You can clone with
HTTPS or Subversion.
I'm trying to load the file provided here: http://spark-public.s3.amazonaws.com/sna/datasets/week1/LadaFacebookAnon.gml (sha256:6fd9c8dbb75415006af1db1c85664477771931c568cbd33e073e9597b3717deb)
I downloaded the recent Gephi from the website (0.8.2), and opened the file. Gephi doesn't seem to react, i.e. I don't receive an update to the UI, esp. no data is loaded or anything. In addition, and that's scary, it unlinks the file!
I didn't expect Gephi to delete files at all. And I expected it to open the file just fine. Python's networkx does.
I think this is related to opening files from temporal folder (in your case possibly by selecting open directly in the browser).
Please try to download it to any other folder and then open it.
It does indeed not delete the file when it is supposed to load it from my home directory.
FWIW: I did download the file to /tmp/ on purpose.
Needless to say that I consider it a serious issue that Gephi deletes my files. No matter whether the file is in a folder named /tmp/ or not.
Yeah, I agree this is an issue and keep it open until solved.
I can confirm this happening too. I'll try to fix it though.
I experienced this issue too, and it's really annoying: I'm generating temporary data from some scripts, thus working in my /tmp, I just spent half an hour trying to figure out WTF happened when some files suddenly disappeared..
The issue is still there (Gephi 0.8.2-beta). I would never expect a program to delete (temporary) files it hasn't created itself.
Trying to figure out a solution to this issue here, since it is very annoying for some people.
I could find 4 different points where Gephi deliberately deletes a file from the "java.io.tmpdir" folder:
1, 2, 3, 4
I'm not sure about the initial reason to this operation, but it seems related to the temporary extraction of compressed files. If my assumption is correct, could we get rid of these deletes and just rely on the deleteOnExit operation?