AquaMail Forum
English - Android => Development builds => Topic started by: Kostya Vasilyev on May 27, 2016, 11:28:04 pm
-
https://www.aqua-mail.com/download/AquaMail-market-1.6.2.0-4.apk
---
+ Includes changes from 1.6.2.0-3:
http://www.aqua-mail.com/forum/index.php?topic=4711.0
+ An optimization for Gmail, scheduled mail checks:
The app is now able to determine when a folder / label is not changed, even when Gmail says it is.
When there are no changes in any folder / label, the app has been able to "do nothing" very quickly, for a long time now.
New case:
Something changes in a few folders (new / deleted message, read / unread / star), but not in others. All folders will be marked as changed (a uniquely Gmail thing).
Before:
Actual re-sync for every folder / label.
Now:
Actual re-sync for the changed folders / labels, but others are skipped.
Although this still takes more time than "skipping everything when nothing at all changed", it's still faster than re-syncing those (really unchanged) folders.
To test:
Refresh an account a few times until it goes really fast (no changes).
Change a message's unread/star state in one folder, or send yourself a test message, or delete one, etc. -- so that only some folders / labels are touched.
Refresh the account again.
You should see the unchanged folders take less time than the changed ones, although you will see the "refresh wheel" move through every folder (on the app's main screen).
Planning to update in Play on Sunday.
---
+ Включает изменения из 1.6.2.0-3:
http://www.aqua-mail.com/forum/index.php?topic=4711.0
Обновление Play планируется на воскресенье.
-
re: You should see the unchanged folders take less time than the changed ones, although you will see the "refresh wheel" move through every folder (on the app's main screen).
Yep - nice!
-
And less time means better (lower) battery use...
-
And less time means better (lower) battery use...
Yes - I believe I am seeing this although only a few days of testing with lower than normal volumes (holiday weekend in the U.S.).
Question specific to Gmail: When an item arrives via IMAP push are the remaining folders checked/synced? Not all obviously...just the ones that are set up for sync.
For every new item in Gmail at least two folders are touched: Inbox and All Mail. Given other weirdness I am wondering of Gmail flags the rest as modified too (specifically in the IMAP push scenario) which is what I believe your code is trying to mitigate.
The above is more of a curiosity question as I use a combination of IMAP push, timed pulls, event triggers and 'pull to refresh' on different devices. I would probably figure it out in time but hope you might know the answer.
-
Re: When an item arrives via IMAP push are the remaining folders checked/synced
Push works independently of scheduled sync, and is per-folder.
Re: combination of IMAP push, timed pulls
I recommend keeping "scheduled mail checks" enabled:
1 - it will get the folders which are not push enabled (and not all really need to be, usually just some)
2 - it will act as a watchdog for push
-
re: I recommend keeping "scheduled mail checks" enabled
Yep, that's what I typically do, either explicitly or with event based sync. On my primary mobile I find IMAP push is more efficient than hourly scheduled sync despite typically receiving multiple mails per hour at different times. I am twiddling some network settings to tune performance to my usage. None of f this represents a problem/issue - I enjoy optimization as time and interest permit.
As an aside, this morning I received an unexpected acknowledgement from the spouse that Aqua handles large conversations better than her clients. She is a die-hard Gmail user with primarily 1st party apps. Expanding a large conversation with Google properties can be an exercise in frustration, especially using a web browser. One tap in Aqua and all the contributors are immediately visible/accessible. Nice.
-
Re: Expanding a large conversation with Google properties can be a exercise in frustration, especially using a web browser
I believe they consider it a feature (the code uses terminology like "collapsed headers" and "the super-collapsed header").