That last point deserves underlining, so I'm just going to post a separate comment about it. Even though this particular link has been removed from this particu

That last point deserves underlining, so I'm just going to post a separate comment about it. Even though this particular link has been removed from this particular page, the first result on Google Scholar is still "Object Oriented Software Construction" and if you click on Google's helpful "all 95 versions" link, you will still find PDF versions of the document. I skipped the first one (which was an FTP url) but ran an HTTP HEAD request against the second PDF listed there (from stttelkom.ac.id) to determine that it was 16.6 MB long, presumably that is also the whole work. The same file was also available a few links below that from uettaxila.edu.pk and then on the next page from fs1.bib.tiera.ru . (There are other things labeled as PDFs with wildly-too-small file sizes to be a 1000-page book; I'm not sure what they are.) So saying that 'Yes' Google really cares seems too simple. They've never cared enough to, for example, store a database of hashes of forbidden files. Their efforts on YouTube are significantly more extensive, where any audio or video which remotely resembles some registered property can be taken down automatically.