So they used it as a manually synchronised dump from the same internal system & processes they had before it, and since that didn't work well and required extra

So they used it as a manually synchronised dump from the same internal system & processes they had before it, and since that didn't work well and required extra effort (surprise!) they're stopping again. I'm going back to managing my own rack, because logging in to rsync the new versions to S3 and then manually updating my Lambda definitions in the console is too much of a burden for the benefit running my app serverlessly in the cloud as well gives me.