Ian's questions are interesting. Before writing http://catalogo.pusc.it/mobile/ I had a Q&A with... myself: Q: An app? A: Very beautiful appearance. It can easily store preferences, credentials, data and offline functionalities. It requires a lot of experience for developing and testing on many present -and future- smartphones and tablets. And certifications of Apple Store and Android Market. Apps increase in number day by day. Maybe this model will have an end is few years. My university has/will have its own app. There will be two apps for our students? Or will I develop part of the institutional app? Q: Ad hoc CSS to change Koha appearance? A: Seems difficult without changing templates. Web pages for smartphones must be very simple, to avoid zooming too often and allow easy finger taps. Q: A web app? A: Doesn't require Apple Store or Android Market certifications. Every mobile device has a (good) browser. Using jQuery Mobile, basically only HTML(5) and javascript skills are required. No way for integration with other apps. Login is required for user features (maybe cookies can help). After analyzing these pros and cons, I choose to write a web app. However, Koha runs in very different institutions. I think the community could try to develop both a web app and an app. Maybe their core could be the same, to share code as much as possible. Stefano Ian Walls wrote:
Nucsoft OSS Labs mentioned at KohaCon 2012 that they're commissioned to write a Mobile OPAC for one of their customers. Hopefully one of their team can provide some more details on what this will entail.
BibLibre also indicated they were working on a mobile CSS. Two different ByWater partners have also expressed interest in making this happen.
I think what needs to happen next is the community needs to get together and come up with a list of features and design styles that should make up a Mobile OPAC, and put together a implementation plan on which to do first. If we have a unified specification out there, all these different parties can work together towards the same goal, rather than having half a dozen different implementations floating around out there.
Some questions to ask:
- CSS-based, or a new template? - What features first? Search and biblio details seem obvious. What about user logged-in features? - Specific features for mobiles, like integration with GPS, compass, accelerometer and camera? This may require OS-specific apps, but would make the model tool that much more appealing
Cheers,
-Ian