You could argue that using one-off purchases rather than subscriptions is leaving money on the table. But if I’ve got people using Cheyenne for just a couple of weeks and then canceling, multiple times, this pollutes churn statistics. And I think it’s betyer to be up-front with users about how the service will likely be used, instead of trying to get them to sign up for Yet Another Subscription.
For other use cases, a subscription does make sense because you’d be actively monitoring your site with Cheyenne. Maybe not every day, but certainly a couple times a month, for the indefinite future.
Indeed, “subscription” might not even be the right word for what I have in mind. Basically, it’s a one-off purchase. You get a seat, but just for, say, a month. No renewal, though you can of course come back again and use Cheyenne for another month later.
The idea is that, for some use cases, it makes sense to use Cheyenne for, say, a couple of weeks to get your site in shape; after that, you don’t need it. This is a scenario where auto-cancellation might make sense. You may want to just use the service once or twice a year, and that’s it. No need to get locked into a monthly subscription you don’t use.
For Cheyenne, I’m thinking of building in some logic for auto-canceling subscriptions. If you don’t use the service for, say, 90 days, you’re wasting money. Your account doesn’t get deleted, but your subscription gets cancelled. A bit of email automation can help survey the customer, find out why they’re not using the service.
Really enjoying lazygit. Magit in Emacs—my main tool for using git— seems to be suffering from a bug in a recent release, so I fired up lazygit just to see what it is. Pleasantly surprised!
In my view, validation is one of those topics on the web that bores most people, but actually gets me fired up. I can think of all sorts of ways validation can be improved.
If there are some pains you've encountered with HTML validation, reply here or DM me. I can think of all sorts of possible features, but I would rather launch Cheyenne in a minimal form before adding features few people care about.
If you do HTML, let me know what you think about markup validation. The first version of Cheyenne will be pretty bare bones, but takes on a couple core pains I've had with validation:
(1) how to validate a whole website (not just a single page), and
(2) ignore certain validator messages.
I'm about to launch a new HTML markup validator I call Cheyenne. It's made for web folks of all kinds. I've been working on this for a while and am excited to get it out there.
* Racket is a modern lisp and a descendent of scheme. The racket distribution includes an incremental native-code compiler*, the Racket language, Typed Racket, Datalog & several other languages, an IDE**, documentation and a variety of other tools. The distribution is suitable for new learners, application development, or language design and implementation.
** Racket works with a variety of editors, and has an LSP implementation
#emacs Book mode is an attempt at offering an interface for reading org-mode files with a clean layout. It is based on the use of large margins, headline and mode-line, with leading stars and bullets in the left margin using right alignment and different symbols. All decorations are part of the header-line (including upper left colored part and right upper glyph) and messages are displayed in the mode-line. Code at https://github.com/rougier/book-mode
@mattmight Doing anything in Racket these days? Racketfest, the little Racket conference I organize, is coming up in March — we’d love to have you! (It’ll be in Berlin, far from AL, but I thought I’d ask.)
Racket enthusiast, shell scripter, math nerd. JS hacker, theorem prover. Web. Compilers @ Igalia
A Mastodon instance for programming language theorists and mathematicians. Or just anyone who wants to hang out.