Fork me on GitHub

Actually I haven’t had the coffee yet - still on the train. Thankfully the kid was relatively good last night, so I probably had 6 hours of broken sleep - but final presentation on current project bid today, so will need caffeine in lavish amounts


after a few beers last night at the ldn clojurians event, I need coffee also


Anyone using Mount/Integrant/Bounce over Component?


@jonpither didn't get a chance to catch up last night...full of cold. Maybe next time?


yeah defo 🙂


i did give a try to mount a while ago. liked the simplicity. but recently all microservices with aws lambdas mainly so no need to add any component stuff


@jonpither i'm using my own bounce-inspired thing instead of component


just doing some more work on it now after talking with @jarohen last night triggered some thoughts


@jonpither I've done extensive research into mount & the philosophy behind it. Because I was really confused before I did. Namespace vars?! For state?! ARE YOU MAD! What I got out of it is that mount doesn't complect state management & dependency injection together. You are a grown up, if you decide you want DI, you can do that through proper programming yourself (I get the impression that the proponents of mount actually use it this way). Mount is really fun, you don't need a special reset, you use your plain old tools.refresh (or just re-eval!) and it will only restart the components that have updated. This is quite a nice feature. Mount is definitely written about in this article by it's author (though not really mentioned until the end). Mount now supports re-binding in tests (I believe it just uses with-redefs under the surface or something similar). I've used it in pet projects. I think it's quicker to start with & with self discipline perfect for teams. Especially on projects where you catch yourself passing around state to 90% of your code base (cough cough it begins with l and ends with m_cough cough_)


just spent several hours finding the place where to add width: 200 in the code... I just love to computers...

Rachel Westmacott13:02:16

does anyone know of a good way to find unused requires/imports in a project?

Rachel Westmacott13:02:52

..but obviously thank you for the swift reply!


you’re welcome, any time 🙂


we just lost over an hour because the curl command we gave someone to run, had something like “curl blah blah ‘foo=bar’ blah"


… something somewhere changed the quotes to pretty quotes.


@korny heard a similar story where something got pasted into chat and added “invisible” non printable chars, so the code barfed with weird errors. The lesson: secret gist over cut & paste.


@peterwestmacott should be available in vim too...

Rachel Westmacott14:02:00

I’m on Cursive, and in fact asking the question has prompted me to rummage through the menus and find some pretty good stuff...

Rachel Westmacott14:02:37

I was hoping for more of an automatable solution, eg. a lein plugin


i thought cursive had this functionality tho


i am not really savy with cursive naturally 😉 so it does not have clean ns or similar?

Rachel Westmacott14:02:26

it lets you know when your requires and imports are unused by highlighting them differently in the editor

Rachel Westmacott14:02:17

and there’s IntelliJ functionality to run the inspection over the whole project and get a list of unused imports/requires (and a host of other warnings/errors)

Rachel Westmacott14:02:49

I’m not aware of a namespace tidy up function in cursive


@peterwestmacott have you tried eastwood? seems like there is :`unused-namespaces` linter. Never used this myself tho

Rachel Westmacott14:02:20

oh, that’s a good idea


you can also use refactor-nrepl from the repl if you really push it 😉

Rachel Westmacott15:02:28

alas eastwood’s :unused-namespaces appears to be broken.

Rachel Westmacott15:02:13

it claims that there are unused things, but they are clearly used.


time to switch to emacs then troll 😉

Rachel Westmacott15:02:57

well yes, I’ll put that on the list with all the other reasons...

Rachel Westmacott15:02:14

for now Cursive does a very nice and reliable ad-hoc analysis though


is there a scenario where “smart” quotes improves understanding over normal straight ones?


perhaps if you had nested quotes it would allow disambiguation of start/end ? though the convention seems to be also to alternate single/double quotes for that case -