-
Notifications
You must be signed in to change notification settings - Fork 310
msglist: Open near first unread #1169
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
2ef3c40
to
f6b697f
Compare
Thanks @rajveermalviya for building this! For getting the sticky headers working properly in this new more-complicated setting, I have a draft branch from… oof, a year ago, it looks like. (I think things got hectic with the start of GSoC season, and I let it slip off my queue and never picked it back up again until now.) I just rebased that and did some quick manual testing, and I think it basically solves that problem. I'll spend some time fleshing that out, adding tests, and refreshing myself on exactly what it does and doesn't do. That draft branch doesn't do anything to MessageListView — it just sticks the latest 10 messages into the bottom sliver, as a way of making things interesting enough to exercise the sticky headers. I think we could merge that once I've polished it up, but it won't have any user-visible effect by itself; it just lays part of the foundation. Meanwhile I tried out this branch in some manual experimentation, and it seems like it basically works great except for the sticky headers. (In fact if I revert the bits here that cut out the sticky headers, those work great too… except in the screenful immediately above the sliver boundary.) So that's encouraging! I think this basically supplies what we need in the view-model, MessageListView. I did spot one bug: after some rapid scrolling around back and forth, both up and down, I found myself at the bottom of the list when there was definitely later history still to be found and it wasn't updating. The log looked like this:
So that seems like a bug in perhaps the backoff logic — we sent way too many requests, got rate-limited, and perhaps kept sending lots of requests after that. I looked at the code and compared (In case it helps with reproducing: I did that running it as a desktop Linux app, using the scroll wheel to scroll.) Anyway, if you feel like taking a break from the TeX work on another task, you might investigate and see if you can diagnose and resolve that behavior. You can leave the slivers just as they are in this revision, to give a scaffolding for experimenting with the view-model, and assume that the slivers part will be replaced later with the branch I'm working on. |
932ffbc
to
d45d8b3
Compare
d45d8b3
to
354e06c
Compare
Closing this in favor of #1517. Thanks @rajveermalviya for the experiment. |
Fixes: #80