🫖 Not to spoil my forthcoming article and interview with @Nathen Harvey BUT oh DANG 2024 DORA found that platforms can increase burnout and instability?! 🍵

increase or decrease? 🙂

Not a typo @Jordan Chernev

“Platforms” like actual platforms, or platforms like “I need to hit my target so let’s slap backstage on our pipelines” target? 😁 Looking forward to the full article!

@Sam Barlien I mean IDPs in general though they aren’t sure exactly why. Article I think will come out Monday but the 120 page DORA 2024 report is already out!

need to throw that 120 pager in chatgpt to give me the gist

Burnout by platform teams or their users (eg, developers)?

i’m not sure about instability but the burn out part definitely resonates with me.. extremely challenging to be the team providing a platform.. keep in touch with users, adding new features, fixing issues safely, working with platform partner teams

i read the DORA report now and the summary is a bit sensationalized… are you really doing platform engineering if you are forcing your users to use it? no wonder that will increase burnout

Platform engineering not done right = burnout, cognitive overload,headaches, resignations, etc etc for all involved - Devs and DevOps engineers ... basically a total mess.

mostly a thought - how's that different from any other discipline?

Platform engineering not done right...

The problem is, outside of our bubble, it isn't being done right. More than 50% of IDPs are mandatory... which is like Platform Engineering 101 stuff being totally ignored. The time it takes to get from "great idea" to "totally oddball real-world versions of the idea" is getting super-short. You used to have a clear run of a decade before the "broad but thin adoption problem".

or could it be by the time a platform gets put into place most engineering team are already burned out, from what a lot of what I have read it seems like putting a platform into place is a solution places come up with as a last hailmary when things are already disfunctional

Here’s my exclusive with @Nathen Harvey on why platform engineering (and less surprisingly AI) are making software LESS stable and slower to release! https://thenewstack.io/dora-2024-ai-and-platform-engineering-fall-short/

Doc quality and code quality makes perfect sense, llms are really good at that, code review not so much because it's hard to get the entire repo context correct ( personally I think llm code reviews don't make sense, it should start at the ide level ), technical debt also makes sense, llms are really good for juniors, for senior level users their better for scaffolding and low level tasks at least for now - at least this is what I have found from my own use and our engineering team ( no you can't use llms to replace your senior devs 😂 )

I heard at a lean coffee last week that amateurs look at LLMs as an expert and experts look at them as amateurs so the first is dangerous

That makes sense, very confident experts 😂

AI is generally a better fit for net new things or one-offs

It's not bad bad for small code explanations but again for bigger codebases it's hard to get context

as a "senior", i use it as a starting point of clay that i subsequently model to a pot / bowl / what have you

Interesting, so Platform Engineering is having a positive impact on individuals/teams, but a negative impact DORA metrics. thinking

@Shane Dowling they perceive they are more productive

Ahh... 🤔

Are you using AI in your platorm engineering @Jordan Chernev?

when i was doing platform engineering, yes