Thread (5 messages) 5 messages, 4 authors, 2023-06-23

Re: [PATCH v4 1/4] gitformat-commit-graph: describe version 2 of BDAT

From: Derrick Stolee <hidden>
Date: 2023-06-20 13:23:59

On 6/13/2023 5:58 PM, Junio C Hamano wrote:
Jonathan Tan [off-list ref] writes:
quoted
+	in Probabilistic Verification". Version 1 bloom filters have a bug that appears
"bloom" -> "Bloom", probably, as the name comes from the name of its
inventor (just like we spell "Boolean", not "boolean").
The ultimate recognition comes from when the term named after you
becomes lower-case ("boolean" is sometimes in this category, but
definitely "abelian" is an example).

In this case, you are right that we should capitalize Bloom.
quoted
+	when char is signed and the repository has path names that have characters >=
+	0x80; Git supports reading and writing them, but this ability will be removed
+	in a future version of Git.
Makes sense.
I also like how you organized this: "We support version 2. 1 is
still around but not for long."
 
I wonder if we want to mention what the undesired misbehaviour the
"bug" causes and what we do to avoid getting affected by the bug
here.  If we can say something like "When querying for a pathname
with a byte with high-bit set, the buggy filter may produce false
negative, making the filter unusable, but asking for a pathname
without such a byte produces no false negatives (even though we may
get false positives).  When Git reads version 1 filter data, it
refrains from using it for processing paths with high-bit set to
avoid triggering the bug", then it would be ideal.  Or "When the
repository has even a single pathname with high-bit set anywhere in
its history, version 1 Bloom can give false negative when querying
any paths and becomes unusable.  You can use $THIS configuration
variable to disable use of Bloom filter data in such a case" would
also be fine.  The point is to give actionable piece of information
to the readers.
This is definitely helpful, but if someone is having issues we
would say "try version 2 and see if it still happens" and not
over-index on the underlying reason.

That's to say, I'm OK with the shorter description of the problem.
Feel free to expand if you're interested, though.

Thanks,
-Stolee
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help