Do you need urlgrabber?

[Tomas Radej]

On 09/22/2014 05:16 PM, Anders Blomdell wrote:
> On 2014-09-22 16:11, Tomas Radej wrote:
>> Hi,
>>
>> Pykickstart uses urlgrabber on multiple occasions to download URLs
>> either into memory or as files in filesystems. I would like to know
>> if you need this particular library, or if it can be replaced by e.
>> g. pycurl or even urllib. The reason is that the packaging team is
>> planning to drop support (and probably even ownership) of
>> python-urlgrabber as of Fedora 22.
> Well, I kind of like the mirroring support in urlgrabber, but besides
> the pain to rip it out from where I use it, I could live without it :-(

Well, I expect someone to take it and keep it alive, so it's not like 
you'd have to stop using it altogether, but from pykickstart's and other 
important components' perspective, it's not ideal to keep it included.

>> I am invested in this matter due to my participation in the 'Python 3
>> as default' initiative [1], and pykickstart's Python 3 compatibility
>> [2], which is better achieved without urlgrabber than with it, is a
>> part of it.
> Does better mean simpler in this case?

Exactly.

If there are no objections, I'll make a patch that removes the usage of 
urlgrabber and replaces it with whatever is suitable and supported, and 
post it to this list.

Cheers, TR