3 ms·

>The approach was the same as Cloudflare’s vinext rewrite: port the official jsonata-js test suite to Go, then implement the evaluator until every test passes.

>The approach was the same as Cloudflare’s vinext rewrite: port the official jsonata-js test suite to Go, then implement the evaluator until every test passes. the first question that comes to mind is: who takes care of this now? You had a dependency with an open source project. now your translated copy (fork?) is yours to maintain, 13k lines of go. how do you make sure it stays updated? Is this maintainance factored in? I know nothing about JSONata or the problem it solves, but I took a look at the repo and there's 15PRs and 150 open issues.

simonw 7mo ago
That's only important if the plan is to stay feature-compatible with the original going forward. For this case, where it's used as an internal filtering engine, I expect the goal is fixing bugs that show up and occasionally adding a feature that's needed by this organization.
- shimman 7mo ago
  This case looks like pure marketing fluff rather than sound engineering tho.
- torben-friis 7mo ago
  >expect the goal is fixing bugs that show up and occasionally adding a feature that's needed by this organization. Even if we assume a clean and bug free port, and no compatibility required moving forward, and a scope that doesn't involve security risks, that's already non trivial, since it's a codebase no one has context of. Probably not 500k worth of maintainance (because wtf were they doing in the first place) but I don't buy placing the current cost at 0.
- PetahNZ 7mo ago
  If the original released a bunch more features that you wanted why wouldn't you just redo the conversion against the latest version?
- kikimora 7mo ago
  In practice the biggest issue will be documentation and tutorials. If JSONata diverges from their fork users will have problems reconciling what they see online with their engine capabilities.
delijati 7mo ago
it is all yolo from here on out ... every major ai decision we're making today feels like a bet that agi will eventually show up and clean up the mess
- 52-6F-62 7mo ago
  There is a choice, yet.
bawolff 7mo ago
I mean, my first question would be how good the test suite on this project is.
Herring 7mo ago
The full translation took 7hrs and $400 in tokens. Applying diffs every quarter using AI is much easier and cheaper. Software engineering has completely changed.
- aniceperson 7mo ago
  except there are 2 go implementations already, and he burnt 500k per year to have a kubernetes clusters to parse json (???), so the total gain is -500000*year - 400 + 1 (deducting prompt to use existing implementation)
- throwaway613746 7mo ago
  [dead]
saadn92 7mo ago
> the first question that comes to mind is: who takes care of this now? probably another AI agent at their company, who I'm sure won't make any mistakes

simonw 7mo ago

That's only important if the plan is to stay feature-compatible with the original going forward. For this case, where it's used as an internal filtering engine, I expect the goal is fixing bugs that show up and occasionally adding a feature that's needed by this organization.

shimman 7mo ago
This case looks like pure marketing fluff rather than sound engineering tho.
torben-friis 7mo ago
>expect the goal is fixing bugs that show up and occasionally adding a feature that's needed by this organization. Even if we assume a clean and bug free port, and no compatibility required moving forward, and a scope that doesn't involve security risks, that's already non trivial, since it's a codebase no one has context of. Probably not 500k worth of maintainance (because wtf were they doing in the first place) but I don't buy placing the current cost at 0.
PetahNZ 7mo ago
If the original released a bunch more features that you wanted why wouldn't you just redo the conversion against the latest version?
kikimora 7mo ago
In practice the biggest issue will be documentation and tutorials. If JSONata diverges from their fork users will have problems reconciling what they see online with their engine capabilities.

delijati 7mo ago

it is all yolo from here on out ... every major ai decision we're making today feels like a bet that agi will eventually show up and clean up the mess

52-6F-62 7mo ago
There is a choice, yet.

bawolff 7mo ago

I mean, my first question would be how good the test suite on this project is.

Herring 7mo ago

The full translation took 7hrs and $400 in tokens. Applying diffs every quarter using AI is much easier and cheaper. Software engineering has completely changed.

aniceperson 7mo ago
except there are 2 go implementations already, and he burnt 500k per year to have a kubernetes clusters to parse json (???), so the total gain is -500000*year - 400 + 1 (deducting prompt to use existing implementation)
throwaway613746 7mo ago
[dead]

saadn92 7mo ago

> the first question that comes to mind is: who takes care of this now? probably another AI agent at their company, who I'm sure won't make any mistakes