Android - major problem with speed using Amharic

[josh]

Hi Keith. I'm getting ready for a workshop so I handed this off to a colleague earlier in the week but he has been giving me updates. We suspect that there is a problem with memory not being deallocated with the Amharic characters and this is consistent with what you have found with the partial save. Unfortunately these kinds of memory issues are very hard to figure out so my colleague has been going through the code trying to find out where this is hapenning but has not been fix it yet. He will continue to look at it an we will keep you updated.

It is possible to launch a new data entry session using the function execpff. We usually this technique to have a main "menu" application launch different survey instruments. It may be possible to do a partial save (using the savepartial() function) and then relaunch the same application using this technique. I have never tried it but it should work. The one problem will be that there isn't a way that I know of to get it to automatically return to the last position. It will always ask you if you want to. Here is a link to an example of using one application to launch another: https://www.dropbox.com/s/g2n1phw0hl8x2 ... r.zip?dl=0

[Keith Tomlin]

Hi Josh

I've tried the following in one of my items. This saves and closes the open case, but it doesn't attempt to re-open that same case and ask if I want to go to the last position - it just leaves me in the data entry window with the list of cases (which I guess savepartial would do on its own). Not sure what I'm doing wrong. The path in the execpff is the folder where all the files are.

Thanks

Keith

PROC Q107
postproc
savepartial();
execPFF("E:\Dropbox\CSPro\HH\2015\150227-1418\IDEAS_HH.pff",stop)

[Keith Tomlin]

Oh, I see...I need to write the pff file using filewrite and include my ID items (cluster & household number) as parameters so that it knows what case to try to open...and I guess at that point it will ask if I want to go to the last position in the case...

[savy]

Keith,

Sorry about the problems with Amharic. We are actively working on this. Your observation is correct that the behavior gets worse over a period of time.
We will let you know once we have a fix for this.

Thanks

Savy

[savy]

We have a fix for this problem! We will post a new beta ASAP.

Thanks

Savy

[Keith Tomlin]

Fantastic!!! I can't wait to see it. Hopefully we can repeat the field testing with it early next week...

Many thanks

Keith

[Keith Tomlin]

Hi Savy and Josh

I just wanted to say a big Thank You for resolving this problem so quickly. We have carried out more field testing this week using the Amharic CAPI, and the speed of movement between the screens is very good now.

We are carrying out a further test of adding 100 cases to one tablet to see if this has any effect upon overall speed, and I will let you know how that goes.

Thanks again

Keith Tomlin
London School of Hygiene & Tropical Medicine

[Guest]

Good morning:
I had a similar problem and found that the slowness was the definition of number of records that occurs when creating the dictionary.
I had given him many records in population 700, then reduced to 25 and the application was accelerated.
I hope you serve.
Best regards,
Maria.

[Maria]

Good morning:
I had a similar problem and found that the slowness was the definition of number of records that occurs when creating the dictionary.
I had given him many records in population 700, then reduced to 25 and the application was accelerated.
I hope you serve.
Best regards,
Maria.