@Silvester-Purple-Leopard#148374 Dat is gewoon mogelijk. Ben zelf nog nooit tegen een limiet aangelopen, meestal was het ergens een foutje (bijv. in de applet filter code). Op de wiki staan een aantal voorbeelden met dergelijke filter code helemaal onderaan de volgende pagina:
https://github.com/woudt/bunq2ifttt/wiki/Examples-and-Tutorial

Als er iets fout gaat kun je in het IFTTT activity screen een eerste indruk krijgen, en daarna zul je in de logs van bunq2IFTTT moeten kijken. Post gerust je filter code en eventuele meldingen in de logs hier als je er niet uit komt.

    HI @Edwin-Red-Lion , super thanks voor je werk. Ik heb alles up and running.
    Maar 1 vraag. Bij de endpoint tests krijg ik een error bij 'queries/list_all_things'
    Deze zijn allemaal rood:
    status code is 200
    body is properly JSON-encoded
    'data' property exists and is an array
    returns at least 2 items
    each 'list_all_things' object has all ingredients

    Volgens mij zit het m hierin:
    {
    "queryFields": {},
    "user": {
    "timezone": "America/Los_Angeles"
    },
    "ifttt_source": {
    "id": "#########",
    "url": "http://example.com/#########"
    }
    }

    De ######### heb ik even zelf gedaan.

    Kan het kloppen dat de example.com URL moet worden vervangen ergens? Ik kan het niet echt plaatsen.
    Thanks!

      @DeInteractieOntwerpers-B-V-Green-Puma#148544 bunq2IFTTT ondersteunt geen queries, alleen triggers en actions. Vermoedelijk heb je een voorbeeld service van IFTTT hergebruikt. Ik raad je aan de list_all_things query te verwijderen.

        Really? Thought I just used your tutorial but diving into it!

          @DeInteractieOntwerpers-B-V-Green-Puma#148552 I understand IFTTT has changed something on their side that for new users they now create an example service rather than starting from an empty sheet.

          I will update the wiki to reflect this. Thanks!

            @Edwin-Red-Lion#148618 It was a default feature indeed. I have removed it. Problem now is i discover the filter feature are part of a 200 dollar account in ifttt. Bit of a bummer now

              @DeInteractieOntwerpers-B-V-Green-Puma#148671 You can create personal filters without paying. You cannot publish them to other users however (but why would you do that?)

                @Edwin-Red-Lion#148694 Not the intention. I found out I used company profile instead of personal. Now it works!

                  @Edwin-Red-Lion#148400 Het is gelukt. Geen idee wat er eerder mis ging. Ik zal later de filter code posten. En wat logging om te laten zien welke fout codes er waren.

                    18 days later

                    So I have this a applet set up that will trigger on a specific event (receive salary) and move the balance before that transaction to a savings account.
                    This month it triggered and failed. 10 minutes later it triggered again, and this time succeeded.
                    Even though the first time it failed, the amount was still transferred.
                    The failed ifttt event shows the same time as the bunq transaction of salary received. The first amount was transferred at that same time.
                    The second transfer was at the same time as the second ifttt event.

                    So there are three things that have me wondering
                    is it standard behavior of ifttt to retry if an applet fails
                     how come ifttt shows a failed event, but the action was performed correctly
                    * is there any way to see the actual error in ifttt? Now it only shows "failed"

                    Any help and suggestions are much appreciated!

                      @DickZeeman#151483 While I have seen triggers fail now and then, I have never seen an action fail, so I haven't observed this behaviour from IFTTT. Perhaps IFTTT received an error or thought the request had timed out, and retried it. This is of course nice if it really failed, but in your case this of course resulted in a bad outcome...

                      To see more logs, here are the things you can do:
                      1. Click on the item in the IFTTT event log, sometimes it shows some more information
                      2. Go to IFTTT platform, then analytics -> API health. All errors are shown at the bottom.
                      3. Go to the Google Cloud Console log viewer: https://console.cloud.google.com/logs/viewer

                      Let me know if you get anything from those. I am of course interested to know if there is anything I can do to fix this!

                        @Edwin-Red-Lion#151495
                        Thanks for the suggestions. No 1 and 2 don't show anything helpful. However, the Google Cloud log shows
                        "Error received from peer ipv4:216.58.205.42:443
                        too much contention on these datastore entities. please try again"

                        The IP address mentioned in the log message is a google IP, so I'm guessing the cloud platform was too busy for my free tier work.
                        The log entries end with "Retries left: 1" which confirms my suspicion that it will automatically retry on failure.

                        As I understand it so far
                        Line 3-4 it signals a mutation
                        Line 6 it matches the mutation against my rule set for the trigger
                        From line 10 - 17 are all related to the error

                        I'll have to study the log a bit more, since several ifttt actions are triggered at the same time (balance, mutation), making the log a bit confusing.

                          @DickZeeman#151514 These logs are not related to the (failed) payment action, but are all related to the incoming mutation. For some reason bunq is calling you twice for the same mutation (line 1 to 4). I have built in duplicate detection, which means the first one is successful (line 5 to 9) and the errors are a result of the second one being unable to get a lock for the same transaction (line 10 to 18).

                          Are there any logs for the payment action?

                            @Edwin-Red-Lion#151516
                            Ah, sorry I could have seen that myself :)

                            After the post of the internal payment method a similar error occurs: "too much congestion". It looks like a timeout as the error is raised by , so maybe the Google cloud gives up after 200 miliseconds while bunq is still processing the action.


                            The error is raised at 10:14:33.077. I can see the mutation-trigger that is caused by payent arrive on 10:15:17.209, 44 seconds later.

                            I'll just wait and see what happens next month!

                              a month later

                              Just a reminder to everyone: if you have installed bunq2IFTTT before the 30th of January and not upgraded after that date to release 2.2 or 2.3: please upgrade in the next week.

                              On the 28th of April bunq will introduce a breaking change to the API which will mean any older releases will no longer work.

                                11 days later

                                Hi, I installed bunq2IFTTT on my raspberry. Wehen I start it with
                                Python3 Main.py I get the following response

                                • Serving Flask app "main" (lazy loading)
                                • Environment: production
                                  WARNING: This is a development server. Do not use it in a production deployment.
                                  Use a production WSGI server instead.
                                • Debug mode: on
                                • Running on http://localhost:18000/ (Press CTRL+C to quit)
                                • Restarting with stat
                                • Debugger is active!

                                Can you please point me into the right direction?
                                Also I can‘t connect to <ip>:18000 from another device.

                                Trank you very much :)

                                EDIT

                                I managed to run it without development by using:
                                env FLASK_ENV=development FLASK_APP=main.py flask run

                                But I still can't connect from another machine.
                                curl 127.0.0.1:5000 over another ssh connection shows me the html so it's defenetly running.

                                EDIT2
                                env FLASK_ENV=development FLASK_APP=main.py flask run --host=0.0.0.0
                                makes a flask server visible in the local network. So this is working now.

                                But I think I will try to use NGINX

                                  24 days later

                                  @Hermiot#162255 Hi Martin, sorry I missed your post. Have you been able to run it successfully?

                                  It is indeed not recommended to run the flask web server in production.

                                    13 days later

                                    @Edwin-Red-Lion thank you for your work here, I was able to port the project to Heroku because you took the time to both create it and to separate the storage concerns out, reimplementing the storage backend in postgres only took a few hours :)

                                      2 months later

                                      Release 2.4 has just been released.

                                      Thanks to a suggestion by Olaf Haalstra, I have implemented two new actions to ensure an account has a certain target balance, e.g. by topping up from another account or skimming excess balance to another account.

                                      If anyone else has any suggestions, please let me know!

                                        3 months later

                                        @Edwin-Red-Lion#191960 I've just updated to version 2.4. I don't think this is related to that, but when I now perform the Endpoint test, validation of "actions/bunq_change_card_account/fields/card/options" fails at "valid request". The failing part is "returns at least one option". Have you maybe seen this before?