# Error Correction Algorithm - questions

**URL:** <https://kopia.discourse.group/t/error-correction-algorithm-questions/1585>\
**Category:** General Topics\
**Created:** [November 29, 2022, 8:44am UTC](https://kopia.discourse.group/t/error-correction-algorithm-questions/1585 "2022-11-29T08:44:14Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![kapitainsky](https://yyz2.discourse-cdn.com/free1/user_avatar/kopia.discourse.group/kapitainsky/32/535_2.png) [@kapitainsky](https://kopia.discourse.group/u/kapitainsky)\
**Post date:** [November 29, 2022, 8:44am UTC](https://kopia.discourse.group/t/error-correction-algorithm-questions/1585/1 "2022-11-29T08:44:14Z")

</div>

Have discovered kopia recently and I am truly impressed by design and functionality available. One feature that caught my eye is experimental Error Correction Algorithm. I have some questions which are not explained anywhere as I am aware (did some google).

I can choose 1, 2, 5, 10% overhead for Reed-Solomon (RS) ECC coding but what does it translates to in terms of error corrections capabilities? What does it protects me from? What I am missing is some quantified data.

I imagine that most likely error on HDD would be reallocated bad sector - let’s say 4 KiB nowadays. Would it be possible to support RS overhead able to correct such situation (I guess 10% can not fix 32768 consecutive bits missing)? It would make interesting option for many of us not using e.g. ZFS RAID NAS.

I think there is no other program available today which can make it possible. kopia already has all required bits in place - now it would be only question of math. Overhead definitely should be less than 100% but would cover most likely issue people experience backing up to local disk. Clearly something to make kopia unique among backup programs.

Let’s say required overhead is 70% (ballpark figure - do not know enough about RS) - for many it would perfectly acceptable to have 10TB disk for kopia repo destination where I could host 6TB backup with some cheap person protection against most likely failure - bad sector.

---

<div class="post-metadata">

**Author:** ![rhinomagnetar](https://avatars.discourse-cdn.com/v4/letter/r/a88e57/32.png) [@rhinomagnetar](https://kopia.discourse.group/u/rhinomagnetar)\
**Post date:** [June 6, 2023, 9:45pm UTC](https://kopia.discourse.group/t/error-correction-algorithm-questions/1585/2 "2023-06-06T21:45:19Z")

</div>

I have the same question. I have also tried researching about the different percentages but have failed to find any answers either.

Another question I would like to ask is why is it still experimental. Are there any bugs that have not been addressed?

Btw there is another app with that feature - Picocrypt, though its main goal is file encryption rather than backups.

---

<div class="post-metadata">

**Author:** ![kapitainsky](https://yyz2.discourse-cdn.com/free1/user_avatar/kopia.discourse.group/kapitainsky/32/535_2.png) [@kapitainsky](https://kopia.discourse.group/u/kapitainsky)\
**Post date:** [June 6, 2023, 9:54pm UTC](https://kopia.discourse.group/t/error-correction-algorithm-questions/1585/3 "2023-06-06T21:54:53Z")

</div>

I think this fantastic project (kopia) is close to be abandoned. If you look at github issues and PRs you will see that there is no activity (but bots). Multiple forum posts here just stay unanswered - there is no really active community. I have had great hopes about kopia but I do not think it has energy to carry on.

---

<div class="post-metadata">

**Author:** ![rhinomagnetar](https://avatars.discourse-cdn.com/v4/letter/r/a88e57/32.png) [@rhinomagnetar](https://kopia.discourse.group/u/rhinomagnetar)\
**Post date:** [June 6, 2023, 10:35pm UTC](https://kopia.discourse.group/t/error-correction-algorithm-questions/1585/4 "2023-06-06T22:35:13Z")

</div>

Yeah, I noticed that too. Hopefully not as I’m just starting to use it. Though according to one of the replies of the developer, they are just busy.

> [@Is this project dead?](https://kopia.discourse.group/t/is-this-project-dead/1166/2):
>
> Ah, the first sign of a project maturity .D But a bit more (but just a bit) seriously, maybe people are too busy committing code: [Commits · kopia/kopia · GitHub](https://github.com/kopia/kopia/commits/master) .P

Though it was more than a year ago, I think it still applies as the last update is as recent as last month. However, there seems to be a need to improve the docs and the GUI as the lack of replies might be due to the lack of people who understand how Kopia works.

---

<div class="post-metadata">

**Author:** ![kapitainsky](https://yyz2.discourse-cdn.com/free1/user_avatar/kopia.discourse.group/kapitainsky/32/535_2.png) [@kapitainsky](https://kopia.discourse.group/u/kapitainsky)\
**Post date:** [June 7, 2023, 11:24am UTC](https://kopia.discourse.group/t/error-correction-algorithm-questions/1585/5 "2023-06-07T11:24:10Z")

</div>

Hopefully you are right. I think kopia has great potential and won’t slowly [rot](https://en.wikipedia.org/wiki/Software_rot). What worries me is that there are plenty of open issues about potential problems or bugs without anybody taking active interest. Well. I know it is open source:) So I could try myself to help here. Unfortunately I am not programmer.
