Kang-Che Sung (宋岡哲)
Kang-Che Sung (宋岡哲)
@BenBE I intended this to be a meta-issue tracking all potentially unsafe integer downcasts. While you and @natoscott fixed the immediate `-Wshorten-64-to-32` warnings in macOS, there are also similar warnings...
> The use of `uint32_t` was intentional to have a minimum of size guarantees for the structure layout. I can't get it. What I see is the value would never...
I've split the changes of this PR into two commits, so that the part without dispute can be applied (cherry-picked) first. And I'm still not convinced with keeping the `uint32_t`...
> If you argue regarding the size of the member and the actually used range, than even `unsigned int` would be no better than the `uint32_t` it's using now. Actually,...
Is it just I that consider htop not cleaning the broadcast message a feature and not a bug?
@Fade78 Interesting. This means htop tried to update the display with all cursor positions wrong. But on the other hand this shows that `ncurses` is good at determining the "deltas"...
@natoscott There is no "utilization percentage" for NetworkIOMeter at the moment. So I would consider showing the number of interfaces a feature separate from this pull request.
@natoscott What was the ideal approach anyway?
@natoscott Look in my patch carefully. I separated the utilization value into two. `cached_utilisation_diff` would revert to original behavior of allowing more than 100% to be shown; the bar would...
> Regarding your data type question, I think I know what @BenBE is going to suggest: size_t FTW. :) For PCP this would require a downcast from `uint64_t` and an...