Hi,
Me again I'm afraid!
I've now encountered some slightly different speed issues. Extracting using the v0.10 release from the downloads (or from NuGet) seems a lot slower than if I compile my own version using source downloaded from codeplex. (In both cases, using the exact same 'user' code calling into the library).
The version on the source code tab seems to be commit fcca5e95eafd, Apr 08, 2013 - so I'm guessing this is somewhat older than v0.10 for some reason? Maybe 0.81 guessing from the data in the download?
In any case, if I compile that version, I can extract a 11MB 7Z archive (containing 2 files) in ~7.5seconds. If I use the binary of v0.10 the same archive takes ~55s to extract (!)
I'm afraid I don't know what the cause is - since I can't see the latest version of source on Codeplex to download anyway? And am I missing any important fixes by using an earlier version? Obviously a 6x slowdown is a lot, but if the new version actually fixes some critical bugs, I don't want to be extracting corrupted data by staying on the old version...
Thanks
Me again I'm afraid!
I've now encountered some slightly different speed issues. Extracting using the v0.10 release from the downloads (or from NuGet) seems a lot slower than if I compile my own version using source downloaded from codeplex. (In both cases, using the exact same 'user' code calling into the library).
The version on the source code tab seems to be commit fcca5e95eafd, Apr 08, 2013 - so I'm guessing this is somewhat older than v0.10 for some reason? Maybe 0.81 guessing from the data in the download?
In any case, if I compile that version, I can extract a 11MB 7Z archive (containing 2 files) in ~7.5seconds. If I use the binary of v0.10 the same archive takes ~55s to extract (!)
I'm afraid I don't know what the cause is - since I can't see the latest version of source on Codeplex to download anyway? And am I missing any important fixes by using an earlier version? Obviously a 6x slowdown is a lot, but if the new version actually fixes some critical bugs, I don't want to be extracting corrupted data by staying on the old version...
Thanks