- User Since
- Apr 19 2018, 12:50 PM (92 w, 17 h)
Oct 31 2019
- It's flying fine
- Fixes from mlaurent review
Oct 29 2019
All your comments have been taken into account, branch cleaned, rebased on top of origin/master. And it has been dog-fed for more than a year… (so sorry for the delay)
- Further delay loading articles on fetch
- Introduce new Storage::articlesForCache
Laurent prefers it that way :)
Oct 28 2019
Hum, sure it is an optimization on a non-optimizing compiler, but since QModelIndex::row() is const and declared in the headers, thus inlined when called, we don't really get anything here (on the contrary, on a -O3 code with GCC 9.2 on amd64 we get two extra instructions and allocate more stack because the variable is kept).
Are you sure it is worth it?
Oct 26 2019
While writing the explanation for Laurent, I felt stupid for not writing
this patch that way the first time.
Mar 20 2019
Merged in master.
Mar 19 2019
I fixed for mlaurent.
I also fixed a new warning about the .css files that do not exist. Previously, they generated silent errors, but now that the baseUrl does not allow by default file:// urls, it appears in the program output.
- Drop reference to .css files that do not exist
- Output the xmlUri to help debug, as asked by mlaurent
Mar 14 2019
Mar 13 2019
Jun 10 2018
- Fix segfault when adding feed
May 26 2018
- Remove wrong comment
May 2 2018
Apr 30 2018
Nice catch. I'm wondering what was the use of this code 10 years ago.
Apr 24 2018
- Fixes following review
- Use the new FeedStorage::article() call
Apr 23 2018
Rebase on latest master.
With all these changes, akregator starts up almost instantly on my computer. RAM usage is far lower at startup (but sadly still grows once feeds are fetched, there are still ways for improvement here).
Also I suspect the guid 'cache' was a quite nice long term leak in akregator. I have no idea since how long that leak existed…
- Make sure all articles are not fetched uselessly at start/end
Apr 22 2018
Add missing fix
I merged this in https://cgit.kde.org/akregator.git/commit/?id=37f6e3a6f2ddbaba289f1ef7baf54ce71281798a, I did a small mistake in the git message and it didn't close this review, sorry about that.
@knauss yes I can, I am currently merging these 3 reviews.