Re: dev Digest 8 Feb 2013 22:46:15 -0000 Issue 317

[Lewis John Mcgibbney <le...@gmail.com>]

Hi Alfonso,

On Fri, Feb 8, 2013 at 2:46 PM, <de...@gora.apache.org> wrote:

>
>
> Hi all,
>
> I updated GORA-174 issue info about HBase backend at [0]. Any thoughts? I
> think now is better expressed.
>

This is much clearer for me at least. We are always going to have certain
problems (when developing Gora) when intricacies associated with (and which
affect all) datastores are encountered. GORA-174 is a perfect example.
There is no workaround and it is essential to have a thorough understanding
of the problem at individual "datastore level". Thanks for the
documentation, it is really driving this issue forward!


> If no one think is wrong, I will implement solution-1 and solution-2(this
> means maybe quite work, so do we maintain it? -I vote yes).
>

I think the proposed resolutions are certainly attractive and that we
should progress on this basis. When we get to a 1.0 Gora release (please
excuse my wishful long-term thinking) then we can act on completely
removing the deprecated methods from Gora, for the time being I see no
problem (and I certainly would back with my +1) methods being deprecated in
favour of more appropriate mechanisms for data persistence.

I've been talking this issue through with Renato offline and glad to
observe that the HBase and Cassandra stuff seems to be coming along nicely.

Is anyone in a position to address this with Accumulo?
What about DynamoDB?
Does DataFIle/AvroStore(s) support this in their current form?
Thanks
Lewis

Re: dev Digest 8 Feb 2013 22:46:15 -0000 Issue 317

[Renato Marroquín Mogrovejo <re...@gmail.com>]

Hi Lewis,


2013/2/8 Lewis John Mcgibbney <le...@gmail.com>:
> Hi Alfonso,
>
> On Fri, Feb 8, 2013 at 2:46 PM, <de...@gora.apache.org> wrote:
>
>>
>>
>> Hi all,
>>
>> I updated GORA-174 issue info about HBase backend at [0]. Any thoughts? I
>> think now is better expressed.
>>
>
> This is much clearer for me at least. We are always going to have certain
> problems (when developing Gora) when intricacies associated with (and which
> affect all) datastores are encountered. GORA-174 is a perfect example.
> There is no workaround and it is essential to have a thorough understanding
> of the problem at individual "datastore level". Thanks for the
> documentation, it is really driving this issue forward!

+1 (:

>> If no one think is wrong, I will implement solution-1 and solution-2(this
>> means maybe quite work, so do we maintain it? -I vote yes).
>>
>
> I think the proposed resolutions are certainly attractive and that we
> should progress on this basis. When we get to a 1.0 Gora release (please
> excuse my wishful long-term thinking) then we can act on completely
> removing the deprecated methods from Gora, for the time being I see no
> problem (and I certainly would back with my +1) methods being deprecated in
> favour of more appropriate mechanisms for data persistence.
>
> I've been talking this issue through with Renato offline and glad to
> observe that the HBase and Cassandra stuff seems to be coming along nicely.
>
> Is anyone in a position to address this with Accumulo?
> What about DynamoDB?
> Does DataFIle/AvroStore(s) support this in their current form?

We probably should think on fixing this issue inside the AvroStorage
first, but I think it would be wiser to also plan for upgrading Avro
to 1.7 at some point after this.


Renato M.

> Thanks
> Lewis