kantselovich

All docs for evidence.dev are AI-hallucinated …YMMV

Say you have an inventory - list of cloud database instances from all engineering teams in the org.

Is this list growing over time or shrinking or remains the mostly the same?

If the list is growing, is it growing evenly across all teams , or some teams stand out? Can I narrow down to a few teams and compare?

What about overall health of the inventory - how many instances list security or billing optimization findings, how many instances are on the version that would be out of support soon and would need to be upgraded?

A dashboard can answer all those questions immediately and make issues visible, as opposed to a long list in a form of a spreadsheet.

I like the observable framework too - very flexible, you can nail very specific visualization details by dropping down to JS/html.

Maybe this is targeted to pure SQL folks that can get lost in UI details.

PS: one issue that I have with observable framework is authentication and authorization , it requires some system to be built on top of it to handle authn.

If there is no database, where one should persist state?

Most software is stateful and needs to persist state across restarts, so I would argue that one needs at least SQLite.

On SQLite being safe default: in practice it means supporting multiple databases, say SQLite and Postgres, this is more complicated that supporting just Postgres. As soon as a project leaves localhost and enters cloud development you need talk to a database over network, which warrants MySQL or Postgres.

Which is more complicated: supporting a docker container with mysql or Postgres for local development OR supporting multiple databases in the project?

Of course, the answer could be “it depends”, I but I would not call SQLite a default choice. It would be if you are writing desktop or mobile app, but for anything like a web app it’s a questionable choice.

I think the findings that the LLM triggers “desperation” like emotions when it about to run out of tokens in a coding session have practical implications. The tasks needs to be planned, so that they are likely to be consistent before the session runs into limits, to avoid issues like LLM starts hardcoding values from a test harnesses into UI layer to make the tests pass.

Thank you for the detailed write up.

I’m was thinking about building native windows UI, wrapping around cross platform library written in swift. I did not know it was that messy and complicated.

I recall, at the time Bloomberg and their source were taking about tiny chip on the bmc that was masking as a resistor.

However they did not produce any concrete evidence, citing NDA between that security company and their client.

I don't think it was a confirmed story. That is, the tiny "grain of rice" size Ethernet module that CEO of a security audit company allegedly found, was not present in other SuperMicro servers. SuperMicro itself, as well as it's buggest customers did not confirm the findings.

From what i recall, the story was very vague, there were no pictures of the specific chip, no pictures of the motherboard of the motherboard that would include serial, i.e. no details that would accompany a serious security research.

100%! Lots of issues are only discovered when enough code has been written. More than that , other issues are discovered only when the project is actually deployed as MVP.

TDD helps a lot, but it’s no guarantee - LLM is smart enough to “fake” the code to pass tests .

I’m working on project - a password manager, where I have full end to end test harnesses - cli client makes changes, sync them to the server and then observe the data in iOS app running in the emulator. More than once I noticed codex just hard coded expected values from the test harnesses directly into UI layout in iOS app to make the test pass…

Similar issues in the crypto layer - tests were written first , then code was written . During the review I noticed that the code was made to just pass the test - the logic was to check if signature values exists instead of checking if crypto signature is valid.

LLM can help with code reviews as well, but it has to be guided specifically what to look for for. This is with codex 5.4 model

I’m at level 6 according to this article. I have solid harness, but I still need to review the code so I can understand how to plan for the next set of changes .

Also, I’m struggling to take it to multiple agents level, mostly because things depend on each other in the project - most changes cut across UI, protocol and the server side, so not clear how agents would merge incompatible versions.

Verification is a tricky part as well, all tests could be passing, including end to end integration and visual tests, but my verification still catches things like data is not persisted or crypto signatures not verified.

I've glanced over the article - I think using SQLCypher makes sense if it benefits Keepass project internally, meaning makes it easier to implement new features or fix bugs.

I do not see how it would benefits end users in any significant way. Sure, you can look inside the database, how many people need that?

Dumping the database to CSV is not a good backup, schema changes over time, what was dumped from one version of the app would not work for importing into another version if schema changes. Backup it needs a versioned schema format, which would actually look like KDBX format if implemented in XML :-)

I do not want to self host either, for exactly the same reasons.

However, I do want to have full control of my secret data beyond the secrets themselves, ideally w/out self hosting, i.e. I want to have crypto-proven control over whom I'm sharing secrets with, I want to have have cryto-level assurance that the service cannot use recovery/escrow mechanism to unlock my secrets data stored on the cloud w/out my consent.

Apple Passwords comes closest to what I want, but it's not cross platform.