @SachaG

before trying and starting to think about using VulcanJS I would like you to ask about the effort is spent on VulcanJS currently and in future.

Is this becoming "a dead project" or are you continuing pushing it forward ?

It would be great to get a simple answer before I may start a project using VulcanJS.

Currently the starter for me does not work and there are a number of old dated issues and PRs.

Thanks in any way Tom

eric-burel commented 4 years ago

Hi Tom,

Will answer for Sacha, I am currently taking care of issues and PR.

Short answer: the project is still active. We are preparing a migration toward NPM and Next.js, as a replacement for Meteor, that will make up the "v2" of Vulcan. The Meteor "v1" will still be maintained, for a very good reason: most of the code will be shared (schema system, graphql code generation, React components....).

You should come to the Slack and describe your project more precisely so we can tell you if the current implementation of Vulcan is suited for you, or what technologies you may want to use instead.

Regarding the broken Starter, the issue is sadly on Meteor side. Server build tend to fail when handling some packages (Material ui icons, no-ssr, but maybe some others), I can't tell why at this point. You should raise this issue to the Meteor community on the forum. Those kind of annoying build issues are the very reason we have been wanting to leave Meteor for a while. Also note that we just updated to Meteor latest version just a few days ago, which explains a few problems during install. We don't do that often hopefully so expect things to improve within a few days, when we figure all possible errors that can happen depending on the environment.

Regarding old open issues, this makes sense as Vulcan is one of the oldest framework around. Some issues have been around for years simply because we lacked the correct technological answer to them (need more maturity of either React or Apollo, not emergencies, feature requests, unclear issues...). A good example: e2e testing. The ecosystem was just nonsensical until Cypress came out, so we left a PR open for basically 2 years because they simply were no good lib to handle e2e tests. Current version of Vulcan is relatively stable feature wise, we did code most of the things that are actually necessary to develop an app quickly, and we fix bugs as they come.

Also we have a policy of not closing issue when they are not active, contrary to many frameworks and lib that closes issues eagerly. They have better stats but they just don't care about legitimate issues that do not receive enough attention.

TomFreudenberg commented 4 years ago

Hi Eric,

Thanks a lot for your answer in details.

I came along after reading this article from Sascha (again)

https://medium.com/@sachagreif/an-open-letter-to-the-new-owners-of-meteor-353d64780b20

So yeah, leaving meteor now makes sense and is consequent. Very sad story for that "unicorn" project.

See you on Slack.

Cheers Tom