They don't mention a twinkle that many task runners seem keen to omit: how do you handle things where there are human steps involved and not everything is automated? How do you track what has worked and what is still left to do if things go sidesways?
I built baker (https://github.com/coezbek/baker) for this some time ago (pre-LLM mostly). It uses markdown with embedded bash and ruby commands to give you a checklist which is run both automated for commands or with human in the loop for things which aren't automated (like login to some admin panel and generate that key, copy it here).
The checklist gets checked off both by human actions (you confirm that you did it) and automated e.g. success bash command runs. So you keep a markdown artifact on where you are in your project and can continue later.
You can wrap commands to run via SSH (of course clunkier than what scotty here does, but you can select a port for SSH).
I already commented about Expect elsewhere in this thread, so I should probably pipe down, but thought it might be worth it here as well because Expect has been handling these kinds of exceptions and control transfers/flows with the full power of a robust programming language for decades. You might have a look at it for ideas and inspiration.
Looks neat, thx. It is really interesting how many facets the same kind of issue can produce. I think using expect from baker is a great way to deal with some steps that are hard to automate otherwise.
This is such a neat idea. I am going to adopt this for my own workflows as well, right now I just write private blog entries for stuff I do that I may forget how to do later (provisioning a server, networking, caddy setup, etc etc)
We need a term like potempkin-ware or something to express "I just built a 3 week project in 3 hours and, although it looks nice, there's probably a ton of problems with it because I couldn't possibly review everything Claude puked out properly, use at your own risk".
Ansible exists because it makes things idempotent, which is great when you have to do a thing on 1,000 servers because you can just fix the role and re-run it.
Bash can be idempotent but isn’t by default, so you either spend time making and idempotent bash script or you spend time learning Ansible to accomplish the same thing in a reusable way
Ansible’s idempotency is dependent on the specific module being invoked. What ansible mainly brings to the table is the parameterized modules. Which brings us back to people adopting it because they don’t know how to compose one liners, quote them properly, and wrap them in a for loop.
Many years ago I wrote my own "cloud instance bootstrapper" that would pull a tar off of S3 based on EC2 instance tags / metadata, untar it, then run a script. I never got into Ansible and I hated having to rebuild AMIs for minor changes.
I founded and developed a similar concept many years back of a web-based SSH dashboard and management console (Commando.io; which I sold). Now a days I use Semaphore UI [1] which uses Ansible playbooks under the hood in my homelab. Pretty happy with it, though setup and configuration did take a bit to get up and running.
I don't fully get the negativity here. This seems like a middle ground between quick'n'dirty bash script and a well-crafted Ansible playbook.
Half the time if you want to do something quick'n'dirty in Ansible playbooks you need to use shell anyway..
I participated in a hackathon recently where my deployment process was just a bash script doing scp/ssh to a remote server and it feels like Scotty would fit well to that kind of use-case.
I've been writing my own "task runner" which seems to have some of the same features. I'd say some pros: A nice view of that has run (what has failed, etc.) - which otherwise could be drowned-out by stderr and stdout. Timing information for each "task". Can organize nested tasks. Save all in a structured log.
I built baker (https://github.com/coezbek/baker) for this some time ago (pre-LLM mostly). It uses markdown with embedded bash and ruby commands to give you a checklist which is run both automated for commands or with human in the loop for things which aren't automated (like login to some admin panel and generate that key, copy it here).
The checklist gets checked off both by human actions (you confirm that you did it) and automated e.g. success bash command runs. So you keep a markdown artifact on where you are in your project and can continue later.
You can wrap commands to run via SSH (of course clunkier than what scotty here does, but you can select a port for SSH).