This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2023-09-15
Channels
- # announcements (15)
- # babashka (8)
- # beginners (23)
- # biff (20)
- # calva (6)
- # cider (9)
- # clerk (31)
- # clj-kondo (3)
- # clj-otel (2)
- # clojure (116)
- # clojure-argentina (1)
- # clojure-austin (5)
- # clojure-europe (64)
- # clojure-nl (3)
- # clojure-norway (23)
- # clojure-sweden (40)
- # clojure-uk (1)
- # cursive (16)
- # data-science (2)
- # datahike (8)
- # emacs (3)
- # events (1)
- # hyperfiddle (24)
- # malli (5)
- # off-topic (24)
- # re-frame (9)
- # releases (1)
- # solo-full-stack (25)
- # sql (18)
- # tree-sitter (19)
- # xtdb (10)
I really enjoyed this talk. I generally listen to this kind of thing while driving to and from work. Thanks for setting up the channel and the link to the video - which I hadn't seen before.
I’m sensitive to the fact that I’m creating a commercial app in a FOSS community. Once I have something working, I will publish the bits that I think will benefit the community.
A lot of us are using Clojure for a commercial purpose and only a relatively small number of people produce/maintain FOSS projects. I know it may look like "everyone" does FOSS here from the steady stream of #C06MAR553 and #C015AL9QYH1 but we have about 1,800-2,000 people here who are active in discussions so that should help put things in perspective 🙂
There's absolutely no shame in making a living with your craft, and not contributing or publishing open source things, no worries.
Thank you. The perspective helps. Of course, your contributions intensify my desire to contribute/reciprocate. I go further in debt every day. It’s a healthy debt. I’ll pay back, forward, or both somehow, someway.
I think something that could help me immensely on this channel and might be of interest to anyone here is being able to have some video call to show progress on what I'm making. This might help with some stronger sense of accountability to push progress a bit harder and eliminate some sources of confusion on the product or UX. I previously made http://aceplace.com and http://sayartii.com, almost entirely by myself, the latter of which gets on average of 5M requests per day and it just runs itself. I'm now working on a platform that's a combo of clojure/script, re-frame, malli, and datomic that has user posts and live chat targeted to solve an automotive problem in UAE. If anyone is interested in doing seeing my progress and doing some 1-1 calls with me, I'd love to show off the product and show off any interesting bits going on under the hood. I don't need any technical help, but definitely open to criticism and discussions on the way I'm doing things. Feel free to reply or DM me, I can entertain a call or two today if anyone is keen.
Going on a walkabout with you would be interesting to me. If you have some sort of minimum expectations of my experience, it’d be better to understand them ahead of time. If that articulation would take too much time, we can also just give it a go with the understanding that the conversation might be very short.
No expectations needed, just happy to show off what I'm doing. Going to have to start onboarding users soon anyway so talking to anyone would be good for me.
To me, https://clojurians.slack.com/archives/C05SVE3P116/p1694777264919879 is amazing. Asking for other’s eyes is brilliant. The very invitation changes the way we see our projects. Brilliant.
This buy-build dichotomy is gnawing at me. By “buy” I mean using libraries and frameworks, commercial or FOSS. Maybe just getting good at out-of-the-box Clojure is the way. Bradford Cross’s https://clojurians.slack.com/archives/C05SVE3P116/p1694737036638269 (build) settles in hard — and so does Sean’s friend’s https://clojurians.slack.com/archives/C05SVE3P116/p1694729341423259?thread_ts=1694712184.885969&cid=C05SVE3P116 (buy). It certainly extends to hire-DIY decisions. So often, fast is slow and slow is fast — and leverage is a thing. I don’t know how this will shake out. It will likely have to find its own balance. Rich’s https://youtu.be/m1tZEn_NAqg?si=8vDPpTpo0tVimw2j&t=153 is certainly a buy-build hybrid that arose out of seeing many different buy-build ratios.
My kneejerk reaction is that you want the thing you want to control in your own code. Then you can use ready-made stuff off in the peripheral vision, for problems you just want solved, and isn’t core to the experience you provide. (though I too find it hard to navigate that boundary. For example RSS feeds, should you find an RSS library or understand the Atom XML spec? Borkdude and Jack Rusher https://clojurians.slack.com/archives/CLX41ASCS/p1693746554325989?thread_ts=1693746239.511099&cid=CLX41ASCS and learning the Atom specification rather than some library’s interpretation)

for generating rss/atom feeds, for sure no need for a dedicated lib. learning the lib won't be any less work than just reading the spec anyway. for parsing feeds I'm much less certain. if you're building some kind of rss reader and thus will be parsing lots of random feeds, many of which will not conform perfectly to the spec, using an RSS library means you'll benefit from whatever quirks-handling it includes. I don't have direct personal experience on if parsing the feeds without an RSS lib is actually a problem in practice, since I've always used a lib (ROME on the jvm via the Remus Clojure wrapper, and feedparser in python). but on multiple show HNs I've posted, people bring up how parsing rss feeds is a pain because of spec non-conformance (my response is always "I just use a lib and it works fine 🤷"). if there's a single/small fixed set of feeds you need to parse, then doing it with a plain xml lib is fine, but even then, Remus turns the feed into plain data structures anyway.

Omar and I had a great huddle. His sense of cost control, care for quality UI experience, and commitment to simplicity are everywhere. He’s definitely build-first, buy-second. He made fantastic use of our time together. Taking someone on a walkabout allows us to see our stuff through someone else’s eyes. His TODO list grew throughout our visit. Another’s perspective can’t be conjured in isolation. Running solo can make such experiences uniquely effective. Teams can take each other for granted, turning precious resources into boat anchors. Good code reviews are irreplaceable. I can’t encourage walkabouts enough. I am left wondering about innovative safe space. It’s tempting to have some sort of NDA structure in place just so we are as free as possible. Law is not my favorite subject — and we live in the world we do. > WHAT TO DO WITH SECRETS > If you find a secret, you face a choice: Do you tell anyone? Or do you keep it to yourself? > > It depends on the secret: some are more dangerous than others. As Faust tells Wagner: > > The few who knew what might be learned, > Foolish enough to put their whole heart on show, > And reveal their feelings to the crowd below, > Mankind has always crucified and burned. > > Unless you have perfectly conventional beliefs, it’s rarely a good idea to tell everybody everything that you know. > > So who do you tell? Whoever you need to, and no more. In practice, there’s always a golden mean between telling nobody and telling everybody—and that’s a company. The best entrepreneurs know this: every great business is built around a secret that’s hidden from the outside. A great company is a conspiracy to change the world; when you share your secret, the recipient becomes a fellow conspirator. > > Thiel, Peter; Masters, Blake. Zero to One (pp. 103-104). Crown. Kindle Edition. I’d appreciate your thoughts.
I have a question. What do you use to track how many users you have? The obvious "free" choice is google analytics but then that corporation has the data in their database. Does anyone know any good alternative that doesn't take much time to operate?
I have a self-hosted https://plausible.io/ instance that I use for page views on my static websites. for my web apps I just have an admin page that queries the database to see how many signups there are etc.