From: "Sean E. Russell" <ser@...>
Date: 2008-04-11T10:32:21+09:00
Subject: Re: REXML in 1.8.6 p114

--nextPart4034975.zXS2vhDo2j
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Thursday 27 March 2008, Michael Klishin wrote:
> In REXML in 1.8.6 p114 call site to
> REXML::Formatters::Transitive#initialize in REXML::Document#write uses
> old method signature (thus wrong arity) on line 188. Simple script to
=2E..
> Could someone shed some light on it?

Yes.  A patch fix supplied by a contributor was applied that introduced thi=
s=20
regression.  There was no unit test that would have caught this, so it ende=
d=20
up being propagated back to the Ruby trunk.

This has been fixed in the REXML repository, and a unit test has been=20
introduced to catch any future regressions (of this feature).  It hasn't ye=
t=20
made it into the Ruby repository, but will soon.

Of course, special dispensation will need to be gained for code freezes. =20

=2D-=20
### SER  =20
### Deutsch|Esperanto|Francaise|Linux|XML|Java|Ruby|Haskell|Aikido|Iaido
### http://www.ser1.net, http://www.ser1.net/Security/ser_public.gpg
### Jabber: seanerussell@gmail.com (OTR enabled)

--nextPart4034975.zXS2vhDo2j
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQBH/r+hP0KxygnleI8RAu4BAKCaaza1bxasq2Km3giDoAWjmJG4MwCfeFvJ
QrootgvOydG6CeibenifLAI=
=1Wrg
-----END PGP SIGNATURE-----

--nextPart4034975.zXS2vhDo2j--