urllib2 - 403 that _should_ not occur.

[Philip Semanchuk]

On Jan 13, 2009, at 9:42 PM, ajaksu wrote:

> On Jan 13, 1:33 am, Philip Semanchuk <phi... at semanchuk.com> wrote:
>> I don't think I understand you clearly. Whether or not Google et al
>> whitelist the Python UA isn't a Python issue, is it?
>
> Hi, sorry for taking so long to reply :)
>
> I imagine it's something akin to Firefox's 'Report broken website':
> evangelism.
>
> IMHO, if the PSF *cough* Steve *cough*  or individual Python hackers
> can contact key sites (as Wikipedia, groups.google, etc.) the issue
> can be solved sooner.
>
> Instead of waiting for each whitelist maintainer's to find out we have
> a new UA, go out and tell them. A template for such requests could
> help those inside e.g. Google to bring the issue to the attention of
> the whitelist admins. The community has lots of connections that could
> be useful to pass the message along, if only 'led by the nose' to
> achieve that :)
>
> Hence, the suggestion to raise a bug.

Gotcha.

In this case I think there is no whitelist. I think Google has a  
default accept policy supplemented with a blacklist rather than a  
default ban policy mitigated by a whitelist. As evidence I submit the  
fact that my user agent of "funny fish" was accepted. In other words,  
Google has taken explicit steps to ban agents sending the default  
Python UA. Now, if the default UA changed in Python 3.0, maybe the  
best thing to do is keep quiet and maybe it will fly under the Google  
radar for a while. =)

Cheers
Philip