Conversation

Notices

1. Embed this notice
   ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Monday, 23-Jun-2025 13:27:31 JST ✧✦Catherine✦✧

   so why on earth were CNCF/K8s kicked off Slack's enterprise plan, anyway?

   cost? how much can it cost to store a bunch of chat messages for two moderately popular OSS groups, ten dollars a day? all of the capital cost has been paid at this point, it's just the marginal one at question, which cannot be high unless Slack's infrastructure is dramatically worse than any reason could allow

   the amount of goodwill they know they'll lose can't possibly compensate the cost

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Monday, 23-Jun-2025 13:28:14 JST ✧✦Catherine✦✧

     in reply to

     is it just that some suit is upset about an open source group getting nominally expensive services for free?

   • Embed this notice
     mcc (mcc@mastodon.social)'s status on Monday, 23-Jun-2025 13:46:37 JST mcc

     in reply to

     @whitequark Kubernetes is really significant. So much of the global economy depends on it. Ignoring all social angles to this, it would be absurd if this ultra-significant project, who partners with a whole host of major corporations, could not afford an enterprise plan for their chat server.

     Problem: This is also the precise group best equipped to self host a chat server, if they could find one

   • Embed this notice
     mcc (mcc@mastodon.social)'s status on Monday, 23-Jun-2025 13:49:15 JST mcc

     in reply to

     @whitequark 8 ball says yes

   • Embed this notice
     Wulfy—Speaker to the machines (n_dimension@infosec.exchange)'s status on Monday, 23-Jun-2025 13:59:16 JST Wulfy—Speaker to the machines

     in reply to

     • mcc

     @mcc @whitequark

     On the upside, the #devops cloud infrastructure is such that it's much easier to spin up iron...

     ...or to coin a word, "precipitation" (if it's in the cloud).

     I got my #VPS builds to about 30min...that's from the time you register an account, to when you have a fully functional system.

     Now, it takes longer if you have a load balanced stack and distributed database, and while I have no first hand knowledge of complex #cloud infrastructure, from my cursory investigation, the most time demanding sink would be moving your massive database.

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Monday, 23-Jun-2025 13:59:16 JST ✧✦Catherine✦✧

     in reply to

     • mcc
     • Wulfy—Speaker to the machines

     @n_dimension @mcc i don't know how large the CNCF message history is, but they tell everyone involved to back it up themselves

     this strongly suggests that it isn't any worse than "few gigabytes" range. which isn't going to cost a noticeable amount unless slack is somehow really bad at infrastructure

   • Embed this notice
     Nathan Sorenson (takeoutweight@mastodon.social)'s status on Monday, 23-Jun-2025 14:52:07 JST Nathan Sorenson

     in reply to

     @whitequark ya.... And maybe not even so personal as a suit being upset at free riders... I put odds on quarter end push for an ill-considered third KR deep in some excel spreadsheet where fiddling with who gets what plan shifted some numerator to get a row from "at risk" to "on track"

   • Embed this notice
     JP (froztbyte@mastodon.social)'s status on Monday, 23-Jun-2025 22:49:20 JST JP

     in reply to

     @whitequark looks like SFDC is clamping down across the board: the local tech slack (sub-5k users) also got one of these (and can confirm this isn't hearsay, have seen the mail myself)

     has a 5-figure (USD) "for your billable monthly active users" proposition to that org

     (org was, to date, on the nonprofit-freebie golden path)