Today, I learned that the quickest way to get to all tasks in my Things3 todo app is saving a *.url file with a link. I open the tag view in Things3, click on three dots, click on "Copy link" and paste this after URL=. This is a workaround for not having saved searches or bookmarks in Things3.
[InternetShortcut]
URL=things:///show?id={id}
OpenCode with Deepseek Flash v4 (and other non-OpenAI/Anthropic models) do work reliably in code bases now. I have been trying every month as experiment and now they don't stop midway or create strange results, but produce good enough output. I feel prepared for price raises and political cut-offs with the new options!
When Derek Sivers has an issue, he follows this process:
In a more detailed description by Derek:
I have three mentors. When I’m stuck on a problem and need their help, I take the time to write a good description of my dilemma, before reaching out to them. I summarize the context, the problem, my options, and thoughts on each. I make it as succinct as possible so as not to waste their time. Before sending it, I try to predict what they’ll say. Then I go back and update what I wrote to address these obvious points in advance. Finally, I try again to predict what they’ll say to this, based on what they’ve said in the past and what I know of their philosophy. Then, after this whole process, I realize I don’t need to bother them because the answer is now clear. If anything, I might email to thank them for their continued inspiration. Truth is, I’ve hardly talked with my mentors in years. None of them know they are my mentors. And one doesn’t know I exist.
Source: https://sive.rs/ment
When the room was asked who felt they had agents operating at scale, no one raised a hand. In a smaller agentic workflow breakout, only 5 of 25 said they had agents in production at all. The concerns were consistent: security everywhere, not just in the CISO office; ROI that is still hard to measure; legacy systems that must be modernized before AI can truly scale; unclear governance and ownership models; and rising questions around vendor claims, data readiness, build-vs-buy decisions, and agent lifecycle management.
Pretty no company at all has active agents in production. And for the very same reasons that IT systems before did not get live fast.
Ed Sim, via What’s 🔥 in Enterprise IT/VC #497
Capture your thoughts on Mac into one file with the Burst Notes app by the maker of the productivity app Vitamin-R. Stays out of your way, hence similar to my open source app Quick Capture. Difference: Burst Notes saves into single files while Quick Captures saves to multiple files, either daily notes or Zettelkasten notes. I've been a long-time user of Vitamin-R and it is super reliable and gets frequent updates. Burst Notes is a pay-once app, 15€.
Repeat after me: most software spends the majority of its time in the maintenance phase.
Frontend is never right, it will always look and behave different from what you expect.
This is already true for one person, but 2+ people using the same interface will satisfy all needs.
Text is much more forgiving here, it's asking for a result rather than a representation.
Text is declarative like SQL, you tell what you want.
So text as interface is flexible as it will use the words to explain that were used to ask. Problem then is how to make result "repeatable" and "reliable".
My idea right now is to skip frontends as often as possible and not hardwire data and presentation but to give the leanest tools possible.
So it could mean for a personal CRM to have a database, APIs, a Bearer token in your ENV and a skill that your coding agent can use to interact with your CRM.
Benioff’s push into agents could put Salesforce at existential risk because once the front-end gets reimagined with AI, the application becomes just a database and customers migrate to cheaper, faster, more agile systems.
Applications might all become a (hosted) database with a thin layer of APIs that AI agents can interact with via text.
Now, everything starts and lives in an LLM (Claude Code / Cowork / Codex / etc.) and the tools that used to “own” a step in the process are now plugins into your workflow.
Text input in LLMs is the new interface that swallows all other products.
Source: metedata #006
Between stimulus and response there is a space. In that space is our power to choose our response. In our response lies our growth and our freedom.
Unclear origin via QuoteInvestigator