Hi Ian, Please see my inline comments. Thanks!
The original app was designed for Android 18 although there are now various SDK levels mentioned in various manifest files. Minimal testing on other API levels was done, pretty much just the developers phone. Targeting as low an SDK version as possible would be desirable. Obviously not way back to Android 2 but something which allows a reasonable percentage of devices to use the app (http://developer.android.com/about/dashboards/index.html). 4.1 (SDK 16) may be a good target. No problem, My office's Mobile dev team has got some variety of Android devices (from low end to high end ranges) which we use to QA/Test our app releases. I would make the app support for minimum SDK as possible and test it on different resolutions and platforms for an smoother release. The app uses various dependencies, listed in the build files. However, these dependencies may be out of date and in fact may be unused by the code. They need to be cleaned up. ok The original app used a modified version of the apache http libraries to talk to the taverna server code because the android sdk did not include an up to date version and the latest versions clashed with the android package namespaces. Android compatible versions of recent versions of the apache http libaries are now available (https://hc.apache.org/httpcomponents-client-4.3.x/android-port.html) and I tried to use them in my 'fork' although testing was minimal. alright The mygrid taverna server library (https://github.com/myGrid/t2-server-jar) is used to access the taverna server API. This library is not actively supported and in fact may be too heavyweight for the app. It is not even clear what branch is used. The taverna server API is just a REST based API over HTTP (http://dev.mygrid.org.uk/wiki/display/tav250/REST+API) so some refactoring is a good idea. It has to be able to handle talking to a taverna server over HTTPS with a username/password that is not the default taverna/taverna one. The HTTPS may have a user created certificate which is not from a CA. ok Some of the UI relies on the sliding menu library by Jeremy Feinsten which is included directly in the src rather than via gradle. The provenance of the code is unclear, possible from https://github.com/jfeinstein10/SlidingMenu. There may be a more modern Android way to do that kind of UI design but I am not convinced that the app needs it. ok The PullToRefresh library by Chris Baines (possibly https://github.com/chrisbanes/Android-PullToRefresh) is used via gradle but again there may be a modern Android way to do things. ok One of the important tasks is to ensure that the app can be supported beyond the project end and that means relying on 3rd party libs that are actively supported. ok Think of the current app as a demonstrator. I would be happy for the it to be developed from scratch. There is no need for it to even be an Android app, a web app would be just as good as long as the functionality is there. In fact, that is a good way to target multiple mobile platforms perhaps using some apache cordova (http://cordova.apache.org/) for any native functionality required. i have some exposure with cordova, but i think we wouldn't be getting that cool native android experience for the end user if we develop the app using cordova. so IMHO redesigning the Android app would be the way to go. is that ok? The basic app functionality required (not an exhaustive list): View workflows from http://www.myexperiment.org Log in to http://www.myexperiment.org and view your workflows Talk to required parts of taverna server API over HTTPS. Run a workflow. Requires upload of input data and the workflow to a taverna server instance. Monitor workflow run. View workflow run results. Download results. Also from JIRA - > Secondary requirements: Favourite workflows > Link to dropbox and save workflow results there. ok Thank you, Harsha On Fri, Mar 6, 2015 at 6:52 PM, Harsha Siriwardena <[email protected]> wrote: > Hi Ian, > > Thanks lot for the detailed information, I will go through this and let > you know if i need any clarifications. Thanks! > > Thank you, > Harsha > > On Fri, Mar 6, 2015 at 5:50 PM, Ian Dunlop <[email protected]> > wrote: > >> -----BEGIN PGP SIGNED MESSAGE----- >> Hash: SHA512 >> >> Hello, >> >> Here is some more background info on why the app needs updated and >> what I think is the main functionality required. >> >> The original app was designed for Android 18 although there are now >> various SDK levels mentioned in various manifest files. Minimal >> testing on other API levels was done, pretty much just the developers >> phone. Targeting as low an SDK version as possible would be desirable. >> Obviously not way back to Android 2 but something which allows a >> reasonable percentage of devices to use the app >> (http://developer.android.com/about/dashboards/index.html). 4.1 (SDK >> 16) may be a good target. >> The app uses various dependencies, listed in the build files. However, >> these dependencies may be out of date and in fact may be unused by the >> code. They need to be cleaned up. >> The original app used a modified version of the apache http libraries >> to talk to the taverna server code because the android sdk did not >> include an up to date version and the latest versions clashed with the >> android package namespaces. Android compatible versions of recent >> versions of the apache http libaries are now available >> (https://hc.apache.org/httpcomponents-client-4.3.x/android-port.html) >> and I tried to use them in my 'fork' although testing was minimal. >> The mygrid taverna server library >> (https://github.com/myGrid/t2-server-jar) is used to access the >> taverna server API. This library is not actively supported and in fact >> may be too heavyweight for the app. It is not even clear what branch >> is used. The taverna server API is just a REST based API over HTTP >> (http://dev.mygrid.org.uk/wiki/display/tav250/REST+API) so some >> refactoring is a good idea. It has to be able to handle talking to a >> taverna server over HTTPS with a username/password that is not the >> default taverna/taverna one. The HTTPS may have a user created >> certificate which is not from a CA. >> >> Some of the UI relies on the sliding menu library by Jeremy Feinsten >> which is included directly in the src rather than via gradle. The >> provenance of the code is unclear, possible from >> https://github.com/jfeinstein10/SlidingMenu. There may be a more >> modern Android way to do that kind of UI design but I am not convinced >> that the app needs it. >> The PullToRefresh library by Chris Baines (possibly >> https://github.com/chrisbanes/Android-PullToRefresh) is used via >> gradle but again there may be a modern Android way to do things. >> One of the important tasks is to ensure that the app can be supported >> beyond the project end and that means relying on 3rd party libs that >> are actively supported. >> >> Think of the current app as a demonstrator. I would be happy for the >> it to be developed from scratch. There is no need for it to even be an >> Android app, a web app would be just as good as long as the >> functionality is there. In fact, that is a good way to target multiple >> mobile platforms perhaps using some apache cordova >> (http://cordova.apache.org/) for any native functionality required. >> >> The basic app functionality required (not an exhaustive list): >> View workflows from http://www.myexperiment.org >> Log in to http://www.myexperiment.org and view your workflows >> Talk to required parts of taverna server API over HTTPS. >> Run a workflow. Requires upload of input data and the workflow to a >> taverna server instance. >> Monitor workflow run. >> View workflow run results. >> Download results. >> >> Cheers, >> >> Ian >> >> On 05/03/2015 11:41, Harsha Siriwardena wrote: >> > Hi everyone, >> > >> > My name is Harsha Siriwardena from Colombo, Sri Lanka - I am a >> > masters student pursuing Msc in Enterprise applications from >> > Sheffield Hallam university. I also work as a Mobile engineer >> > specialised in android development & altogether I had experience in >> > Android development over 2 years now. So I'm really interested on >> > Taverna Android App enhancement project idea and would like to work >> > on it this summer. >> > >> > Currently I have able to checkout the android source >> > <https://github.com/ianwdunlop/tavernamobile> code and successfully >> > ran the app in my android device. >> > >> > From: https://issues.apache.org/jira/browse/COMDEV-123 >> > >> > The code relies on some APIs from google and apache which have >> > moved on >> >> since then eg the https libraries which needs updated. The >> >> project would be to: 1) fix the ssl/https parts and use the >> >> latest version of taverna server 2) update the code to remove >> >> some legacy dependencies and use the latest APIs 3) improve the >> >> UI and responsiveness >> > >> > 1) & 2) are the most important bits with 3) being a bit extra >> > depending on >> >> how the first bits progress. >> > >> > >> > I understand the objectives slightly, but could you please >> > elaborate more about the project goals; may be about the >> > deliverables, objectives & things like that, so that would helpful >> > for me in getting started & also to write a good proposal. >> > >> > Thank you, Harsha >> > >> -----BEGIN PGP SIGNATURE----- >> Comment: GPGTools - https://gpgtools.org >> >> iQEcBAEBCgAGBQJU+ZtxAAoJEPK45GBX+Cy5VhwH/0kwtEgOdZ1CaMw9nh/zMwzZ >> lrQItNZgg3SknZqq6GV27zif9nNlh+wrYLdyu+Ms+bhy/TKSlzsAYtTHL40pg3qU >> MB4tbib5TYxLa8vYOWs12NpMAHtQ7DiD6upoOJRtUMvWmkSneqvlD8sjEVeC4768 >> a2BbI8nRoKFPI7WawPU8EgfX1JEo+i5+CVJ4Qe2cDG5fPs1CJ9IAqQp1bpi9i6ql >> LO9KhYQo/vninCnDRT9mMb6ve2szKZpMMqmMVh/VMjyyVEt5RCgwZsFs55sWN0L/ >> ih6n4TCeHr8XqcW4o7HpJtm3x4Yiwp5PYW4HjhH4lsaJFLKllroMIGreCkw2YIc= >> =CZsf >> -----END PGP SIGNATURE----- >> > >
