You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@clerezza.apache.org by Reto Bachmann-Gmür <re...@apache.org> on 2013/07/15 15:47:29 UTC
Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Hello
As some refactoring is needed to correctly support the default graph
I'm wondering what's the justification for having 2
SingleTdbDatasetTcProvider. I understand
ScalableSingleTdbDatasetTcProvider works better for large number of
graphs, but what are the advantages of the original one?
Cheers,
Reto
Re: Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Posted by Reto Bachmann-Gmür <re...@wymiwyg.com>.
Ok, done with CLEREZZA-807.
Cheers,
Reto
On Mon, Jul 29, 2013 at 7:26 AM, Rupert Westenthaler
<ru...@gmail.com> wrote:
> Hi Reto,
>
> I can not remember changing anything in the
> SingleTdbDatasetTcProvider. I have only fixed bugs in the
> IndexedMGraph (e.g. the comparator bug with bNodes you reported)
>
> best
> Rupert
>
> On Sun, Jul 28, 2013 at 5:28 PM, Reto Bachmann-Gmür <re...@wymiwyg.com> wrote:
>> Hi Rupert
>>
>> I agree re. the name. IIRC you recently fixed an issue in the current
>> SingleTdbDatasetTcProvider. Do you remember what this was about?
>>
>> Cheers,
>> Reto
>>
>> On Tue, Jul 23, 2013 at 8:45 AM, Rupert Westenthaler
>> <ru...@gmail.com> wrote:
>>> Hi all,
>>>
>>> I am fine with replacing the SingleTdbDatasetTcProvider with the more
>>> advanced ScalableSingleTdbDatasetTcProvider. However I would suggest
>>> to keep SingleTdbDatasetTcProvider as name for the component.
>>>
>>> best
>>> Rupert
>>>
>>> On Mon, Jul 15, 2013 at 8:41 PM, Reto Bachmann-Gmür <re...@wymiwyg.com> wrote:
>>>> On Mon, Jul 15, 2013 at 5:53 PM, Minto van der Sluis <mi...@xup.nl> wrote:
>>>>> Hi,
>>>>>
>>>>> The SingleTdbDatasetTcProvider probably performs a little better for a
>>>>> small number of graphs. Otherwise I see no real benefit.
>>>>
>>>> This seems to be a week reason for keeping those 817 lines of code.
>>>> Maybe @Ruperts knows more reasons?
>>>>
>>>> Cheers,
>>>> Reto
>>>
>>>
>>>
>>> --
>>> | Rupert Westenthaler rupert.westenthaler@gmail.com
>>> | Bodenlehenstraße 11 ++43-699-11108907
>>> | A-5500 Bischofshofen
>
>
>
> --
> | Rupert Westenthaler rupert.westenthaler@gmail.com
> | Bodenlehenstraße 11 ++43-699-11108907
> | A-5500 Bischofshofen
Re: Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Posted by Rupert Westenthaler <ru...@gmail.com>.
Hi Reto,
I can not remember changing anything in the
SingleTdbDatasetTcProvider. I have only fixed bugs in the
IndexedMGraph (e.g. the comparator bug with bNodes you reported)
best
Rupert
On Sun, Jul 28, 2013 at 5:28 PM, Reto Bachmann-Gmür <re...@wymiwyg.com> wrote:
> Hi Rupert
>
> I agree re. the name. IIRC you recently fixed an issue in the current
> SingleTdbDatasetTcProvider. Do you remember what this was about?
>
> Cheers,
> Reto
>
> On Tue, Jul 23, 2013 at 8:45 AM, Rupert Westenthaler
> <ru...@gmail.com> wrote:
>> Hi all,
>>
>> I am fine with replacing the SingleTdbDatasetTcProvider with the more
>> advanced ScalableSingleTdbDatasetTcProvider. However I would suggest
>> to keep SingleTdbDatasetTcProvider as name for the component.
>>
>> best
>> Rupert
>>
>> On Mon, Jul 15, 2013 at 8:41 PM, Reto Bachmann-Gmür <re...@wymiwyg.com> wrote:
>>> On Mon, Jul 15, 2013 at 5:53 PM, Minto van der Sluis <mi...@xup.nl> wrote:
>>>> Hi,
>>>>
>>>> The SingleTdbDatasetTcProvider probably performs a little better for a
>>>> small number of graphs. Otherwise I see no real benefit.
>>>
>>> This seems to be a week reason for keeping those 817 lines of code.
>>> Maybe @Ruperts knows more reasons?
>>>
>>> Cheers,
>>> Reto
>>
>>
>>
>> --
>> | Rupert Westenthaler rupert.westenthaler@gmail.com
>> | Bodenlehenstraße 11 ++43-699-11108907
>> | A-5500 Bischofshofen
--
| Rupert Westenthaler rupert.westenthaler@gmail.com
| Bodenlehenstraße 11 ++43-699-11108907
| A-5500 Bischofshofen
Re: Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Posted by Reto Bachmann-Gmür <re...@wymiwyg.com>.
Hi Rupert
I agree re. the name. IIRC you recently fixed an issue in the current
SingleTdbDatasetTcProvider. Do you remember what this was about?
Cheers,
Reto
On Tue, Jul 23, 2013 at 8:45 AM, Rupert Westenthaler
<ru...@gmail.com> wrote:
> Hi all,
>
> I am fine with replacing the SingleTdbDatasetTcProvider with the more
> advanced ScalableSingleTdbDatasetTcProvider. However I would suggest
> to keep SingleTdbDatasetTcProvider as name for the component.
>
> best
> Rupert
>
> On Mon, Jul 15, 2013 at 8:41 PM, Reto Bachmann-Gmür <re...@wymiwyg.com> wrote:
>> On Mon, Jul 15, 2013 at 5:53 PM, Minto van der Sluis <mi...@xup.nl> wrote:
>>> Hi,
>>>
>>> The SingleTdbDatasetTcProvider probably performs a little better for a
>>> small number of graphs. Otherwise I see no real benefit.
>>
>> This seems to be a week reason for keeping those 817 lines of code.
>> Maybe @Ruperts knows more reasons?
>>
>> Cheers,
>> Reto
>
>
>
> --
> | Rupert Westenthaler rupert.westenthaler@gmail.com
> | Bodenlehenstraße 11 ++43-699-11108907
> | A-5500 Bischofshofen
Re: Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Posted by Rupert Westenthaler <ru...@gmail.com>.
Hi all,
I am fine with replacing the SingleTdbDatasetTcProvider with the more
advanced ScalableSingleTdbDatasetTcProvider. However I would suggest
to keep SingleTdbDatasetTcProvider as name for the component.
best
Rupert
On Mon, Jul 15, 2013 at 8:41 PM, Reto Bachmann-Gmür <re...@wymiwyg.com> wrote:
> On Mon, Jul 15, 2013 at 5:53 PM, Minto van der Sluis <mi...@xup.nl> wrote:
>> Hi,
>>
>> The SingleTdbDatasetTcProvider probably performs a little better for a
>> small number of graphs. Otherwise I see no real benefit.
>
> This seems to be a week reason for keeping those 817 lines of code.
> Maybe @Ruperts knows more reasons?
>
> Cheers,
> Reto
--
| Rupert Westenthaler rupert.westenthaler@gmail.com
| Bodenlehenstraße 11 ++43-699-11108907
| A-5500 Bischofshofen
Re: Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Posted by Reto Bachmann-Gmür <re...@wymiwyg.com>.
On Mon, Jul 15, 2013 at 5:53 PM, Minto van der Sluis <mi...@xup.nl> wrote:
> Hi,
>
> The SingleTdbDatasetTcProvider probably performs a little better for a
> small number of graphs. Otherwise I see no real benefit.
This seems to be a week reason for keeping those 817 lines of code.
Maybe @Ruperts knows more reasons?
Cheers,
Reto
Re: Do we need Scalable and non-Scalable SingleTdbDatasetTcProvider?
Posted by Minto van der Sluis <mi...@xup.nl>.
Hi,
The SingleTdbDatasetTcProvider probably performs a little better for a
small number of graphs. Otherwise I see no real benefit.
ScalableSingleTdbDatasetTcProvider is/was my first contribution intended
to replace the original SingleTdbDatasetTcProvider. However replacing
from the start might have been a to big change, so I created sort of a
duplicate. If others are satisfied with the scalable version as well
then we could drop the old on. Please not that I still occasionally get
a concurrent modification exception, but I think the original has the
same flaw.
Also note that the Sesame based Provider (rdf.sesame.storage) has the
exact same scalability issues. When I have time I will redo the same
changes for that implementation.
Regards,
Minto
Op 15-7-2013 15:47, Reto Bachmann-Gmür schreef:
> Hello
>
> As some refactoring is needed to correctly support the default graph
> I'm wondering what's the justification for having 2
> SingleTdbDatasetTcProvider. I understand
> ScalableSingleTdbDatasetTcProvider works better for large number of
> graphs, but what are the advantages of the original one?
>
> Cheers,
> Reto
--
ir. ing. Minto van der Sluis
Software innovator / renovator
Xup BV
Mobiel: +31 (0) 626 014541