24 votes

Make it ephemeral: Software should decay and lose data

23 comments

  1. kollkana
    Link
    This is a really weird take. Only the final paragraph makes any kind of sense. "I don't want to make my own deletion schedule or change how I organise my files/systems, so make all software delete...

    This is a really weird take. Only the final paragraph makes any kind of sense.

    I will forget to delete, and I'm lazy and given the tools available I rarely clean up. Yet many of the things I create I already know I really only need for a week or [two]. So give me a button I can press to schedule deletion. Then I don't have to remember to clean up after myself a few months later, but I can make that call already today when I create my thing.

    "I don't want to make my own deletion schedule or change how I organise my files/systems, so make all software delete stuff automatically"

    42 votes
  2. [4]
    MimicSquid
    (edited )
    Link
    As someone in a finance role, I completely disagree. It's important that if I need to prove that someone has a sandwich with a particular vendor 5 years ago I can lay my hands on that receipt and...

    As someone in a finance role, I completely disagree. It's important that if I need to prove that someone has a sandwich with a particular vendor 5 years ago I can lay my hands on that receipt and a note that explains the purchase. No data should ever be removed without a human choosing to delete it. Even after the 3 or 7 year line where you don't need to save things for taxes there will be documents you want to hold on to. Imagine losing the original decade-old lease agreement because your software decided to purge. No thank you.

    21 votes
    1. [3]
      Johz
      Link Parent
      I mean, this is kind of exactly what the author is talking about. Keeping data for five years, seven years, ten years - whatever is necessary to ensure compliance - makes sense. But it's still...

      I mean, this is kind of exactly what the author is talking about. Keeping data for five years, seven years, ten years - whatever is necessary to ensure compliance - makes sense. But it's still personal and private data, and it still needs to be deleted at some point.

      The danger is always that this data ends up leaked, either through negligence or attacks or whatever else. Reducing the amount of data that exists to be leaked in the first place should be a key goal of any security measures.

      Yes, there are always going to be some exceptions, but designing tax and controlling laws around the idea that data doesn't need to be kept eternally to still be acceptable is important.

      2 votes
      1. [2]
        MimicSquid
        Link Parent
        The laws already have deadlines for how far back you need to be able to provide documentation, and it's generally 3 or 7 years. Practically, I sort documents in subfolders by topic and then year...

        The laws already have deadlines for how far back you need to be able to provide documentation, and it's generally 3 or 7 years. Practically, I sort documents in subfolders by topic and then year specifically for this purpose. When it's the right number of years old, it's two clicks to delete the appropriate years. But it only happens when I want it to, and that's important. Imagine a 3 year automatic deletion trigger that goes off when you're being audited for the last three years?

        Maybe future generations will have solved things, but the thought of automatic data deletion makes me itchy.

        5 votes
        1. Johz
          Link Parent
          I kind of get the worry about a fully automated system, but fundamentally my data is my property, and if you're holding it then you have a duty to me to handle that data correctly, and not hold it...

          I kind of get the worry about a fully automated system, but fundamentally my data is my property, and if you're holding it then you have a duty to me to handle that data correctly, and not hold it any longer than you need to.

  3. [14]
    TurtleCracker
    Link
    I mostly agree with this take. It would be great if transactional email or marketing email for example automatically deleted itself after a year. How much energy are we wasting on storing coupons...

    I mostly agree with this take. It would be great if transactional email or marketing email for example automatically deleted itself after a year. How much energy are we wasting on storing coupons that expired years ago?

    7 votes
    1. [8]
      kollkana
      Link Parent
      There are already tools to do that sort of thing anywhere it would be useful. Email clients allow for filtering rules and auto-deletion within folders etc. That's what I find strange about this...

      There are already tools to do that sort of thing anywhere it would be useful. Email clients allow for filtering rules and auto-deletion within folders etc. That's what I find strange about this opinion piece, it seems to be talking about ignoring existing tools (because that requires setup and effort, I guess) in favour of software systems deleting stuff based on vague criteria that the end user doesn't need to set.

      19 votes
      1. Akir
        Link Parent
        Personally speaking that sounds like a nightmare. I for one do not want my software to decide for me that I don’t need to see any information that I feed into it. If that is a functionality I...

        Personally speaking that sounds like a nightmare. I for one do not want my software to decide for me that I don’t need to see any information that I feed into it. If that is a functionality I want, it needs to be based on my personal criteria rather than something some faceless organization had determined for me.

        This entire philosophy strikes me as essentially lazy. They talk a lot about clutter. But the thing about clutter and junk is that you are supposed to clean it up. Don’t blame the fact that your data is bad on the software; you’re supposed to be the one maintaining that data! They mention SaaS and dashboards as being their area of concern, but I honestly can’t find any excuse why someone would just use whatever default dashboard for information instead of configuring it to be useful to them. If your reports are showing old irrelevant data, the solution is not to have it throw out that data, it is to narrow the report to the data that is actually relavent. Going the direction the author suggests is madness.

        15 votes
      2. [6]
        TurtleCracker
        Link Parent
        It’s more around the default behavior. I think the default should be expiring data. You should have to take specific action to preserve. It’s one of the reasons I use Signal for messaging.

        It’s more around the default behavior. I think the default should be expiring data. You should have to take specific action to preserve. It’s one of the reasons I use Signal for messaging.

        2 votes
        1. [5]
          kollkana
          Link Parent
          Why do you think the default should be expiring data?

          Why do you think the default should be expiring data?

          8 votes
          1. [4]
            TurtleCracker
            Link Parent
            A default of expiring data provides more security and privacy to the user.

            A default of expiring data provides more security and privacy to the user.

            1 vote
            1. [2]
              kollkana
              Link Parent
              That depends heavily on the data in question. If something like my opt-out from the public electoral roll automatically expires that frees up my name and address to be listed online by any...

              That depends heavily on the data in question. If something like my opt-out from the public electoral roll automatically expires that frees up my name and address to be listed online by any business that buys the list. If files on my own computer expire that means I need more physical backups of important documents or have less security that e.g. I can make an insurance claim, prove my employment history, prove medical diagnoses etc.

              4 votes
              1. TurtleCracker
                (edited )
                Link Parent
                The default is more in the context of SaaS applications and/or cloud based services. Not local files. I do have my default Downloads directory automatically clean up most file types on a pretty...

                The default is more in the context of SaaS applications and/or cloud based services. Not local files. I do have my default Downloads directory automatically clean up most file types on a pretty regular basis though.

                2 votes
            2. papasquat
              Link Parent
              A default of not receiving email at all provides more security and privacy to the user as well. The tradeoffs aren't worth it though. I like being able to find an Amazon shipping notification in...

              A default of not receiving email at all provides more security and privacy to the user as well. The tradeoffs aren't worth it though.

              I like being able to find an Amazon shipping notification in my email from 10 years ago. Stuff like that comes in handy all the time. Do I really care if someone gets access to that decade old shipment notification? No, probably not.

              I would say that the majority of the time, stale, old data is far more valuable to me than it would be for any attacker, and I can honestly say I've never had a moment where I've looked at data and thought "man, I sure wish my computer automatically deleted that after an arbitrary timeframe"

              4 votes
    2. [5]
      Omnicrola
      Link Parent
      Overall I also mostly agree with the idea of some degree of ephemerality in data. For email specifically, I do very much appreciate being able to dig out receipts for things. I've occasionally had...

      Overall I also mostly agree with the idea of some degree of ephemerality in data. For email specifically, I do very much appreciate being able to dig out receipts for things. I've occasionally had a need for a receipt from 4+ years ago. If I had to intentionally archive or physically file such things I would never have managed it.

      I remember a long time ago (10+ years) being very intrigued by a similar suggestion but for Facebook. If you didn't interact in a "meaningful" way with a friend in your friends list, their profile image would slowly fade into sepia tones, and eventually automatically un-friend. The article was centered around how this more closely parallels how some real life relationships fade away even though nothing is "wrong", and that this is completely normal and arguably healthy.

      7 votes
      1. [4]
        TurtleCracker
        Link Parent
        I think the email service Hey kind of addresses with some of the features like sorting into "Paper trail" etc... I agree being able to look up receipts in the future is valuable, but the majority...

        I think the email service Hey kind of addresses with some of the features like sorting into "Paper trail" etc...

        I agree being able to look up receipts in the future is valuable, but the majority of my email is not receipts. A lot of data loses the majority of it's value in a remarkably short period of time. Maybe it would be nice if services let you set a default preference for data.

        1. [3]
          Akir
          Link Parent
          All email are receipts. If it’s correspondence then they are merely receipts of conversation. I for one appreciate that I have an archive of messages from people who are no longer are alive, or...

          All email are receipts. If it’s correspondence then they are merely receipts of conversation. I for one appreciate that I have an archive of messages from people who are no longer are alive, or who I no longer have any contact with.

          4 votes
          1. [2]
            TurtleCracker
            Link Parent
            I just popped open my email and looked at the 10 most recent messages. 2 Digests/Summaries of activity 4 Marketing emails 2 Shipping notifications 1 Recap of a support chat 1 Request for a survey...

            I just popped open my email and looked at the 10 most recent messages.

            2 Digests/Summaries of activity
            4 Marketing emails
            2 Shipping notifications
            1 Recap of a support chat
            1 Request for a survey

            Of the 10 most recent emails I received only one of them holds any value beyond about a 1 week period, and that value is minimal at best.

            My friends don't email me. They text me, message me on Signal, or message me on Discord. The majority of the communication is through Signal and it self deletes everything after 4 weeks. I have never wished I could reference a message on Signal that had already deleted itself.

            I think we may structure our communications very different.

            1. Akir
              Link Parent
              It sound to me that you only use email like a repository of conversations with robots instead of an actual communications method. I can't help but wonder what the last 10 emails you sent were? No...

              It sound to me that you only use email like a repository of conversations with robots instead of an actual communications method. I can't help but wonder what the last 10 emails you sent were? No need to answer that, though. Of the last 10 emails I got in my work inbox, only 3 of them were sent by robots.

              For my personal email, I get a lot of robot email, true, but it's important to keep them for an extended period because of Murphy's law: anything bad that can happen, will. Having a paper trail is usually required to get any real action done.

              1 vote
  4. googs
    (edited )
    Link
    These are good thoughts, at least for some cases. Obviously auto-deleting data is not going to be a good fit for every system. With that said, lots of SaaS and cloud tools have this pattern...

    These are good thoughts, at least for some cases. Obviously auto-deleting data is not going to be a good fit for every system.

    With that said, lots of SaaS and cloud tools have this pattern built-in already! One example I'm familiar with is Azure Storage. You can set up what they call a "Data Lifecycle Policy" that will basically move your data to cheaper "Cold" storage after some set amount of time, or you can configure it to just delete the data out right after some time.

    It's more a matter of getting buy-in to turn this stuff on. Companies can be pretty protective of their data, no matter how useless it is :). But the technology of an auto-delete solution is not so complex in most cases.

    The article mentions Datadog notebooks. I'm not really familiar with Datadog, but maybe it would be possible to set up a cron job that calls the Delete a notebook API? I'm not sure how realistic of an option this is, and wouldn't be as easy as clicking a "schedule to delete" button, but could be useful?

    4 votes
  5. MechanicalMagpie
    Link
    I'm pretty sure they're saying that "the trash bin pretty much only exists in email clients and operating systems, for the purpose of cleaning up old emails and files, respectively" but i read it...

    Yes, the trash bin was created as an appoximation of this, but the bin seemingly did not make it farther than file or email management software.

    I'm pretty sure they're saying that "the trash bin pretty much only exists in email clients and operating systems, for the purpose of cleaning up old emails and files, respectively" but i read it very differently bc the number of people I've encountered who store things in the recycle bin/deleted items folder and then get upset when those things are deleted is greater than zero and endlessly frustrating. if you wouldnt use your physical trash can as a file cabinet, dont use the digital trash can as a file cabinet either.

    All that aside, i dont really agree with their premise. Maybe i just dont use my technology the same way as them, but i cant think of a single piece of software that i would want to (or trust to?) auto-delete my stuff without my input. would i like my software to give me the option to easily set up rules/tasks to delete old files? i mean, maybe. but i feel like if i truly wanted something to be ephemeral, id just???? use paper???? like paper still exists. Something something use the right tool for the job or whatever.

    4 votes
  6. papasquat
    Link
    The authors premise doesn't even hold true here. Physical data doesn't auto delete, and in fact is often times far more resilient than digital data. Egyptian hieroglyphs and sumeran poetry exists...

    The authors premise doesn't even hold true here. Physical data doesn't auto delete, and in fact is often times far more resilient than digital data. Egyptian hieroglyphs and sumeran poetry exists carved in stone from thousands of years ago. There are physical, readable books that are older than my great great grandfather.

    On the other hand, the copy of office space I downloaded into a hard drive in 2005 is almost certainly unreadable due to bitrot.

    Even mirrored, backed up data in cloud storage providers likely won't exist there anymore if I don't interact with them in a couple decades as their licensing agreements change, ownership shifts, and monetization strategy is flipped around.

    Compare that to the sticky note the author referenced. If I wrote that note and then didn't touch it, it would almost certainly outlive me.

    Digital data is inherently extremely ephemeral. Users only get the impression it isn't due to gargantuan, monumental efforts. At the end of the day it's still more ephemeral than most physical records.

    If you're not looking at the modified fields of files and cleaning up useless ones, thats on you, just like keeping a messy desk is.

    4 votes
  7. Notcoffeetable
    Link
    So at base I kind of disagree with the track the author has gone down. A lot of software does auto-delete, and in fact I've had to turn it off on my cell phone for legal reasons. But since I saw...

    So at base I kind of disagree with the track the author has gone down. A lot of software does auto-delete, and in fact I've had to turn it off on my cell phone for legal reasons.

    But since I saw this topic I've been wondering: what would an ephemeral internet look like? We have Snapchat and instagram stories. Can we small webify ephemeral social media?