Building upon AndroidAPS? [EN]

Because of my blog post „Completing AndroidAPS objectives – the developer way“ many people contacted me to complain about the objectives. As I wrote in my article I see the need that people using AndroidAPS should have some kind of training, but I don’t like the way how it’s done today.

Also there are many other „functions“ in the app I find annoying (or patronizing). For instance:

  • Objectives (of course)
  • Spamming me with text walls when I want to give a second meal bolus
  • Try to kill me with SMB after eating something as hypo treatment
  • Many restrictions in the „Actions“ plugin
  • Warnings with 1g carbohydrate suggestions (seriously? 1g ??)
  • Version update warnings

Also I find there are missing features, such as:

  • Logging fitness tracker data
  • Using fitness tracker data for looping
  • Interface for „Alexa“ commands
  • Some kind of useful reminder for missed meal bolus

I thought about building a modified AAPS to takle these issues but the developers have some measures to make this harder than expected. For example, I just found a module to check the signature of the APK and compare it to a black list. If your signature is on the black list, the app will stop functioning.

In summary, using AndroidAPS as basis for your own app might be tricky. It takes a substantial amount of work to clear the code from all the „safety“ mechanisms to make sure the devs can’t pull a self destruction trigger, or leaking data, or something like that. I still don’t know if it’s worth my time to follow the „own app“ path…

What do you think about the measures of the devs? Do you think that’s the good right of the devs or is it violating the spirit of open source? Let me know in comments down below!

9 Kommentare zu “Building upon AndroidAPS? [EN]

    • To make this clear: I truly appreciate the work of all APS contributors. I think it is a really great movement and improves the lives of many diabetics. That’s why I also started contributing a few years ago.

      That being said, that doesn’t mean that it’s perfect and I think factual criticism is very important. Using a DIY APS system is very dangerous in all cases. There is no quality insurance process, no 100% test coverage, no certification, nothing. What makes it feeling secure, is that many people are using it and are still alive…

      To answer your second question: as far as I know, there is only AndroidAPS, Loopkit and OpenAPS accessible to the public. There are other systems from some people build their own software stack, but they don’t want to give it away.

      I personally like OpenAPS more. It simply doesn’t annoy me as much. But I don’t want to carry more devices than need and that’s why I switched to AndroidAPS.

  1. What I miss in your criticism is how you would address what the objectives are attempting to achieve: to reach the countless people with comparatively little experience, knowledge or talent who would benefit incredibly from looping.

    You can say you don’t care about reaching such users, and make the system very inaccessible, or you can attempt to make the system much more accessible IN A SAFE WAY, so many people can benefit. What you cannot do is make an easy-to-use system that makes it likely even a small ratio of your users will be put into a coma (I assume you agree).

    The „accessible in a safe way“ approach is what AAPS is trying to achieve. This is hard.

    So until you can come up with a realistic alternative on how to achieve that, or you are clear about how and exactly why you would formulate project goals differently, deeply involved community members will not take you seriously.

    • I don’t agree with your assumption what the objectives should achieve. In my opinion they should build a barrier and transfer responsibility. I think basically it should avoid a lawsuit.

      And I do care about people who don’t have any clue about how the looping system works, but in a complete different way: I think they shouldn’t do it! AAPS and OpenAPS are (from my experience) not perfectly save. If people don’t understand the mechanism of the algorithm they are prone to harm themselves.

      And last, I don’t agree with your assumption the objectives or any other text wall will make AAPS safer. A typical user reaction on a text wall is to not read it, and klick on any button until the program does what the user want.

      As there are reasonable good commercial loops available by now, I suggest leave the inexperienced users behind and go for the users who want the extra portion of control over the technology.

      • > I think basically it should avoid a lawsuit.

        This is important, but less than so than preventing harm in the first place.

        > And I do care about people who don’t have any clue about how the looping system works, but in a complete different way: I think they shouldn’t do it!

        This is simplifying reality to a degree I find unhelpful. We are neither thinking of non-expert users as „not having any clue“ nor do we think that AAPS will every be „perfectly safe“. You are arguing for an expert app with no concessions for non-experts. That doesn’t contradict your claim about caring about people who don’t understand. It also doesn’t contradict my hypothesis that you do not care about REACHING those people with AAPS.

        > And last, I don’t agree with your assumption the objectives or any other text wall will make AAPS safer. A typical user reaction on a text wall is to not read it, and klick on any button until the program does what the user want.

        The overwhelming feedback I have read and my own personal experience directly contradicts your statement. People care about their own personal safety and that of their close ones, and they succeed learning about AAPS with the objectives. Most (myself included) are incredibly grateful for the investment made into teaching.

        > As there are reasonable good commercial loops available by now, I suggest leave the inexperienced users behind…

        Thank you! This is what I was looking for. The reasoning I would suggest why this may be even harder to achieve is that successful open source software needs a community, and yours will be very small (I would disagree regarding the quality of commercial loops).

        I am not saying it is not possible, indeed go for it. If you do, please prove your own claim about caring about „people who don’t have any clue“ by doing all you can as an author to prevent them from using it!

      • P.S.

        > In my opinion they should build a barrier, and transfer responsibility

        The quality of a teaching tool is not judged by how well it keeps people out, but by how well it teaches. This is what the objectives are trying to do.

        Regarding responsibility: IANAL, but my impression is that successfully transferring all responsibility to the user for looping pretty much negates open source development. You can put as many „danger“ signs you want in front of a hand grenade, but if you leave it lying on the sidewalk I suspect the authorities will want to have a word with you.

  2. > The overwhelming feedback I have read and my own personal experience directly contradicts your statement. People care about their own personal safety and that of their close ones, and they succeed learning about AAPS with the objectives. Most (myself included) are incredibly grateful for the investment made into teaching.

    That’s pretty interesting to me, as I have the feeling it is exaclty the other way round.

    > Thank you! This is what I was looking for. The reasoning I would suggest why this may be even harder to achieve is that successful open source software needs a community, and yours will be very small (I would disagree regarding the quality of commercial loops).

    I suggest to take a look to OpenAPS. The software does not have objectives and (at least to my knowledge) nobody is working on something like this. On the other hand the software is only useable if you spend some time and effort to make it work for you, which I like more. And the community is big enough to drive the project forward. Also, the contribution of „people who don’t have any clue“ is negligible.

    However, at this point I can only thank you for sharing your perspective on the topic. I guess we simply have different opinions, which is fine to me.

    But the least we can agree on is that DIY loops are in essence a very good thing.

    • Agreed, yay for DIY loops! I don’t know much about OpenAPS and have no idea how it would hold up liability-wise, but hopefully it forces people to learn by virtue of being hard to configure.

  3. Hallo, Grusse aus Brazilien!
    (Mein deutch get nur bis, so I will continue in English)

    What I have seen following lots of groups and chats is that the majority of people have become radicals about AAPS. Rejecting all ideas, thoughts, aps and solutions other than AAPS.

    Despite all that, I understand why the devs did it. They have to protect their fellow T1Ds and nowadays hackers are trying to put malicious code on everything they can get their hands on.

    But as it is on a git repo, pull requests should be encouraged and would allow for several further improvements.

    Also the tricky questions of objective 3 are just silly.

Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:

WordPress.com-Logo

Du kommentierst mit Deinem WordPress.com-Konto. Abmelden /  Ändern )

Google Foto

Du kommentierst mit Deinem Google-Konto. Abmelden /  Ändern )

Twitter-Bild

Du kommentierst mit Deinem Twitter-Konto. Abmelden /  Ändern )

Facebook-Foto

Du kommentierst mit Deinem Facebook-Konto. Abmelden /  Ändern )

Verbinde mit %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.