On Friday, 30 January 2026 at 03:36:32 UTC, Ki wrote:
On Friday, 23 January 2026 at 14:51:26 UTC, Mike Parker wrote:
The D Language Foundation’s July 2025 monthly meeting took ...
I have a few questions:
1. This summary is not written by hand, is it? If the meetings
are recorded, an LLM could be used to generate a draft summary
from the audio or video. A second LLM call could then flag
potential inconsistencies or gaps for manual review, helping
streamline the workflow.
I record every meeting, then produce a transcript using Davinci
Resolve's AI generator. It includes speaker names and time codes,
and I can use it inside Resolve to easily jump to specific points
in the timeline to clarify whatever is garbled.
Until recently, I would then take that transcript and draft a
summary by hand into the the form you see in the posts. I'm not
going to publish raw transcripts for several reasons.
I experimented a few times with feeding ChatGPT examples of past
summaries, then giving it a transcript to generate a first draft
that I could then edit. It required too much revision to be
worthwhile. On my most recent attempt, I found that it's much
improved. Now it can produce a first draft that isn't garbage. I
still have to do extensive revisions to get it into the shape I
want, but I'm able to keep a significant portion of it. This is
saving me a nice chunk of time and will help me get caught up to
where I want to be, which is publishing every month the previous
meeting's summary just before the next one.
2. I think this would be a great blog post series. What do you
think?
Maybe name it like this:
- D under the hood: [title]
- DFL (D Language Foundation) monthly: Behind the Scenes
- Inside D: Monthly Foundation Updates
If I had infinite time, I might consider it, but it's not
realistic. And anyway, that seems redundant to me since they're
posted here on the forums and in this git repository:
https://github.com/dlang/meetings
Once I get caught up, I'm going to go back and add the older ones
that aren't there.
---
Thanks for sharing this! I really enjoy reading summaries like
this from time to time - they are both interesting and
insightful, and shows the complex issues the language
maintainers and developers face. I think it would be great to
give it its own space.
I'm glad you enjoy reading them.
