[Go-essp-tech] [esg-node-dev] Re: Inconsistent search results from PCMDI9

[Sébastien Denvil]

Hi Luca

Le 25/01/2012 15:35, Cinquini, Luca (3880) a écrit :
> Hi Sebastian,
> 	thanks for testing the system - we really want to do this as much as we can before we open it up to external users...
>
> Now, what you found is not unexpected, and it originates from the fact that our unit of search is the dataset (aka the THREDDS catalog), not the file... So when you make a selection for "variable=pr", you find the dataset that contains, among other things, that file and variable.

Yep.

> When you place the dataset in the cart, you place it there as a whole - i.e. with all the files it contains, which you can then deselect if you wish.
>

With the current pcmdi9 deployment it seems you don't place the dataset 
in cart as a whole (only ten files appears per dataset and unluckily not 
the one I'm interested in). I can't figure out how to expose all the 
files in each dataset. May be the real bug (due to a hardcoded 10 files 
limit?) is here.

> That said, some scientists at PCMDI have expressed the same usability concern, and in fact the web UI used, till a few days ago, have an option called "Filter search constraints down to Files" which, if selected, would allow you to expose in the data cart only the file that contained variable=pr. This option was temporarily taken out for two reasons:
>
> a) it was buggy :)
> b) it was kind of confusing
>
>  From this email it sounds like you would advocate placing the option back in, and fixing the bugs ?

Yes I would advocate so. As it's a subjective Human Machine Interface 
question having both options in place seems to be the best solution.

Thanks.
Sébastien


> thanks, Luca
>
>
> On Jan 25, 2012, at 4:33 AM, Sébastien Denvil wrote:
>
>> Hi all,
>>
>> making some testing on pcmdi9.llnl.gov we encountered a worrying bug. We
>> used chrome 16.0.912.77 and firefox 9+.
>>
>> Please see the attached picture and the corresponding wget (from WGET
>> All Selected).
>>
>> On the search screen we made such a selection:
>> (x) time_frequency:mon
>> (x) experiment_family:Paleo
>> (x) experiment:midHolocene
>> (x) variable:pr
>>
>> 1) Search results
>> As a result we ended up with datasets but (once expanded) a wrong
>> selection of files and variables.
>> Instead of getting just pr (precipitation flux) as expected we ended up
>> with selection of several.
>>
>> For IPSL-CM5-LR the result should have been only one file:
>> http://vesg.ipsl.fr/thredds/fileServer/esg_dataroot/CMIP5/output1/IPSL/IPSL-CM5A-LR/midHolocene/mon/atmos/Amon/r1i1p1/v20111119/pr/pr_Amon_IPSL-CM5A-LR_midHolocene_r1i1p1_230101-280012.nc
>>
>> Instead we got 10 files (like all the other model), but no file
>> containing the precipitation flux. We know the precipitation flux is
>> available for all the resulting datasets.
>>
>> Regards.
>> Sébastien
>>
>> -- 
>> Sébastien Denvil
>> IPSL, Pôle de modélisation du climat
>> UPMC, Case 101, 4 place Jussieu,
>> 75252 Paris Cedex 5
>>
>> Tour 45-55 2ème étage Bureau 209
>> Tel: 33 1 44 27 21 10
>> Fax: 33 1 44 27 39 02
>>
>> <Capture-ESGF Portal - Google Chrome.png><corresponding_wget.sh><smime.p7s>_______________________________________________
>> GO-ESSP-TECH mailing list
>> GO-ESSP-TECH at ucar.edu
>> http://mailman.ucar.edu/mailman/listinfo/go-essp-tech
>


-- 
Sébastien Denvil
IPSL, Pôle de modélisation du climat
UPMC, Case 101, 4 place Jussieu,
75252 Paris Cedex 5

Tour 45-55 2ème étage Bureau 209
Tel: 33 1 44 27 21 10
Fax: 33 1 44 27 39 02


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4172 bytes
Desc: S/MIME Cryptographic Signature
Url : http://mailman.ucar.edu/pipermail/go-essp-tech/attachments/20120125/b849bbfd/attachment-0001.bin