Open GoogleCodeExporter opened 9 years ago
Another example is attached. The original text:
In a follow-up to their previous performance tests, long-running, independent
testing site AV-Comparatives.org loaded up anti-virus and malware prevention
suites with definition updates from February 2010, but slipped them malware and
other bad stuff that had been detected and defined since then. This was to test
the heuristic and analytic powers of these security suites—in other words, to
see if the software could figure out up-to-the-minute malware, even if their
providers haven't cataloged them yet in an update, or if the update hasn't hit
your computer yet.
As you move the slider around, the document gets out of sync with the window.
The database file and js file is attached.
Original comment by esperte...@gmail.com
on 9 Jun 2010 at 7:59
Attachments:
So annoying, but rather fixable. The problem is this: if a patch can be cut,
one of the patch options can be an empty string. When this occurs, we see the
problem above. For example, the first 5 tick marks are for different versions
of patch 1. The 6th tick mark has the original version of patch 1 and a
different version of patch 3. However, the results for this tick mark include
the smallest version of patch 1 (the empty string) instead of the original
version. The reason for this? When patch 1 becomes zero length, patch 2 "eats"
it. So when patch 1 is later expanded again, patch 2 encompasses the new text.
Then when patch 2 is written, it overwrites the new data (patch1 + patch2) to
be just patch2. In this way, patch1 is never again expanded.
Solution: I guess we could include some kind of place-holder indicating "This
patch can be cut". This could be "|", " ", or something else. This is not
ideal, obviously. Perhaps we could just make these things temporary, i.e.
existing only when the ShortnDialog window is open. Upon window close we could
remove them?
Original comment by dcrowell...@gmail.com
on 15 Jun 2010 at 7:19
Waiting on Issue 28 in order to fully test. The case needing to be tested:
[Patch1][Patch2][Patch3]
Patch1 and 2 are removed. Then, when Patch1 is replaced with text, not only
does Patch2's start point need to be moved, but Patch3's as well.
Original comment by dcrowell...@gmail.com
on 16 Jun 2010 at 8:16
Original issue reported on code.google.com by
dcrowell...@gmail.com
on 4 Jun 2010 at 10:18