Hi @TSC, as mentioned at the last call, I'd like to give you a feedback from users that came out from three workshop on ODK I facilitated in Singapore and Indonesia.
Overall users are very happy with the new ODK Collect and especially with the form update feature, very useful when used together with csv preload using the username metadata field as ID for a specific enumerator that can get a specific list of choices filtering the data on the csv by his ID and then continue with a questionnaire "prepopulated" with specific info. I can go more on detail about CRVS and VA if someone is interested on a future call.
Of course a user when learn about a new feature look forward asking for an extension of the feature. So they asked if it would be possible to at least notify on the ODK collect when a totally new form is added to Aggregate. For new forms a notification only would be ok, then the enumerator woul if necessary download the new form.
Next is the appearance column that is very useful to adapt a form to a device (e.g. a smal screen phone would work better with no-calendar for dates or minimal for answer choices etc). Of course if I have big screen tablet I'd prefer other appearance. What about when on my project I use several kind of devices (tablets of diff screen size and phones)?
It would be useful to have the possibility on the settings to override the appearance of some widgets so you can define those on the xlsform and in the tablet someone can change this to something like "compact" or "normal (same as xlsform definition)" or "large".
I added this on the "Features" category Collect: make better use of larger screens.
Lot of people are looking forward to see the p2p Share app and the sms one. I already mentioned on the call of 30th May that if p2p can be on the intranet as well this would be faboulous.
Requests on GPS background feature is always coming from several users.
Some requests on Aggregate that I suppose will be addressed to Central are:
export to csv should leave as headers the variable names as defined in the csv and not have on it all the groups/sections part. This is a big headache for statisticians that need to clean up columns names before start analysis.
If I use many calculations on my form all of them will be part of the export. If in the future ODK Central or Briefcase would give the possibility to choose which variable to export this would be very useful.
Aggregate at the moment would support integration with Enketo, that many users see a very good added value to the system. Would Central support this as well? Maybe like with kobo, it would be interesting to ship central with Enketo (or think about a possibility to also provide web data entry to a form).
Briefcase is getting better and better. When it will support the scheduling features, then it would be perfect.
If any other feedback is arriving the following days weeks, I'll add it to this post.
Let me knwow if you have any thought or comment.
A feature about to be worked on is to extend form audit log to add a location column that is updated opportunistically if a “location-minimum-accuracy” or “location-update-frequency” parameter is set in the form. Is that the kind of background GPS feature that would satisfy the requests?
I'm very intrigued about this. I see a lot of users consuming Aggregate exports instead of using Briefcase exports, which are much better suited for forms with repeat groups, etc. I would love to know the rationale behind the decision of using Aggregate instead of Briefcase. Is it maybe the web API? What is that that Aggregate has and Briefcase lacks?
Based on the feedback from last weeks, I'm starting to think about backloged and new features that might be handy for these use cases:
Export of flat CSV files
Export media attachments as download URLs instead of writing the actual files into the output folder
If I use many calculations on my form all of them will be part of the export. If in the future ODK Central or Briefcase would give the possibility to choose which variable to export this would be very useful.
Hi @ggalmazor,
I'm not talking about repeat groups but grouping questions (begin-end group).
When you export (also in Briefcase) the name of the column is not simply the variable name, but it contains the group it belongs to in the name.
If you for example look at the attachment for me a variable name is Id10003, Id10004 etc, but as these variable are inside group presets in the xlsform then the column name get the name presets-Id10003, presets-Id10004.
And you can see long variable names when you have a complex form with group within groups.
Hope now I'm clearer.
I see now that those group prefixes have some tradeoffs. I think they would be always needed in case two groups have a field with the same name, though.