Step through time. Each time point should load faster than normal.
Re-enable the disabled channel. The channel should load into its current timepoint.
What this PR does NOT do:
This is a relatively small change to address the "letter" of the linked issue, which falls just a bit short of the "spirit" of the issue for the following reasons:
Channels are initialized as enabled and only disabled after loading has begun, so this feature gives no benefit to the initial load of a volume. As a consequence, because the test viewer does not have a JSON volume that includes both multiple time points and enough channels to require more than one atlas image, I had trouble testing how the JSON atlas loader behaves when loading a subset of channels.
Though I spent a long time trying to clean up how channel enabled/disabled state gets updated to fix a bug, which turned out to be caused by something unrelated and silly, I didn't manage to find a change that I could reasonably earmark onto this PR. Until channel state is a bit more reliable, this feature may be a bit fiddly, though the point above is the only behavior I've seen that I'd describe as a "bug."
This PR doesn't include the optimization of only loading newly-enabled channels when possible. I'd rather address the subtleties that come with that optimization (described in the linked issue) when state updates are a bit more coherent. (Caching may make the benefits of this optimization minimal anyways.)
Addresses #158: don't load disabled channels.
To verify:
What this PR does NOT do:
This is a relatively small change to address the "letter" of the linked issue, which falls just a bit short of the "spirit" of the issue for the following reasons: