-
Notifications
You must be signed in to change notification settings - Fork 16
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
[email protected] not available on npm #15
Comments
Argh, that is really odd. For publishing both v0.3.0 and v1.0.0, I used the same method, which relies on Travis CI deployment after running And as you can see at the end of https://travis-ci.org/astorije/chai-immutable/jobs/65006776:
Also:
This is really odd... Does anyone have an idea of what happened here? I'll investigate, thanks for the report! |
Huh. That is strange. Can you try bumping to 1.0.1 and publishing again. If you can, use |
@seldo I am publishing automatically following this guide. I wish the I'll trigger v1.0.1 manually and hope that I can still using Travis CI to deploy v1.0.2+. |
Actually, after looking around, it might be possible. I gave it a shot in my latest commit, we'll see when Travis CI kicks in. |
Argh, it failed at the coveralls step as you can see here, for the same reason :-( Well, nevermind, I published manually and it seems to have worked. Assuming I was right when making sure there is no sensitive information in the output, here it is:
@seldo , do you have any idea of what happened for version 1.0.0? Do you think it came from npm or Travis CI (the latter will prevent me to publish automatically, but I'd be surprised if such feature is and stays broken)? Finally, is there any way I can re-build the tarball? It would be very unfortunate to have versions that fail every now and then on the list of releases... |
Hiya, I wanted to thank you for bringing this to our attention! What was happening was a race condition in the registry when identical versions of a package were being published at the same time from two different places. The first would succeed, the second would partly succeed and then fail at the end, and then roll back its work. In this case, it would end up rolling back the tarballs that the first publication had put into place, because they were for the same version. We're now detecting this situation & not nuking the tarballs from the first, successful publish, so we think this won't bite you again. Thanks again for telling us-- it was staring at the logs from the publications of |
Waouh @ceejbot that was nasty! Last question before I close this ticket: Is there any way to regenerate the Thanks!! |
I would suggest that you unpublish the version and just move forward by bumping the patch number. Versions are immutable in the registry, so there's no way to replace an existing published version with new data. Instead, just let semver help you here: if you have published |
Aaarrrrggghhh! What an unfortunate outcome:
Well, I guess |
My apologies for the inconvenience! It was definitely our bug :( |
Oh no @ceejbot I wasn't grumping at you, I was grumping at me :-) Just my two cents though, this seems like a behavior you would not expect (cf my terminal paste above)... |
Get following error when attempting to install chai-immutable:
Seems like https://registry.npmjs.org/chai-immutable/-/chai-immutable-1.0.0.tgz does not exist.
npm install [email protected] works correctly
The text was updated successfully, but these errors were encountered: