zemptime.bsky.social
@zemptime.bsky.social
This might just be a personal thing, but I really don't like rspec's shared behaviors/shared examples most of the time.
Find it makes specs really hard to understand, obscures relevant line #'s on failure, and generally adds a layer of anti-useful indirection
February 26, 2025 at 8:38 PM
One technique which seems to be working well for me is specifying a source requirements section. the bullet points after may vary, but this is always the first one:

Source Requirements:
- Minimum 3 independent sources for major findings
This is a pretty good guide to why people may misunderstand Deep Research, and what it is actually very good at & what it is not from @afinetheorem.bsky.social
February 9, 2025 at 3:00 AM
Been using cline for a little bit now. it can write some code!

the profession of software development is going to (slowly?) shift from directly building systems to building systems which build systems.

now to begin the search for approaches, techniques, and principles beneath it all...
February 9, 2025 at 2:43 AM
Just listened to @syntax.fm's episode on state management. Wasn't all that satisfied with the definition of state, but also, couldn't think of anything better. Until... it just hit me.

'State' is information in your application which could have been something else.
November 28, 2024 at 6:12 PM