luke_redroot

Approximate knowledge of many things.

[Isaac King 🔎 @IsaacKing314]

Javascript is light years ahead of every other programming language.

[Babble @Babble____]

12 people have been to the moon and only 8 people have won Takeshis castle. Really makes you think

[Charlie Holtz @charlieholtz]

David Attenborough is now narrating my life

Here's a GPT-4-vision + @elevenlabs python script so you can star in your own Planet Earth:

[matt swanson 😈 @_swanson]

Sometimes I write a TODO comment when we are waiting for a library to release a new version to fix a workaround or patch

But even better is a test that will fail when you've upgraded the library so you remember to go back

[luke_redroot @luke_redroot]

@cokofreedom However, we only have so much agency in this respect, and having that lack of agency all other our board, in the first column we check, is way more demoralising than the approach you suggest

[luke_redroot @luke_redroot]

High WIP / Kanban strategy q: working in a team where we work with many 3rd party suppliers, tasks often involve investigating our side, submitting a ticket to them, waiting 2-3 weeks for a resolution, then picking up the work again. Makes even JIY pull based priority tricky 1/2

[luke_redroot @luke_redroot]

@cokofreedom That’s a good plan, less disruption to our day to day board. Additional dimension however is that most the tasks are driven by a drop in our KPIs, dues to issues / bugs with the suppliers systems or upstream systems to them. So there is some urgency to follow up if they need info

[luke_redroot @luke_redroot]

@cokofreedom I think that might be better, I think the fact we have it on our board every morning as a card we’ve committed to investigate and fix is the problem, there’s a risk we wouldn’t prioritise as quickly once we can pick it up but that’s happening anyway

[luke_redroot @luke_redroot]

@samwhoo Yeah that’s a good idea, unfortunately not predictable at all, trying to negotiate SLAs but no luck so far

[luke_redroot @luke_redroot]

@samwhoo We can do that, and that’s our currently strategy, I guess my annoyance there is the cycle time for these cards is so long already, anything we can control to help without trashing other cards. Perhaps just the nature of the work

[luke_redroot @luke_redroot]

@samwhoo Better SLA on resolutions would be great, but generally these are middlemen suppliers who integrate with the companies who provide the data / assets we need, so extra hop required to resolve anything sometimes oo

[luke_redroot @luke_redroot]

@samwhoo So far my ideas are a) make WIP limit a function of team size plus number of cards blocked in this way, so when stuff is unblocked there’s less queue time to get picked up, good for time to value or b) have a discovery stream before delivery (useful anyway), this work sits there

[luke_redroot @luke_redroot]

@samwhoo Ultimately, tasks getting blocked by us or others once we commit to them is a pain, something you want to avoid in a pull based Kanban system which I like, as it’s a planning overhead to have work arrive in multiple ways, plus …

[luke_redroot @luke_redroot]

Should be JIT not JIY in first tweet *

[luke_redroot @luke_redroot]

As tickets can become unblocked at any point, and often then falls behind work we’ve started even though it’s already committed too. Other work that doesn’t involve suppliers is fine and has decent cycle time. Often this leads to high WIP as feel obligated to either 2/3

[Sam Rose @samwhoo]

Do I know any other programmers who have special needs kids in the UK? I’d like to chat with you. I think there’s good that can be done in this space.

Retweets appreciated 🙏

[Matt Fuller @matthewdfuller]

"It's called Kubernetes, and Joe, get this, it'll fix basically all of our problems. I haven't dealt with it myself, but I did read about it in CIO magazine and immediately told all our developers to start migrating. Apparently you just give it a container and it handles everything for you. You don't need to worry about security or scaling or anything else."