Cassie Heart on Deescuss

It has been a full week now since GitHub disabled our public repository for Quilibrium. Here's what we know.

1. After pushing through multiple connections to escalate the case, we finally received a support response after the first 48 hours:

We had an auto upgrade script that checked the repo and pulled the latest for updates. We weren't aware this was a violation, as there were many other repos existing for years without issue. That being said, we also have had over 200,000 nodes, so we might just be more popular than those.

Suffering from success, but the "following reports" line gives us pause this may have been a malicious report. There have been many times that the ticket had an updated time, but with no visible info. We replied to their message, and at this point, contacted their sales team to see if Premium support could expedite.

Three days later, we finally get another response, refusing to budge on this, and not being willing to reinstate the repo. The next day, I got a message from the sales rep to schedule a call.

Taking the call with the sales rep was very revelatory. This representative, who I will not name because I don't believe they deserve to be fired for telling the truth, said that there was a chance premium support would potentially get this unstuck, but it would cost "several tens of thousands of dollars a year".

They additionally revealed two other bizarre things going on with the state of GitHub: 1. Copilot is absolutely destroying their resources for non-enterprise customers, and they've been stretched too far to properly provide regular git services for non-enterprise users, causing several outages.

2. Management of support has told them in order to alleviate budget and time allocation, to stonewall anyone asking to have a repo restored after automated triggers disabled it. Further, reinstantiating the repo on a different name will subsequently also issue a full account ban.

Armed with this knowledge, we pushed harder against support yesterday, offering them two choices: 1. acknowledge this was a simple mistake and with no communication provided, they put themselves and us into a serious bind and specifically caused us significant harm, and restore the repo so we can resolve this.

2. Alternatively, if they will not, we gave them formal notice to preserve all records associated with this repository takedown, because if this was sourced from external reports, we have the right to know who caused us deliberate harm and will pursue this.

At this time, we have still received no response. Magically, github issues had an incident this morning. But that call with the sales rep revealed issues are serviced via Zendesk on the backend, so I went over to Zendesk to see if there were any active incidents. There were none.

We hope GitHub will be cooperative, but if not, we'll be very curious to see if the timestamps I captured on their ticket updates line up with internal notes they were asked to retain. Because if they don't, we'll have firm proof of destruction of evidence.