Add explanation of vector holes/in-operator changes to NEWS [skip ci]

This commit is contained in:
Jon Siwek 2021-04-12 18:03:53 -07:00
parent 0c93e2fcb8
commit 361c99cc26
3 changed files with 18 additions and 1 deletions

View file

@ -1,4 +1,8 @@
4.1.0-dev.501 | 2021-04-12 18:03:53 -0700
* Add explanation of vector holes/in-operator changes to NEWS (Jon Siwek, Corelight)
4.1.0-dev.500 | 2021-04-09 12:18:06 -0700 4.1.0-dev.500 | 2021-04-09 12:18:06 -0700
* Reference the manual's explanation of originator and responder from conn_id (Christian Kreibich, Corelight) * Reference the manual's explanation of originator and responder from conn_id (Christian Kreibich, Corelight)

13
NEWS
View file

@ -52,6 +52,19 @@ New Functionality
Changed Functionality Changed Functionality
--------------------- ---------------------
- The ``in`` operator does not any more track whether a given vector element
has been assigned too. It now simply equates to "is this index between zero
and the size-of-the-vector-minus-one, inclusive". Previously, the ``in``
operator could have detected whether, by purpose or accident, a "hole" was
created in a vector by way of assigning to an index that's more than 1 beyond
the end of the vector (with intermediate elements becoming "holes") and
whether a "hole" had ever been subsequently filled by an assignment.
The behavior of creating a "hole" in a vector was never explicitly
documented, and should be thought of as undefined behavior and not something
the ``in`` operator can be used to detect. For this kind of usage, a
``table[count] of X`` is more appropriate than a vector.
Removed Functionality Removed Functionality
--------------------- ---------------------

View file

@ -1 +1 @@
4.1.0-dev.500 4.1.0-dev.501