run some optimizations on #crisp image resize service, following the migration of all stored files to a new S3 backend, so that if cloudflare / cloudflare tunnel is down in AMS DC we can still serve images wordwide

Add diff viewer in #crisp helpdesk editor for upcoming #hugo feature

almost finish migrating all small #crisp S3 buckets, very large one (150M objects) planned to be migrated next Monday

migrate first #crisp user-facing S3 bucket from good old Minio to DigitalOcean Spaces

route some #crisp static assets to the new S3 powered by DigitalOcean Spaces, out of our legacy Minio setup — user uploads progressively coming next

finish last fixes on new #crisp s3 storage backend to support hybrid between our old school Minio over LAN, and DigitalOcean Spaces S3 buckets, and any other S3 provider — used to route certain path prefixes to certain S3 providers for future cost optimization and to ease with the production migration

deploy new #crisp file upload backend, invisible for users but now supports many different mapped external S3 bucket providers so that we can scale file uploads to infinity!

Review 3rd party plugins #crisp

Merge stuff, fix bugs #crisp

migrate #crisp to aws-s3 v3 library from v2

check #crisp old file server for amount of files stored and the count process is still running after 10 hours lol cause we have so many files maybe 100M+

investigate if it's possible to migrate all #crisp s3 objects from our current minio on network volume deployment, to digitalocean spaces because they have a limit of 100M objects and we may be well past that... lol - wrote a small C helper to count faster than Linux's ls command cause it's so slow when recursive on millions of files

Bump Vite on #crisp projects

finish simulating s3 migration for #crisp 14TB of user uploads to new system

finish removal of legacy #crisp ai training pipeline

#crisp accounting

merge 10+ merge requests on #crisp and feel head is 🤯

Improve UX during contacts import #crisp

decommission legacy #crisp analytics backend

store workspace identifiers in #crisp refund invoices so that we can post-process refund invoices for churn analytics