[Cosmo-dev] Bug 10304 and base64 encoding

Travis Vachon travis at osafoundation.org
Mon Sep 10 11:43:38 PDT 2007


To be clear, when I said "-0 for doing this for 0.7.2" I meant  
Bobby's proposal.

Sorry about the convolutions!

-Travis

On Sep 10, 2007, at 11:41 AM, Travis Vachon wrote:

>
> On Sep 10, 2007, at 11:28 AM, Bobby Rullo wrote:
>
>> The reality is that this there is no client who can do much of  
>> anything with EIM+JSON except for the WebUI, and the primary  
>> (only) consumer of it is the web browser. The EIM+JSON feed would   
>> never have existed if not for the web UI. So the needs of the web  
>> UI are of primary importance here.
>>
>> Now, Cosmo as an open service is important as well. To this end,  
>> complying with the standard is important and therefore so base64  
>> encoding the Atom content.
>>
>> Travis suggested in IRC having base64 encoding on by default, but  
>> having the ability to turn it off via a header or query string.  
>> This solution would seem to me to satisfy everyone's needs.
>>
>
> Actually, for 0.7.2 I'd rather just revert changes instead of  
> making more, though I'm only -0 on doing this for 0.7.2 if other  
> folks are behind it. I think Bobby's proposal is worth considering  
> for trunk, though I'd rather find a solution that works more in the  
> spirit of the way we're doing things now (Atom + Google style  
> projections). Perhaps some way of specifying the <content> element  
> should contain "text/eim+json" vs "application/eim+json" should be  
> considered for 0.8.
>
> -Travis
>
>> bobby
>>
>>
>> On Sep 10, 2007, at 11:15 AM, Brian Moseley wrote:
>>
>>> On 9/10/07, Travis Vachon <travis at osafoundation.org> wrote:
>>>
>>>> Unfortunately, I'm -1 on leaving the code as-is in 0.7.2. I'm +0 on
>>>> your proposal, but am still confused why any of these changes  
>>>> belong
>>>> in this particular release. To be clear, I would be perfectly  
>>>> content
>>>> moving all of these changes (base64 encoding included!) into trunk.
>>>
>>> one could argue that .7 is broken in this regard. what if somebody
>>> needed to build a client and the only library available to them
>>> automatically attempted to base64 decode our entry content  
>>> because of
>>> its mime type?
>>> _______________________________________________
>>> cosmo-dev mailing list
>>> cosmo-dev at lists.osafoundation.org
>>> http://lists.osafoundation.org/mailman/listinfo/cosmo-dev
>>
>> _______________________________________________
>> cosmo-dev mailing list
>> cosmo-dev at lists.osafoundation.org
>> http://lists.osafoundation.org/mailman/listinfo/cosmo-dev
>
> _______________________________________________
> cosmo-dev mailing list
> cosmo-dev at lists.osafoundation.org
> http://lists.osafoundation.org/mailman/listinfo/cosmo-dev



More information about the cosmo-dev mailing list