The author of the code can't know if it was infringed without seeing the source.
I am the author. And yes it was infringed. I can know with 100% confidence, without seeing claymore's full source.
How can I know? Because part of his source leaked in his binary. We can see the content of the assert statements, the name of the variables in these asserts, he name of the functions, the order, type, and number of arguments. Everything matches perfectly my blake.c implementation. And I wrote it myself, from scratch, with no inspiration from anyone else.
Now, I don't personally believe Claymore copied all of silentarmy. I think he just took blake.c which is only ~100 lines of code. That's why his miner is way different than mine.
The authors solver is slower, uses more VRAM, uses a lot more CPU, and PCIe bandwidth, runs on fewer cards, and crashes a lot. Why would you want to copy that?
Actually Silentarmy is faster (by a hair) than Claymore's
But I do use more CPU and more PCIe bandwidth :-(That said, I encourage everybody to CHILL OUT here. I didn't want this issue to be public in the first place (I messaged Claymore privately), but some started discussing here... Give Claymore time to respond. All he has to do is include a copyright notice in his miner, and he will be compliant with the MIT license. That's all.
For speed, vram usage and such, I was going with what was available when it was released. The current version is similar speed, but each miner has different speeds on various cards. (I don't know about vram and other factors because I don't want to fool with linux....). For code, I was thinking more of the birthday problem part, which is probably the main performance factor. I was just noting different characteristics to note that the main meat seems a lot different. Wanting acknowledgement for your work is fine, I was just speculating the issue probably isn't as big as people make it.