On 19/06/2025 18.27, Rob Hallam wrote:
On Thu, 19 Jun 2025 at 19:33, Mark Filipak
<markfilipak.imdb-at-gmail....@ffmpeg.org> wrote:
I have a story. I'm corresponding with a fellow who wrote a popular subtitling
program. He wrote it
in 2000, rewrote it in 2013, and is currently rewriting it, again. He seems to
have little idea
regarding what what makes a good user interface, good. He doesn't think
user-interface thoughts. He
thinks coding thoughts. To me, user interface is design, not coding. Like all
good design, good UI
design takes experience and lots of time and lots of thinking (and compassion).
Hard agree.
I don't know what that means. Sorry. Did you leave any words out?
There's a huge number of people writing code. Millions. They're a
spectrum from meticulous planners with assiduous attention to
documentation to 'vibe coders' who let ChatGPT generate code and go
"hmm, looks about right". It sounds like you had experience with
people who were "code first, think later"; I'd like to think most
prefer the reverse, I know I do.
That's interesting. First, my friends in Silicon Valley were all professional programmers of the
first rank, working for companies of the first rank, for example: Seagate, Western Digital, Oracle,
Hewlett Packard, and Atari of course. They were not taught design in university. They were taught
coding. They were taught languages. They were taught a little about systems. They did not think
about their jobs as machine design. But machines are not just hardware. Even the most interactive
program is a machine. Second, when you're in a project, what percent of total time do you spend
thinking and sketching and what percent of total time do you spend coding? Do you feel any urgency
to get into implementation (coding) and showing something? Do you 'punch in' at the place you're
most confident about or do you start by mocking up and interconnecting what you think will become
the big chunks? I don't mean that as hard and fast rules; everybody sometimes does a little of
'this' and a little of 'that', especially when there's unknowns.
So, Rob, describe FFmpeg as a machine. Can you do it? What are the parts and their inputs and
outputs? How do they interconnect? I'm going to leave it at that and not make suggestions. I want to
see what you think.
I think that one of wisest things MPEG did was to create a reference decoder.
_______________________________________________
ffmpeg-user mailing list
ffmpeg-user@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-user
To unsubscribe, visit link above, or email
ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
