Edwin :
Hi,
Any chance that you can work on a portable version as well?

Please see the TestFiles Applet function of ExactFile.

john doe :
Hi Brandon!
Have you ever compare speed of EF with other software?
Recently I made some extensive benchmarks and results show up that EF is terrribly slow.
1) In singlethread-mode it is 3-4 times (!) slower than competitors.
2) Multithread is even worse: 4 times slower than singlethread (!).
BTW: Multithread is bad idea, because HDD is bottleneck, not CPU.
Feel free to contact me for more details.
PS
Sorry for my English.

Multithreaded is faster than non-multithreaded in my own testing, but it will always depend on file sizes and hash methods. You should post your benchmarks on a blog somewhere so people can see what you are doing — just be sure your tests are repeatable.

If I have four 1 gigabyte files to hash with MD5, it takes 1/4 the time with four threads than it does with a single thread. HD is not the bottleneck in such a case. But if you switch to a computationally easy hash like Adler32, then that’s a different story.

The ONLY time I have seen multithreaded take *longer* than a single thread is on CD-ROM — and that’s because seek times is a bottleneck. That’s why Exactfile defaults to a single thread on CD media.

Finally, it’s pointless to use more hashing threads than you have CPU cores, so be sure you aren’t trying to hash with four threads on a single-core CPU. ExactFile defaults to using as many threads as you have cores.

Any chance that you can work on a portable version as well?

Have you ever compare speed of EF with other software?
Recently I made some extensive benchmarks and results show up that EF is terrribly slow.
1) In singlethread-mode it is 3-4 times (!) slower than competitors.
2) Multithread is even worse: 4 times slower than singlethread (!).
BTW: Multithread is bad idea, because HDD is bottleneck, not CPU.

Feel free to contact me for more details.

PS
Sorry for my English.

Brandon :

unagi :
Even though ExactFile has been out for a while I still like to use filecheckmd5 in certain scenarios. I’ve been using it with a new external drive lately and it seems like it gives me different results every time I run it. Sometimes “every file appears OK” then after a repeated run (with the same files) some files appear to be corrupted/changed. And every run seems to point out different files as being corrupted.

Strange. If you have any antivirus or shadow backup services running, try disabling those and see what happens.

Hi Brandon,

thanks for replying. I don’t have any backup services running but I turned off my antivirus and made a few test runs. Again, every test run claimed different files to be corrupted. It is really strange as I’ve been using filecheckmd5 for quite some time now and never had this happen before.

I know filecheckmd5 sometimes recognizes XLS-files as changed although they are not (that happened quite a few times) but this time the claimed corrupted files are of all sorts: PDF, TXT, XML, etc.

David Parvin :
I did not mean to pressure you into work on this, but the last comment or change I saw was over 6 months ago and I didn’t know what was going on.

Heh, no worries! I like working on ExactFile — but it doesn’t pay many bills so it’s pretty low on the priority list.

unagi :
Even though ExactFile has been out for a while I still like to use filecheckmd5 in certain scenarios. I’ve been using it with a new external drive lately and it seems like it gives me different results every time I run it. Sometimes “every file appears OK” then after a repeated run (with the same files) some files appear to be corrupted/changed. And every run seems to point out different files as being corrupted.

Strange. If you have any antivirus or shadow backup services running, try disabling those and see what happens.