Hi,

I've only just now uploaded it - it might take a few hours to reach all mirrors. In the meantime it's available from github:

git clone git://github.com/ingydotnet/inline-pm.git   Inline

Only difference between 0.45_01 and 0.45_02 is that untainting when running the Inline script with -T now works as documented. With earlier versions of Inline, you simply couldn't run an Inline script under -T, despite what the documentation claimed. (Some would argue that was a good thing.)

Because this version of Inline untaints %ENV and the Inline objects blindly, it issues (and documents) warnings about the wisdom (or, rather, the lack of wisdom) of making use of this capability. Nevertheless, for those who want to risk hanging themselves (or maybe even others), Inline now offers sufficient rope. (The warnings can even be silenced by setting the config option NO_UNTAINT_WARN=>1.)

Patrick provided practically all of the patches that fixed the problems re "untainting" with earlier versions of Inline, but there was one small section of code (provided by Patrick) that I couldn't see the need for. That section can still be found in Inline.pm, though it has been commented out. Search for "# The following small section of code seems".

If anyone can provide a reasonable argument in support of that code being included, I'll be most happy to oblige. It didn't actually break anything for me, but I baulked at the warning about untainting being issued for the running of *every* test script.

I think (hope) I thanked Patrick privately on his work in fixing the "untainting". I'd like to take this opportunity to thank him publicly for doing that. I've just realised that there's no recognition of this contribution of his in the 0.45_02 release. (I'll add an acknowledgement in the Changes file for 0.46 - which will be what 0.45_02 becomes ... all being well.)

As always, thoughts are welcome.

Cheers,
Rob

Reply via email to