Error Checking/auto RE-searching
 

Issue 1: I have been trying to download large files for awhile (~700 MB) and they are consistently filled with errors. I know they are from the download process, because the locations change with each attempted download.

Now, that the Download ALL/ANY feature has been resolved. I suggest you implement a Download EXACT/FAST feature.

Download FAST: normal error checking (if any) currently used for downloading, I want it fast, fsat, fast.

Download EXACT: similar to the techniques employed by the EAC (Exact Audio Copy) program for cd ripping. Download the file twice and compare the results, if any discrepancies exist download that particular block again and check.

Issue 2: Also in my large file download ATTEMPTS (I highlight this word on purpose!!!), I get dropped by server all the time. I find by continuously researching for the file I get larger number of servers to download from and hence faster speeds.

I think Limewire should continuously search for my files using the implemented hash function.

In addition, this may be a bug or something, but the file I am downloading constantly starts over unless I babysit the program and continuously search for the file I am interested in downloading. It just makes the program less of a click-and-go and more of a watch-and-hope program.

Reliability is keeping me from buying the pro version. I am not going to shell out any cash for a program that can be here one day and gone the next.

I thank you for your time,
vossman

Running
Mac OS X 10.1.4
Limewire 2.4.1
dual 450 G4

researching...
 

I think automatic searchs should definetly be implemented even if it is set to an half hour or an hour until it researchs. I noticed a huge difference a long time ago when I switched from limewire/bearshare to phex, when I left each program alone over night to download files. With phex I was more willing to leave it on over night, that includes sharing files, with limewire and bearshare it was pointless for me to leave it on over night having to baby sit it and all.

Downloading a file twice means taking up twice the available bandwidth - not a good idea.

The upcoming hashing implementation will effectively build a checksum for each file - this will be sufficient to tell whether a download has been corrupted (although I imagine a corruption /fixing/ algorithm is a little way off).

Hashing will also reduce/eliminate download restarts. LimeWire should be able tell whether files are the same no matter what source or even if the file has been renamed (I hope).

I'd like LimeWire to be a little more aggressive in its searches as well, but the developers have to strike a balance - too much searching swamps the network.

Paying for the PRO version slightly increases the chances the LimeWire will be around tomorrow. The chances are pretty good anyway - it is open source so that if the developers disappear others can easily pick it up and carry it on.

Mark