This is all guess work as an observer with little knowledge of the theory or practice of Freenet code.
I'm not particularly surprised that files from freenet sites would turn up on one's computer. Unless the freenet code contained it's own browser this seems an almost certainty.
I'm using a Windows 2K system with Earthlink.net access. The browser provided by the latter largely calls segments of regular Window browser for it's work. Windows is noted for putting convenience (so long as that is not a problem to MS) above security until someone with a very broad-band voice complains. But without freenet providing it's own "secure" browser, one would suspect that most of the other browsers would leave local tell-tales. Surely many have noticed that a second call to the same site (which has given up) will almost always load some of the graphics (and text even quicker) much faster the second time. This implies they are laying around ready to be accessed again by the browser.
Recently I was hunting something and found that the search function of Windows is quite capable of pointing to items in the cache (even though it looks quite meaningless if one opens such a folder directly). And it appears that one can copy that file found by search into another location. I'm sure those who regularly try to extract material from confiscated (or even sold) disks know this. That is, it appears that Windows provides the code to break the packing, but that is expected since the browser (or other functions) need to examine items that have been downloaded to the local machine.
And of course, if you have a browser that's not broken (as is my Windows currently) one can extract HTML source code (Netscape and Mozilla can do this just fine) so one too should be able to transfer graphic files from Internet folders even though they look cryptic.
The cheap solution (not in time) might well be to keep an inventory of what is being downloaded by Freenet and then promptly delete those files as soon as the Browser has finished it's work, but does Freenet currently know the final name or location of these files before it sends them to the Browser? Does it have access to that information immediately after the Browser finishes? What happens if on closes the Browser without returning to Freenet GUI?
--- Nicholas Sturm
