-
Notifications
You must be signed in to change notification settings - Fork 10.5k
[release/3.1] Redo yarn install
#42173
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
Conversation
dougbu
commented
Jun 13, 2022
- get latest versions
|
Hi @dougbu. If this is not a tell-mode PR, please make sure to follow the instructions laid out in the servicing process document. |
|
Not sure this is quite what we want but I did the following:
I suspect this isn't enough to clean up the branch because I got lots of warnings in the last step in most projects, mostly |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is the only mention of eventsource in a package.json file in this branch. The mention of @types/eventsource above is also unique. Should one or both be removed while I'm at it❔
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this is supposed to be here - should just update it to 1.1.1
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'll just leave it alone because yarn.lock shows 1.1.2 anyhow.
|
Hey @dotnet/aspnet-build, looks like this PR is something you want to take a look at. |
|
@Pilchie should we treat this as |
|
Build break may be due to an old version of |
Looks like it's because >12.20.20 of @types/node package need Typescript version 3.7+ |
Any downside to updating the TS version❔ |
|
Probably new warnings/errors showing up in the code, but it should be fine otherwise, webpack should be controlling how the app is packaged and what format (ES5 etc.) to use. |
|
Also hitting lots of version resolution issues w/ System.Text.Encodings.Web. Did you make a change in that area recently @wtgodbe❔ Any issues if I pin the newer version in the affected bunch o' projects❔ @dotnet/aspnet-blazor-eng the other warning about STEW is But we create a non-shipping /fyi @BrennanConroy most of those issues are in SignalR test, benchmarks, and sample projects. |
I believe I did, but don't remember seeing any errors - pinning should be fine |
|
Going back to the |
No, there is a reason we depend on it |
|
Well, it is request that's bringing in too new a version of "@types/request@^2.47.1":
version "2.48.8"
resolved "https://registry.yarnpkg.com/@types/request/-/request-2.48.8.tgz#0b90fde3b655ab50976cb8c5ac00faca22f5a82c"
integrity sha512-whjk1EDJPcAR2kYHRbFl/lKeeKYTi05A15K9bnLInCVroNDCtXce57xKdI0/rQaA3K+6q0eFyUBPmqfSndUZdQ==
dependencies:
"@types/caseless" "*"
"@types/node" "*"
"@types/tough-cookie" "*"
form-data "^2.5.0"I'll bump |
|
I think this and related changes are fine for |
Like I said, errors 😆 |
|
Glad you're pleased @BrennanConroy 😃 I could either mosey through the specific errors or try porting the portions of 739f585 that updated the |
|
That change isn't going to be helpful, Typescript was already at 4.x by then. |
I'm inspired but won't get back to it for a couple of hours. @vseanreesermsft @mmitche how much time do I have to get release/3.1 building today❔ |
|
This isn't quite enough but I'll look at the remaining errors tomorrow |
|
Current iteration will likely fail w/ |
|
Errors already visible in macOS build: |
|
@BrennanConroy I also see a lot of new apparent syntax errors starting at https://dev.azure.com/dnceng/public/_build/results?buildId=1825044&view=logs&j=f43b187f-6c15-531c-a59b-e4fba4e9ed20&t=9e357cb9-1774-5509-1b39-7d8136be93e2&l=1122 with Suggestions appreciated❕ |
- get latest versions
|
@dotnet/aspnet-blazor-eng please comment on the many problems compiling in src/Components/Web.JS/Microsoft.AspNetCore.Components.Web.JS.npmproj w/ the new TypeScript version. That's the remaining build problem in this PR. |
|
@BrennanConroy some SignalR tests are also failing, specifically in client-ts.npmproj and SignalR.Npm.FunctionalTests.npmproj. I suspect the issues are compilation errors that aren't caught when building. But one message worries me: Do we need to update ts-jest too❔ |
|
Let's step back here, what is the overall goal? Is it just to fixup things like getting eventsource to be version 1.1.2? If so we should try not to update the entire world in a patch branch and do the minimal updates needed. |
That's my aim at least.
I agree but the rolling stone is gathering moss. |
|
I'm wondering for example if we could undo the TypeScript upgrade and downstream changes. Then, turn left instead of right and for example use |
- avoid package resolution (`MSB3277`) warnings
- avoid version that dropped support for TypeScript < 3.7 - see DefinitelyTyped/DefinitelyTyped#55430
dougbu
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Trying my alternative to bumping the typescript version…
| "resolutions": { | ||
| "url-parse": ">=1.5.8" | ||
| "//": "Avoid version that drops support for TypeScript < 3.7", | ||
| "@types/node": "<12.20.21" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
With this and removing "@types/node": "^10.9.4" under devDependencies above, 12.20.20 gets resolved everywhere. Is that better or worse than current settings, which resolve 10.17.60 for everything except request (which gets 12.20.20)❔
|
General story w/ the last few commits is that we use old enough versions of |
|
@vseanreesermsft @mmitche I don't remember enabling auto-merge and am sorry this went in. Did the new commit cause problems anywhere❔ Should I revert it or perhaps force |
|
I believe the 3.1 build is already staged, in which case we'd be fine - generally though, with servicing builds, the problem is not the new commit, it's the new build which triggers new PRs, so reverting actually makes things worse (and force-pushing causes new problems if the updated dependency has already flowed, since now nonexistant commits are getting referenced). The best thing to do is to cancel the triggered build in time, or close the opened dependency update PRs. If any of those auto-opened PRs have already been merged, it's pretty much too late & we have to let everything flow |