←back to thread

Shift Left Is the Tip of the Iceberg

(semiengineering.com)
179 points yatrios | 7 comments | 19 Nov 24 14:06 UTC | HN request time: 1.03s | source | bottom
Show context
0xbadcafebee ◴[19 Nov 24 15:15 UTC] No.42184298[source]▶
For those not aware, Shift Left[1] is (at this point) an old term that was coined for a specific use case, but now refers to a general concept. The concept is that, if you do needed things earlier in a product cycle, it will end up reducing your expense and time in the long run, even if it seems like it's taking longer for you to "get somewhere" earlier on. I think this[2] article is a good no-nonsense explainer for "Why Shift Left?".

[1] https://en.wikipedia.org/wiki/Shift-left_testing [2] https://www.dynatrace.com/news/blog/what-is-shift-left-and-w...

replies(13): >>42185611 #>>42186878 #>>42187331 #>>42187375 #>>42187393 #>>42187419 #>>42187454 #>>42187463 #>>42187501 #>>42188834 #>>42192801 #>>42194403 #>>42207068 #
leoc ◴[19 Nov 24 17:01 UTC] No.42185611[source]▶
So “shift left” is roughly equivalent to “tests first” or “TDD”?
replies(5): >>42185876 #>>42185932 #>>42185960 #>>42186122 #>>42188425 #
1. hitchstory ◴[19 Nov 24 17:40 UTC] No.42186122[source]▶
That's just one example of it.

Other examples:

* Replacing automated tests with (quicker) type checking and running it on a git commit hook instead of CI.

* Replacing slower tests with faster tests.

* Running tests before merging a PR instead of after.

* Replacing a suite of manual tests with automation tests.

etc.

replies(2): >>42186754 #>>42187010 #
2. 0xbadcafebee ◴[19 Nov 24 18:41 UTC] No.42186754[source]▶
Or, again more generally:

- implementing security features earlier (DevSecOps)

- implement tracing and metrics/analysis tools earlier, use them to test and debug apps earlier (as opposed to laptop-based solutions)

- building the reliable production model earlier (don't start with a toy model on your laptop if you're gonna end up with an RDS instance in AWS; build the big production thing first, and use it early on)

- add synthetic end-to-end tests early on

The linked article is talking about Shift Left in the context of developing semiconductors, so you can see how it can be applied to anything. Just do the needed thing earlier, in order to iterate faster, improve quality, reduce cost, ship faster.

replies(1): >>42191133 #
3. js8 ◴[19 Nov 24 19:12 UTC] No.42187010[source]▶
So aside from automation (which also has tradeoffs), when you shift left, what gets shifted right?
replies(1): >>42187322 #
4. saynay ◴[19 Nov 24 19:43 UTC] No.42187322[source]▶
The knowledge needed to do the 'shift-left' tasks. The downside of making your developers take on more tasks that used to be handled by someone further down the chain is they need to know how to do those tasks.
replies(1): >>42187451 #
5. hammock ◴[19 Nov 24 19:58 UTC] No.42187451{3}[source]▶
Why can’t you shift the “someone further down the line” left along with the task?
replies(1): >>42188527 #
6. ElevenLathe ◴[19 Nov 24 21:56 UTC] No.42188527{4}[source]▶
This is interesting. Seems like it would mean something like having QA or infosec pair program with app devs. I'm sure somebody does this and am curious how it works out in practice.
7. skybrian ◴[20 Nov 24 05:50 UTC] No.42191133[source]▶
Yes, it can be applied to anything, but might not pay off with software like it does with semiconductors? Better tests and more static analysis are useful, but teams will often work on making their releases faster and easier, iterating more frequently, and then testing some things in production. We don’t want our software to be too brittle, so it should tolerate more variation in performance.

But with chip design, they can’t iterate that fast and performance is more important, so they are doing more design and testing before the expensive part, using increasingly elaborate simulations.