me as well. I don’t know npm teams even really notice this :(
I mean it has been a day is it possible it is EOL?
I’ve also been seeing this for a couple of days. I tried using
registry.npmjs.com, but the view I’m executing doesn’t return any results.
I would love to hear from npm support about this.
If it’s EOL, they should at least have notified us ahead.
Any info about it? Was it shut down completely?
This is the only API I have found to get machine readable version information for npm packages . At linuxserver.io we use this to trigger new releases for npm packaged programs and properly tag our containers. If there is an alternative Json endpoint I would really like to swap over.
As of now skimdb is hard coded into our logic.
Can a team member please chime in here?? I cannot find any information anywhere about this issue.
I contacted npm support.
Here is their response.
Thanks for reaching out.
We recommend using replicate.npmjs.com instead. This serves the same purpose as skimdb, but with the added bonus of including scoped public packages.
Please let us know if you need additional assistance. We’re always happy to help.
Thanks https://replicate.npmjs.com/registry/packagename is exactly what I was looking for . Much appreciated .
This is such an asshat move by the NPM organization. This caused havoc for many services, since they never alerted anyone of their decitition to shut down the one server people relied. The guys over at RunKit have had massive issues trying to put out the fires caused by this, disrupting a service many people rely.
Hi there folks.
I’m an SRE on the npm infrastructure team, last week we planned to shut down skimdb.npmjs.com because it’s difficult to maintain and the source of many outages and on-call toil, we made plans to disclose this fact to the community responsibly and repoint the domain to replicate.npmjs.com.
So what went wrong? Human error, with a few other fires happening this week, the work to decommission the service happened out of sync with the rest of the plan.
We’re truly sorry for the distress, anger and extra labor this has caused, and we will carry out a retrospective to document how it happened like this and avoid similar happening again.
I have repointed
replicate and I’m working on a blog post to get the word (and apology) out there.
Thanks for repointing the domain, however, it seems like there is an issue with retrieving data from some of the couchdb views (this was already happening on the replicate domain and I opened a ticket with npm support but haven’t heard back yet).
Will the old endpoints still be valid? If not, is there any documentation on how to get the same information? I’m still able to see my list of repositories through npmjs.com: https://www.npmjs.com/~doowb so I’m assuming there’s still a way to get that list.
can you please open a ticket in email@example.com so we can get high visibility into this and address it properly.
I already opened a ticket but I haven’t heard back yet.
excellent, thank you.
I’ll look up in the system and ensure it gets higher priority.
That’s an exemplary response, Wes. Thank you for disclosing a bit what went on behind the curtain veil.