sysguru2046

5 hari yang laludiscuss

Superbase?

sysguru2046

10 hari yang laludiscuss

Yes. I like the outcome generated by the skill, and it triggers me to learn more terminology that I never heard of. I want to try more for public trading companies. If this skill can be integrated with like 401(K) investment research, it'll be great. Does some of investment companies like Fidelity using such SKILL background?

sysguru2046

19 hari yang laludiscuss

I figure it out. I saved the project to ~/.agents/skills and use Codex /skills and choose Peter Lynch Skill for a quick research.

sysguru2046

19 hari yang laludiscuss

I downloaded the project to my local laptop and let Codex to use downloaded project. Is it a right approach to use the agent skill?

sysguru2046

28 hari yang laludiscuss

I'm very interested in your open-source Shopify project for e-commerce.

Do you consider Event-Source Commerce as part of this project?

Examples of Commerce events:

Catalog events - Product created, price changed

Inventory events - Quantith change

Order events - order created, cancelled

Routing events - order received, routing decision made

Shipment events - shipment created, tracking assigned

Customer events - Customer registered, address changed

Supplier events - Feed arrived

Taxonomy Events - Category created

sysguru2046

30 hari yang laludiscuss

Glad to hear it can support large file system. Is ZeroFS stable with:

Millions of files, Interrupted transfers, Backup Server Reboots, Long-term operation?

sysguru2046

30 hari yang laludiscuss

I assessed the tool on a CentOS 7 system today. It can't run on older CentOS 6. It will generate core dump.

My testing went well. I was able to mount ZeroFS as NFS and write a 100M file to the mount point. The file persists after I deleted ZeroFS cache.

My plan is to use it for backup process. If we don't have such tool, we would need to prepare a large Filesystem stage area, then run rclone to upload to cloud space. With this tool, I can mount it on Backup server, rsync file server a large Filesystem to the ZeroFS mount point, eliminating stage area.

Will ZeroFS support large Filesystem size about 800GB?

sysguru2046

2 bulan yang laludiscuss

Generated spec is 542 lines including 7 sections: - Requirements - Event Schemas - Checkpoint File Schema - Resume Algorithm - Failure handling - Acceptance Tests - Implementation Tasks.

It inspects my existing repo, and identified existing logic, then imported my existing functions, and created a checkpoint_manager.py.

The new function adds additional functions like: - update_started - update_completed - update_failed - install_shutdown_handlers - restore_shutdown_handlers - persist_checkpoint

Below points are generated in Implementation Tasks:

3. Implement checkpoint model: - typed checkpoint structure; - schema version validation; - bounds validation for next_url_index. 4. Implement atomic checkpoint writer: <- "atomic" - write temporary file beside final checkpoint; - flush and fsync; - atomic rename; - preserve old checkpoint on failure. 5. Implement checkpoint loader: - handle missing file; - reject malformed JSON; - reject unsupported schema version; - reject URL fingerprint mismatch.

I don't fully understand what "atomic checkpoint writer" means, but the LLM understands what it means. The functions are likely to implement "atomic checkpoint writer".

The key takeaway is that tokens generated by human usually not restrict enough. When I thought about "checkpoint", I wasn't thinking of any "atomic" shape of checkpointing. I was using my casual language model to generate first version of implementation first. Then using Spec-Driven-Development model, the generated spec has better understanding of what it is. There could be many mutated forms of real life implementation of "checkpoint", the LLM generated code is mostly restrictive one that I will never know in my whole life experience. The LLM training must have undergone very restrictive learning process that I never experienced.

sysguru2046

2 bulan yang laludiscuss

I cloned your Spec-Driven-Development repo to my GitHub repo, use codex to generate a scraper checkpoint manager spec, then use codex prompt to implement the checkpoint logic. The new implementation has significant improvement on top of my original shape of checkpoint logic. It's strict implementation based on generated spec.

comments