Jeff Standen
@jstanden.bsky.social
Founder & Software Architect @ cerb.ai
I've been obsessed with coding since the dial-up BBS days of the early 1990s. My daily stack is PHP/MySQL/AWS/Docker, with increasingly more Python for devops/ML.
#php #mysql #aws #python #docker #german #piano
I've been obsessed with coding since the dial-up BBS days of the early 1990s. My daily stack is PHP/MySQL/AWS/Docker, with increasingly more Python for devops/ML.
#php #mysql #aws #python #docker #german #piano
... the system tablespace could be rebuilt by creating the empty table schemas, using `discard tablespace`, dropping secondary keys, moving .ibd in, `import tablespace`, and adding keys back.
Always use SQL backups!
Always use SQL backups!
May 16, 2025 at 2:42 AM
... the system tablespace could be rebuilt by creating the empty table schemas, using `discard tablespace`, dropping secondary keys, moving .ibd in, `import tablespace`, and adding keys back.
Always use SQL backups!
Always use SQL backups!
I was doing something similar, but switched to Premiere Pro and .srt files.
It creates reliable speech-to-text transcripts, captions, and translations in the same UI. I'm still trying to automate it at the same quality.
That probably fits in your 30 min, but has the usual Adobe disclaimer.
It creates reliable speech-to-text transcripts, captions, and translations in the same UI. I'm still trying to automate it at the same quality.
That probably fits in your 30 min, but has the usual Adobe disclaimer.
April 30, 2025 at 3:41 PM
I was doing something similar, but switched to Premiere Pro and .srt files.
It creates reliable speech-to-text transcripts, captions, and translations in the same UI. I'm still trying to automate it at the same quality.
That probably fits in your 30 min, but has the usual Adobe disclaimer.
It creates reliable speech-to-text transcripts, captions, and translations in the same UI. I'm still trying to automate it at the same quality.
That probably fits in your 30 min, but has the usual Adobe disclaimer.
We have a large helpdesk client (120+ seats) who has been experimenting w/ AI drafts in a separate UI through our API. They could eventually be zero seats with the same high resource usage on our side.
I'm loathe to artificially lock features or charge by resolution. A change is needed though.
I'm loathe to artificially lock features or charge by resolution. A change is needed though.
April 30, 2025 at 3:24 PM
We have a large helpdesk client (120+ seats) who has been experimenting w/ AI drafts in a separate UI through our API. They could eventually be zero seats with the same high resource usage on our side.
I'm loathe to artificially lock features or charge by resolution. A change is needed though.
I'm loathe to artificially lock features or charge by resolution. A change is needed though.
Per-agent helpdesk pricing is definitely on the way out. It's never felt like a fair metric, and many apps reflect that w/ discount "guest" users who collaborate but don't answer tickets: devs, execs, sales.
We've always used floating "seats", but that also undercounts API/AI integrations.
We've always used floating "seats", but that also undercounts API/AI integrations.
April 30, 2025 at 3:17 PM
Per-agent helpdesk pricing is definitely on the way out. It's never felt like a fair metric, and many apps reflect that w/ discount "guest" users who collaborate but don't answer tickets: devs, execs, sales.
We've always used floating "seats", but that also undercounts API/AI integrations.
We've always used floating "seats", but that also undercounts API/AI integrations.
Cerb is another HelpScout alternative for your list. It's free for a single seat like yours, source available, and can be self-hosted in Docker (or cloud). We price by seats (simultaneous users) rather than usage. You'd recognize teams using it.
Disclaimer: I've been working on it for 20+ years.
Disclaimer: I've been working on it for 20+ years.
April 30, 2025 at 3:02 PM
Cerb is another HelpScout alternative for your list. It's free for a single seat like yours, source available, and can be self-hosted in Docker (or cloud). We price by seats (simultaneous users) rather than usage. You'd recognize teams using it.
Disclaimer: I've been working on it for 20+ years.
Disclaimer: I've been working on it for 20+ years.
It's a good rule. I've always admired your approach for maximizing enjoyment and minimizing burnout as an open source maintainer. Thanks for Redis, and continuing to share what you're learning. 👍
April 5, 2025 at 6:25 PM
It's a good rule. I've always admired your approach for maximizing enjoyment and minimizing burnout as an open source maintainer. Thanks for Redis, and continuing to share what you're learning. 👍
AI voice cloning from ElevenLabs (or even Kapwing) sounds a lot more natural. You can train them on 30-180 mins of audio. They both captured the nuances of my speech really well for video voiceovers. ElevenLabs' model is multilingual & they also let you design voices by prompt rather than training.
April 5, 2025 at 4:27 PM
AI voice cloning from ElevenLabs (or even Kapwing) sounds a lot more natural. You can train them on 30-180 mins of audio. They both captured the nuances of my speech really well for video voiceovers. ElevenLabs' model is multilingual & they also let you design voices by prompt rather than training.
Reposted by Jeff Standen
When I worked at an SV startup, we used to call an investment from SoftBank the kiss of death. SoftBank would pump so much cash into startups that it would completely kill that startup’s need to innovate or develop a financially sustainable business. 21/
January 27, 2025 at 2:12 PM
When I worked at an SV startup, we used to call an investment from SoftBank the kiss of death. SoftBank would pump so much cash into startups that it would completely kill that startup’s need to innovate or develop a financially sustainable business. 21/