|
|
|
| Authored by: Anonymous on Thursday, May 17 2012 @ 06:07 PM EDT |
How in the Hell --
Ugh. Were I Google "expert" (I do write my own lexers, parsers and
bytecode interpreters), I would have said this:
"This is a test." <- That sentence is like English byte code.
This is what Davlik does: "Find letters preceding a period and replace them
with "success". This would yield:
"This is a success."
When you do that, it's pattern matching. It's a specific search and replace
operation, nothing more. Even if you don't know English at all you could
perform the search and replace. You don't have to simulate English to find
letters before a period, and after a space, and replace them with a set of other
letters.
Now, I would refine this a bit and make it exactly apply to the byte code, and
show a few diffs of the before and after blobs of codes and the actual search
logic... "find this pattern" not evaluate the meaning of the bytecode
(a simulation).
It's ridiculously how the technical witnesses didn't actually describe the
mechanism at work (I know the mechanisms, I've seen the source, it's not
damaging to explain them).
Oh well. C'est la mort.[ Reply to This | Parent | # ]
|
|
|
|
|
