SendLang: A DSL for Email Automation | Hacker News Reader

SendLang: A DSL for Email Automation

sendlang.com

15 pointsksajadi11 comments

Get me the inverse, I am so done with cold outreach spamming my inbox. We need less garbage automated low effort crap.

I agree that a lot of marketeers send spam but that doesn't mean email as a category is just bad and should be blocked. I am all for anything that can make better use of email as yet another tool in our pockets be it marketing or just run off the mill transactional emails we all send from SaaS (password resets and whatnots).

Is this peak AI?

Turning some email sending library, already included in many modern languages, into 2(TWO!) new languages.

I don't think this is about that. A lot drip campaign and segmentation tools are based on WYSIWYG editors or UI based flow editors, which puts them out of reach of agents. This turns those into agent friendly workflows. The app will send the email however it sends it before: via SES or whatever via your own code.

Maybe I'm really dense, but this is just instructions for agents on how to write code, whatever that code looks like.

What is the difference between asking an agent to write these new two DSL's and asking an agent to create or customize a django template, or whatever?

I know the job market is tough, but devs are not working on the user side of WYSIWYG email editors. They were creating templates before, and they are doing it now, why would they use this?

These DSLs are not there to replace templates. They are to define email lifecycle: those are different things. Email lifecycle dictates "who" gets an email (ie anyone with a Pro plan) and "when" they get an email (after they upgrade to the Pro plan). Template defines "what" they get (the email content etc). Those are different things.

You can do the "what" part (template) via WYSIWYG or django template or ReactEmail,... Sendlang is the equivalent of your django code but for the lifecycle part.

I don't see how that makes any more sense. This is some gastown polecat slinging I can't understand :D

yeah that’s pretty much what i thought as well, this rhymes with the post earlier today about DSLs working well as harnesses. i get the logic but i find it depresssing that even language design is now not just AI aware but including it as a core use case.

Don't we ready have mail sieve

Sieve is for filtering (mostly dispatch and spam blocks) but admittedly it can be used for contact segmentation as well (similar to SendQL). However it doesn't have any workflow (drips for example) idioms.

This slop is disgusting

Reply on news.ycombinator.com