New simpler-dup-finder

Short version:
There's a new experimental page up at
Play with it and tell me what you think.

Long version:
This new page is somewhat like the current simple-dup-finder: the page
layout is similar and when it finds a backtrace in the data you
supply, it searches for other bugs with a similar stacktrace (and both
use the same methods for searching in this case so the results will be
identical).  It's different in that if it doesn't find a stacktrace,
it does a natural language search to check for potential dupes
(employing MySQL's fulltext search capabilities).

The natural language search produces fairly different results from the
stack trace search.  While the latter can be thought of as a method
for just finding potential duplicates, the former is perhaps best
thought of as a way to find the 10 most closely related bugs
word-wise.  This means that the bugs it finds are somewhat less likely
to be duplicates, though in my testing so far I have still find it
useful (especially when I already recognize the bug and know its a
duplicate).  The ability to find closely related non-dupes seems
interesting too.

There's also a simplified version of the page with just the textbox at
which I'm kind of thinking of as a "Check if your issue is already
filed" kind of form.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]