What I did this week (April 30)

--Originally published at Hermes's Blog

I helped marco to finish the Android app for the final presentation, these changes were UI changes, and there are push notifications, that was something more elaborated than I tought, I needed to obtain some keys in our firebase database to let our server comunicate with the gcm service (now it’s called fcm, firebase cloud messaging), then when the user signups or updates their profile, the android device sends a token that fcm generated for the device and we store it in neo4j, then when someone invites that user to a pool or asks them to pay his debt we send a push notification to his phone. Now, to receive push notifications, we needed to register a service in the android app that would listen for the messages, then depending on the messages we would create a different behaviour when the user clicks the notification, or one of its buttons. It was a pain in the ass.

This is the app. https://play.google.com/store/apps/details?id=com.cooper.cooper

I think it does not have the most recent code.

What I did this week (April 30)

What I’ll do this week (April 23)

--Originally published at Hermes's Blog

This week I plan to finish the test and have at least 70% of the api code covered. We already have all signup, login, profile and delete account functionality tested and passing. I tried to include the tests in the travis-ci build but it seems to have problems with the neo4j service, neo4j is throwing this error:

Uncaught error when processing result: Neo4jError: 140223444313984:error:140770FC:SSL routines:SSL23_GET_SERVER_HELLO:unknown protocol:../deps/openssl/openssl/ssl/s23_clnt.c:827

And I just don’t know why, should I just remove the https code when the code is being executed in travis? Maybe it’s because it does not find the key and certificate, but I don’t see why that would be the case, I’ll try again anyway.

What I did this week (April 16)

--Originally published at Hermes's Blog

This week I started helping with the testing. I solved the issues that they were facing with the cookie. Marco is creating the web application and, when he was developing, found some bugs in the api, these are already fixed. Marco (and someone else that I’m not aware of, I guess) integrated the firebase chat to the app. That’s nice, altough it’s kind of ugly, but hey, it works.

We know we may have some security issues, in the chat and some parts of the api, but no can do, we are aware of that and will not fix it, we have to finish the things that we need to present in the demo.

What I’ll do this week (April 9)

--Originally published at Hermes's Blog

I have some things left to work in the api before helping Marco or Francisco with the mobile app and the tests.

  • Pool owners can specify custom amounts for each user, but they cannot make it automatic, I have to put a flag in the pool creation to set if the users will be charged the same amount, and also update everyones debt if a new user joins (This will only be updated if the pool has not started yet, because then users can start paying).
  • I need to notify users if they have a debt that have not paid (when the pool end date arrives).
  • With cash, owners should confirm the amount the users claim to have paid.
  • Store profile pics somewhere.

That’s what I’ll work this week, even if I don’t finish all of this I’ll start working in the Android app with Marco.

What I did this week (April 1)

--Originally published at Hermes's Blog

This week I worked in the flow logic for cash payments, a user can only pay a pool if it has a debt, but he or she can overpay and then the pool owner has a debt with him. Pool owners can edit the debts and amounts of the users but only if the debt is high enough to surpass the pool’s total. I was working implementing stripe, in fact, we were able to receibe payments from users that registered in stripe (we were missing the frontend that would comunicate with stripe for the registration). Anyway we decided that we don’t have the time to finish this, there are other, more urgent things to do before the final delivery, so we will drop the credit card functionality from the app and focus in making better what we already have.

What I’ll do this week (April 1)

--Originally published at Hermes's Blog

This week I’ll be working on the payment flow.

When a pool is marked as cash, users should say how much they paid and then the administrator should confirm the amount.

When a pool is marked as credit we will have to request the payment from the users and then send the same amount to the admin (we need to check if this is possible with stripe).

If the admin wants to update the amounts that the users have debt, we need to make sure that everything is kept within the limits of the initial costs (right now you can update a user’s dept to whatever amount). And maybe we should not be able to update the already paid amount (at least for credit), what’s already paid should not change.

Another nice to have would be to find friends on Facebook, instead of searching them by name or email.

That’s it.

What I did this week (May 25)

--Originally published at Hermes's Blog

Even though we were on vacations I made a lot of progress. From the Cooper API now we can:

  • Use another invitation flow: there is an endpoint where we can find the pools that we are invited to, an endpoint to accept invitations, and another one to decline invitations, this way we can make sure that only those who are invited join and that you can also decline invitations.
  • Disabled sendgrid, we will delete the code that send emails, everything will be done in the app.
  • Friend requests, now we can have friends, send, accept and decline friend requests.
  • Login with facebook.
  • I put a server on Digital Ocean, started using travis for continous integration, everytime we push to the server repository, the app is built and the tests are run, if they pass we can merge, when we create a pull request to master and merge it, the code is deployed automatically to the Digital Ocean Server and restarted.
  • Now there is an https version of the server on port 3443, altough the http server is still running on 3000, this was because facebook only allows login from an https server. We may need to get a real certificate (I created and signed one by myself and the browser shows a warning).

What I’ll do this week (March 12)

--Originally published at Hermes's Blog

The most important thing to do right now (in the api) is to be able to split the bill of a pool. I think the best way to handle this is, when the pool owner says that the bill should be splitted evenly, take all current users and divide the bill evenly, the problem is that not all users have joined by this time, all I can do is that when a new user confirms the invitation the API will update all the relations to users in the pool to fit the new bill split. When the ower sets the bill split to custom it I think it should be able to set the amount that invited users should pay, that will require me to add a new relation from the pool to users where invited users have a special relation with the label `:invited`, just so it can tell the user how much is he spected to pay and then he can accept or reject this invitation.

The TODO list is as follows:

  • Pool owners need to split the bill, they should be given the options of spliting it evenly or with custom amounts (do we need to make sure that the total of all the custom amounts sum up to the total of the bill?).
  • Pool needs to specify who you should pay to, how much money do they owe or who payed more than necessary.
  • The only payment method available right now is cash, the pool owner should be the only one able to change the pool and relation to the pool properties.
  • Maybe invites should be visible through an interface in the app, I think that the best way to approach this is to create a new relation between the pool node and user node that indicates that he’s been invited. Because otherwise there is no other way to find out who you have invited, the email is send and after that it all depends on the user clicking the link.
  • We need a Facebook and Google login (I think this will be done by someone else).
  • User profile images are not critical (because we have default profile gravatars) but would be nice to have, we should store them somewhere outside the database, maybe store the path to the file in the node, or use an external service like Amazon S3.