Hello, I'm worried that we are going a bit of track by trying to use the taverna player and that although the original intentions were noble we are actually making Larry's life harder by using the player. It's just a web app used as a convenience for uploading and running workflows on the taverna server built almost as an aside for another project. You could almost say that we should just improve that and make it mobile friendly rather than writing a native app. I think that the native android tavern app should be using the taverna server and not the player. The player does not have an api and talking to it will require jumping through some ugly hoops like grabbing csrf tokens from web pages and faking html forms for uploading workflows. I really think we are going down the wrong path by trying to use the player here. One of the original aims for this gsoc project was to write some client code that could talk to a taverna server in both http or https and we have lost sight of that.
Cheers, Ian -----Original Message----- From: Larry Akah [mailto:[email protected]] Sent: 22 June 2015 19:04 To: [email protected] Subject: Re: gsoc taverna mobile progress so far 2015-06-22 9:56 GMT-05:00 Ian Dunlop <[email protected]>: > Hello, > > @Larry I have had a look at the latest version of the tavern mobile > app after pull request #4. Here are some thoughts: > > > 1) The Usage and About pages from the left menu don't seem to do > anything. > Yes they don't for now, Actually, i think they are just trivial pages and would be implemented later. > > 2) Which service is the user logging in with? As of now, the login is with the taverna player portal, but as we have discussed, am trying to wrap that so it works for myexperiment > The code refers to both taverna server and taverna player and seems to > be logging in to taverna player using a GET. Does this actually do > anything since I logged in with a random account name and it all still worked. The login works with the taverna player portal but was not attached to allow for testing of some of the inner functionalities > (Actually, I was surprised this wasn't a POST request). We need to > make it clear what a user is logging into since the image on the login > screen implies that we are logging into myexperiment . Ok, understood. I'm not exactly sure what is gained by logging into the player. > This was beccause at first, i thought workflows would be coming from the player also. But as we have dicsussed, these workflows are from myexperiment. So i would do the necessary changes in the upcoming requests. > 3) In the settings we can change both tavern server and player URLs. > Do we need both. Isn't it a bit confusing. > The server and player can have different mount points depending on the organisation who could decide to have theirs installed elsewhere . > 4) Where are the workflows coming from? I thought they were from a > users account on myexperiment. In fact that is what I thought we were > logging into. > As of now, the workflows you see are from the player. Some samples were put there, and that's why you could see just one public workflow, the others are marked as private. But in upcoming versions, i'll make sure they are myexperiment workflows. > > 5) In the workflows tab I only have 1 workflow appearing. The workflow comes from the player. > There are 3 workflows in the favorites tab, although I haven't got any > favorites yet. Are these real workflows or just hard coded for the moment. They shouldn't be there as of the recent code, i'll just check that. maybe i haven't submitted the request yet but normally you would need to mark a workflow as such in order to see them in that section. > > 6) The powered by "myexperiment" logo seems really small. I'm not even > sure we need it at all. Ok, i guess i'll have to get that off then. > The menu screen says powered by Apache Taverna. It's a bit confusing. > This is a placeholder typo. it should read 'powered by MyExperiment' > 7) I clicked on workflow 'favorite' star and although a message > claimed that it had done something it wasn't added to the favorites list. > > Again, this maybe due to the fact that the code is older than what i > have here. A message actually appears indicating that the workflow hs been marked. > This seems like a lot of stuff to think about but that's really > because I want you to succeed. Apologies if you have already addressed > any of these issues in your fork since I appreciate that the apache > tavern repo is always behind yours. How do you feel you are > progressing with respect to the initial plan? Its not going quite perfectly as planned but it isn't going offtrack either. There were a couple of things i had to find out and learn about which were not planned for, but i guess the plan is still working good, thanks. > Is there anywhere where you need more clarification or help? Actually, i have tried to get some response as JSON from myexperiment workflows but have not yet been able. Say for example I need to get a list of all workflows and this is the link http://www.myexperiment.org/workflows.xml . Is there a way myexperiment could give me the response as JSON? Tried setting headers to accept json but it doesn't work. Also from the xml results, there seem not to be all the information about a workflow as is normally present on the myexperiment site itself. Am also wondering if the Taverna player would accept .t2flow files from myexperiment because the API specifies that the body required is in JSON format. Can the .t2flow files be converted as such? > It's almost time for the midterm assessment so we need to make sure > that we are (mostly) on track. > > Cheers, > > Ian > Cheers -- *Akah Larry N.H* *Lead Software Engineer @ GiftedMom* *Founder IceTeck Inc* *www.iceteck.com* Developing technologies for emergence and sustainable development.
