[Linux only] CEF Upgrade to 2785 #626
Conversation
|
Brackets relevant PR adobe/brackets#13975. Both need to be merged. |
…ethip/LinuxCEFUpgrade2785
…ize to repaginate view port content 3) Change the minor version of CEF to 1486
|
@abhijitapte Could you please review this? |
appshell/cefclient_gtk.cc
Outdated
| } | ||
|
|
||
| int RunMain(int argc, char* argv[]) { | ||
| int RunMain(int argc, char** argv) { |
There was a problem hiding this comment.
might not be really needed in this PR's context
There was a problem hiding this comment.
@abhijitapte I will revert this change.
|
Please keep track of 3rd point to pick that up in next sprint. Apart from that it looks good to me 👍 |
| void ClientHandler::OnLoadStart(CefRefPtr<CefBrowser> browser, | ||
| CefRefPtr<CefFrame> frame) { | ||
| CefRefPtr<CefFrame> frame | ||
| #ifdef OS_LINUX |
There was a problem hiding this comment.
is the function's signature like this for linux plaform alone?
There was a problem hiding this comment.
This is required as we have migrated to a newer version of CEF and the signature changed in the newer version.
There was a problem hiding this comment.
got you. this will have to be removed after we are on 2785 on all three platforms.
|
@navch The third option is no more required as the issue was fixed in a different way. |
|
@abhijitapte Addressed the review comments. Please have a look at the PR. |
* Force HiDPI to scale factor 1.0 for now on Linux * Linux only: Changes required for CEF 2785 Integration * 1) Removed force-device-scale-factor code\n 2) Introduced Delayed resize to repaginate view port content 3) Change the minor version of CEF to 1486 * Removing unused variable. * Address review comments. * Reverting the space change.
4 participants
CEF needs to be upgraded to a later version, as the current version of CEF has some issues with menus text getting corrupt. Also CEF needs to be patched wit IME as IME handling is missing in CEF.
The following are the note worthy things.
Global.jsForced the DPI scale factor to 1, as rendering is breaking upon launching. Surprisingly, upon resizing, the view starts to render properly. We need to revisit this in the next sprint.(This is not required anymore, as the rendering issue is now fixed with a different technique with a fake resize)