Make nothing but locking the file "D:\locations.txt"
I assume because of the try to get more File-Information like last modifier, size and so on..
My Suggestion:
In all Network Operation Opus should have a Timeout that should be adjustable from the setting-panel.
I above situation the file meta infos should be skipped by importing, and trying to receive on opening of that collection or even on mouseover or click on specific file.
Opus just asks the operating system for the information. It does not talk to the server directly at all, nor have any control over which timeouts the OS uses.
Opus could have timeouts for not answering operating system commands.
Similar to retry, skip, skip all, abort by copy operation.
In this situation Opus detect not responsible drives.
When i have a not responding network drive opened in one tab, than the whole lister become not responsible.
And i have to kill the opus at all, or open new lister and reopen all tabs manually again. It is not nice.
If i kill Opus, the tabs are not remembered like in firefox. Remembered is ONLY ONE lister not all, and when man exits opus without killing it.
The session should be written on modification not on exit.
I would at least be able to reopen all lister and all tabs after dopus crash or kill.
I checked all setting searching for "networks" in preferences.
Is there other setting that could help me?
We seem to be talking about two or three different things now.
There is no setting in Preferences because, as I said, Opus just asks Windows to talk to network drives (the same as it does for local drives, and the same as any other program does) and the timeouts and so on are all up to Windows, not Opus.
The most Opus can do is make requests on a separate thread and if the user asks to abort then the separate thread is "thrown away", still running and waiting for an answer but with nothing else waiting for it or caring what eventually comes back. (We do this already in some places.) The thread which makes the request will remain blocked until Windows decides to stop waiting. There is no way to cancel most filesystem requests; the thread which makes them simply has to keep waiting until a result or an error comes back.
Yes, a separate thread for each tab or OS-based-network operation would solve the most of the problems.
Would that be easily implemented? Could we expect that in following updates in near future?
Until then, do you know how/where to set the timeouts in the Windows 7 64bit Setting?
I will try to find something by myself, but any suggestion/direction would help me.