Richard Eckart de Castilho

I did an analysis. I pulled the latest changes from git to commit 5853f25a4146e8cc3331e6f4d7d79391021c3543 - and I added a log point in the deferred SHA256 calculation: ``` private final static...

I have built myself a locally modified version with a little cache for the SHA256 hashes. It expires hashes that are older than 30 minutes or if the file modification...

Well you can assume that there are like a couple of hundred Eclipse projects in the workspace, each of which is an m2e module and each of which includes the...

The set of dependencies of all of these projects is probably the ~1000 JAR files that end up in my cache (see comment above) - so if two of the...

This is the makeshift static cache I hacked myself together for the moment - not very well done, but apparently effective... ``` private final static AtomicInteger missCount = new AtomicInteger(0);...

Looking at the code, I note that there is a small bug - the `missCount` is actually not counting misses but lookups. The real miss count would be `hitCount -...

I guess deferring the calculation until it is needed could still save effort in case the value is never needed at all. I did notice a performance improvement when you...

I think instead of removing it, I might add counters that count calls to `addFile()` and counts to the supplier of the deferred value. The difference would be the calls...

So here some statistics - again, opening one of my `test.bndrun` files after a fresh start of Eclipse: * cache size: 1003 * times addFile was called: 52294 * times...

... I tried running Eclipse from the terminal and I do see some messages there, but not from the logger call I added to the cache code. So I used...