GEF5 Moving anchorage of anchored Connections performances [message #1798541] |
Tue, 20 November 2018 09:45  |
Patrick Muscat Messages: 14 Registered: May 2018 |
Junior Member |
|
|
Hello,
After investigating for performance problem with moving anchorage of several connections, we noticed that the checkCursor of ListListenerHelperEx in the bundle org.eclipse.gef.common was the sticky point.
We are using this version: Eclipse-SourceReferences: scm:git:https://github.com/eclipse/gef.git;path="org.eclipse.gef.common";commitId=51a73027b7d9c631d94e98b0ac97ff7a1e83f51e
We simply removed it and all its calls and did not noticed any error until now. Is it a good cure or is there any reason of this check we shoudn't ignore ?
Title [/Edit]
[Updated on: Tue, 20 November 2018 10:16] Report message to a moderator
|
|
|
Re: GEF5 Moving anchorage of anchored Connections performances [message #1802539 is a reply to message #1798541] |
Mon, 11 February 2019 16:34   |
|
Hi Patrick,
thank you very much for sharing!
As far as I can see, it should be safe to remove the checks. Only in the case when a change is not correctly iterated, then you would get an exception if the checks are enabled.
I will investigate w.r.t. the performance implications. However, due to the anchor architecture, I believe in many cases, this change will not be sufficient to resolve performance issues. Nonetheless, I will discuss this change for GEF with a colleague and experiment with it, too.
Best regards,
Matthias
|
|
|
|
|
|
|
Powered by
FUDForum. Page generated in 0.01957 seconds