-
Notifications
You must be signed in to change notification settings - Fork 327
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
WIP: Version 6 #112
WIP: Version 6 #112
Conversation
For anyone still interested, this is on hold until at least March of 2019. Thanks for your interest! |
Interested. And it's March. Questions/comments:
Great "batteries included" stuff here! |
I wish the custom prefix routing #55 could be added in v5, instead of waiting for v6. Or streaming stuff can be a v7? |
Yes, as I stated in #70, I think streaming might be the wrong next step for Twirp. That would simplify this release considerably, changing it down to just being about paths.
@rmasp, I'm not sure exactly. The current branch just removes the Thanks for waiting, and sorry about the long delay on this. The reason for the delay is mostly that Twirp works pretty well today, and these routing changes will take some careful work to make the transition smooth. I think they're basically just cosmetic, too. |
And that's all I want/need. Again, for those of us exposing our Twirp APIs to 3rd parties, the /twirp is an issue. However, I branched That said, why not make the lead path part configurable at protoc code-gen time, with nil or empty string leaving that path piece out entirely? Default would be |
Right, I'd just like to avoid configuration at code-gen time if possible. Code generators are often buried in tool chains, so passing through options to them can be hard for people with any sophistication in their builds. An option in the |
Seems like the approach I like puts the onus of toolchain config on those if us who don't like the default. It's a no-op for anyone happy with the |
Yep, but if I can avoid adding that complexity for anyone, I'd prefer that. Good point though, you've moved my thinking on this a bit. |
Thanks for listening! Fwiw, I'm using twirp in a pipeline with gqlgen, which has a bunch of code gen configurability in a yml: https://gqlgen.com/config/ Imho as twirp sees wider adoption you'll get requests for a more configurable settings. Look at Django, Rails, ... I know it's a fine line between simplicity and configurability. Way to many tools (esp in the golang space) have more dials and buttons than Spock's science station. Have to design for the "just right porridge" and common case defaults. |
Shall we merge newer releases back to edit: I took the liberty created: #173 |
Merge v5.7.0 into v6_prelease
Fwiw, next week I'm moving my servers (and clients) to the v6_prerelease branch because of product requirements that include no /twirp in path. I'd really like to see v6 become mainline. |
Hi @spenczar , Also, thank you, twrip is awesome! |
Sorry, this PR is not going to be merged. The protocol v6 proposal was closed because both issues included in the release could not reach consensus (streaming: #3, and updating the path: #55). This doesn't mean that the features will never be implemented in some form, this only means that they don't depend on each other for v6. |
Issue #, if available:
#3
#55
#70
#80
Description of changes:
This PR holds all the v6 changes, so they can be merged into master at once. While working on these features, they will be merged into the v6_prerelease branch to make them available for adventurous users.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.