Skip to content

prepare next release#275

Merged
piotr-oles merged 18 commits intomasterfrom
beta
May 2, 2019
Merged

prepare next release#275
piotr-oles merged 18 commits intomasterfrom
beta

Conversation

@phryneas
Copy link
Copy Markdown
Contributor

No description provided.

phryneas and others added 11 commits April 22, 2019 23:56
- do not create a IncrementalChecker instance on the test thread
- use worker-rpc so the tests can actually access the checker instance run by the service
- fix include path in "tsconfig-imports.json" so the "should resolve *.vue in the same way as TypeScript" test actually compiles files
- add two counter checks to make sure the compiler actually throws errors it finds at all
actually use the correct `useTypescriptIncrementalApi` value for some of the linter tests that were using the default value before
fix a bug where, as `tslint` was set to true and the files being checked had no parent folder with a tslint.json, tslint in these tests never found any warnings
also fixes a failing test because a syntactic error was counted as linter warning
when the options `checkSyntacticErrors: false` and `useTypescriptIncrementalApi: true` both were active, no semantic errors were emitted as soon as the first syntactic error was encountered
this patches the `typescript` import to override that behavior

see discussion in #257
improve generation of diagnostics, fix linter tests
# [1.2.0-beta.4](https://github.com/Realytics/fork-ts-checker-webpack-plugin/compare/[email protected]@beta) (2019-04-23)

### Bug Fixes

* **tests:** fix linter tests that were doing nothing ([d078278](d078278))
* **tests:** linter tests - useTypescriptIncrementalApi usage ([e0020d6](e0020d6))

### Features

* **apiincrementalchecker:** improve generation of diagnostics ([ae80e5f](ae80e5f)), closes [#257](#257)
Add documentation on how to use hooks on a compiler instance.
Explain the purpose of arguments passed to plugin hooks.
docs: 📖 add hooks documentation
@phryneas
Copy link
Copy Markdown
Contributor Author

So we'd need to make a new release for the changes from #274 to land in the README. Also this lands a few test improvements and feat(apiincrementalchecker): improve generation of diagnostics.

Since the beta branch is protected I don't really know how to best prepare merging this into master @piotr-oles, can you help here?

@piotr-oles
Copy link
Copy Markdown
Collaborator

I will take a look today’s evening :)

@piotr-oles
Copy link
Copy Markdown
Collaborator

I've resolved conflicts in the CHANGELOG.md - now everything should be good. I think semantic-release will release new beta version (1.2.1-beta.1).

piotr-oles
piotr-oles previously approved these changes Apr 30, 2019
# [1.3.0-beta.1](https://github.com/Realytics/fork-ts-checker-webpack-plugin/compare/v1.2.0...v1.3.0-beta.1@beta) (2019-04-30)

### Bug Fixes

* **tests:** fix linter tests that were doing nothing ([d078278](d078278))
* **tests:** linter tests - useTypescriptIncrementalApi usage ([e0020d6](e0020d6))
* **tests:** rework vue integration tests ([5ad2568](5ad2568))

### Features

* **apiincrementalchecker:** improve generation of diagnostics ([ae80e5f](ae80e5f)), closes [#257](#257)
@phryneas
Copy link
Copy Markdown
Contributor Author

Great! Who wants to pull the trigger? 😄

johnnyreilly
johnnyreilly previously approved these changes May 1, 2019
Copy link
Copy Markdown
Member

@johnnyreilly johnnyreilly left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

First time with the semantic release hotness!

@johnnyreilly
Copy link
Copy Markdown
Member

It looks like we can't merge as Travis isn't running or something... @piotr-oles can you investigate please?

piotr-oles
piotr-oles previously approved these changes May 1, 2019
@piotr-oles
Copy link
Copy Markdown
Collaborator

We have all these problems because not all commits were merged into beta :) Travis didn't want to build because when we merged master to beta, semantic-release created new release and updated CHANGELOG.md with [skip ci] in the commit message. That why travis didn't want to start build. I have created a new commit with some spacing changes in the CHANGELOG.md to trigger build again without [skip ci] in the last commit :)

@johnnyreilly
Copy link
Copy Markdown
Member

Cool - is there a way we can avoid this happening in future? I'm keen we have a release process that just works™️ 😊

Is this happening just because this will be our first release end to end with semantic release? So in future these issues won't present?

I'm a little confused / concerned with the problems we've been facing getting semantic release bedded in. Hopefully we can get past this and into a good release flow 🤞

Copy link
Copy Markdown
Contributor

@swashata swashata left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have found a typo in the docs. Please accept the changes in this PR too :)

Comment thread README.md Outdated
@piotr-oles
Copy link
Copy Markdown
Collaborator

The most problematic is generation of the CHANGELOG.md - I think that's why it's not in the default semantic-release setup. After we merge to the master it creates additional commit with updated CHANGELOG.md. This commit doesn't appear in the beta branch and this creates merge conflicts in the next pull request (as we modify CHANGELOG.md in every beta release). We can manualy merge the master barnch into the beta after release but maybe we could automate this also (that was the point of the semantic-release)

@johnnyreilly
Copy link
Copy Markdown
Member

My take away from that is that semantic release doesn't have a good workflow with regards to changelogs. That's super disappointing.

I'm rather hesitant about us creating our own automation around this as that may end up being another thing to maintain as semantic release evolves. If there's not a good out of the box experience that we can have I'm inclined to say "well nevermind - let's just have the GitHub releases instead"

It means we have to be online to see them and if we were ever to move the plugin elsewhere we'd need to recreate manually. Neither of those is a deal-breaker in my eyes.

I value a simple release workflow above other concerns. If you want to drop the changelog generation I'm fine with that. 😊

piotr-oles
piotr-oles previously approved these changes May 1, 2019
@phryneas
Copy link
Copy Markdown
Contributor Author

phryneas commented May 1, 2019

Since the CHANGELOG is automatically generated from commit messages, I think we won't lose any metadata when dropping the automatic generation - even if this project were to move away from github some day in the future. The commit history won't be lost.

@piotr-oles
Copy link
Copy Markdown
Collaborator

I've created PR to drop CHANGELOG.md generation and remove emoji as not everyone is using them 😢

@piotr-oles
Copy link
Copy Markdown
Collaborator

The last thing we should remember is to use "Rebase and merge" when merging from the beta to master - that will not create additional commit and will create a tag with the new version on the last commit that was also on the beta branch. Thanks to that master and beta will be in sync (otherwise master will be 1 commit ahead of beta - the merge commit). I will create PR with docs update for the deployment process.

@johnnyreilly
Copy link
Copy Markdown
Member

Good catch - thanks!

Piotr Oleś added 2 commits May 1, 2019 16:15
Not everyone uses emoji in commits - so it's better to remove them than
having two styles of commits
@piotr-oles piotr-oles merged commit e9ee9ad into master May 2, 2019
@piotr-oles
Copy link
Copy Markdown
Collaborator

🎉 This PR is included in version 1.3.0 🎉

The release is available on:

Your semantic-release bot 📦🚀

@johnnyreilly
Copy link
Copy Markdown
Member

How did you get it to work? Can you just not rebase on a phone?

@phryneas
Copy link
Copy Markdown
Contributor Author

phryneas commented May 2, 2019

Nope, looks like this was one last normal merge. I guess we'll have to rebase that one back to beta one last time and then we'll be golden as long as we don't push to master.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants