Advertisement
Guest User

Untitled

a guest
Jan 27th, 2020
100
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 6.23 KB | None | 0 0
  1. A vision from the Ghost of Urbit Future:
  2.  
  3.  
  4. Ask HN: How come nobody uses my great side project?
  5. 78 points by y33th4x0r June 6, 2021 | flag | hide | 130 comments
  6.  
  7. I used to be one of the proud minority of people who turn off the lights
  8. when they step into the shower. However, I recently had a flash of
  9. inspiration so powerful I jumped up and down in excitement and joy,
  10. disorienting me and causing me to fall. Conveniently, this gave me a
  11. week of sick leave, which I used to hustle my idea into a prototype.
  12.  
  13. I've done everything I'm supposed to do: my project fills a real need
  14. for myself, it doesn't scale, and I've been spending every spare moment
  15. I have going directly to users who claim they have the same problem I
  16. have. I've found a real human problem, and I wrote the code to fix it.
  17. Why am I not rich yet?
  18.  
  19. ---
  20.  
  21. ^ patio11 5 hours ago [-]
  22. v
  23. Sounds like you found a real problem and have built a prototype, but
  24. it's just not easy enough for people to integrate into their lives.
  25. They need to be able to sign up, maybe download an app, and go. You
  26. need to spin up servers and store their data so you can sync it
  27. between their devices, and you need to make an app so you can send
  28. them notifications and track their usage. Of course, for collecting
  29. all this data to be legal, you need to work with a lawyer, and since
  30. you're going to need to hire a mobile developer and dev ops guy,
  31. you'll need to set up a company. Then you'll need an office, and
  32. probably an office manager as well. Your best bet is to pitch some
  33. investors and see if you can raise a million or so, that's probably
  34. enough to test the waters.
  35.  
  36. ^ y33th4x0r 4 hours ago [-]
  37. v
  38. You're probably right that it's not easy enough to integrate into
  39. your life. In truth, I don't even use it much. I should quit my
  40. job and spend a couple years trying to get this off the ground.
  41.  
  42. ^ mnemnion 5 hours ago [-]
  43. v
  44. When you're ready: urbit.org
  45.  
  46. ^ js4lyfe 3 hours ago [-]
  47. v
  48. Yes is 0 and no is 1? Hard pass.
  49.  
  50. ^ y33th4x0r 4 hours ago [-]
  51. v
  52. Never heard of this before, is it performance art? I'll take a look.
  53.  
  54.  
  55.  
  56.  
  57. How using Urbit turned my side project into a multinational sensation
  58. 390 points by y33th4x0r July 14, 2021 | flag | hide | 98 comments
  59.  
  60. Here's an update to [this] Ask HN post:
  61.  
  62. With this prototype, I could tell I'd solved the technical part of the
  63. problem, but the distance from there to a product that was pleasantly
  64. integrated with my life was a long way and was capital-intensive. My
  65. wife told me I couldn't quit my job to commit full-time work to the
  66. project, so I chose an easier option. If I knew how well it would work,
  67. I wouldn't have considered any alternatives.
  68.  
  69. My prototype was a little python script that stored the data locally in
  70. a csv file (I meant to switch that to sqlite for robustness eventually).
  71. You had to go run it when you wanted an update (I thought about running
  72. it as a cronjob and emailing the results, but I never got that hooked
  73. up), and of course it only worked on my desktop.
  74.  
  75. I'd never used Urbit before, but I was told I could use it to add the
  76. quality-of-life features and robustness I needed. I installed it on my
  77. computer and phone and started reading the API docs. First thing I did
  78. was learn how to "poke" the notification service, which popped up a push
  79. notification on both my desktop and phone. I was worried might be too
  80. loud for some users, but I looked at the notifiation app in Urbit, and
  81. there's settings for which devices to pop up the notifications, and
  82. whether they should be persistent. For very little work I was able to
  83. use a well-designed push notification system that worked on all my
  84. users' devices. Nice.
  85.  
  86. Every once in a while my csv file would get corrupted, and I hadn't
  87. gotten around to putting that in a database or something to be more
  88. convenient. I realized that I could just put the data in Urbit, and it
  89. would keep it around for me with database-like robustness. This
  90. required me to define the data type in Hoon, which was a little scary,
  91. but it turned out to be pretty easy to copy an example and modify it for
  92. my needs.
  93.  
  94. Some of my users wanted to have shared pieces of data, and they were
  95. tired of emailing CSVs to each other. Plus, once I put the data in
  96. Urbit, there were no CSVs. Reading further in the docs I realized it
  97. was trivially easy to send the data directly between the different
  98. Urbits, so I added a button for that. Someone showed me the "groups"
  99. app, which meant they could even share their data among a group of
  100. people. Then they started chat rooms connected to those groups, and
  101. they also started writing about what they're doing on the "publish" app.
  102. I didn't have to do anything to support that since they could just use
  103. the Urbit interface, though I did add a little chat area to my front end
  104. that would check for an associated chat group and let you chat right
  105. there. Mostly this was useful for discovering the chat room.
  106.  
  107. Users loved it. Turns out some of the people I was talking to already
  108. had Urbit, and it was really convenient for them since they didn't have
  109. to add anything special. Eventually, some people started spamming the
  110. groups, but it's pretty easy to ban people on Urbit, so that fixed
  111. itself pretty quick.
  112.  
  113. Urbit let me take my prototype and, in my limited spare time,
  114. incrementally build something that people loved with little overhead.
  115. After writing the initial prototype, most of my work was learning about
  116. Urbit and reading its docs so that I could find all the features I could
  117. add.
  118.  
  119. My next step is to learn a bit more Hoon so I can translate my algorithm
  120. into Hoon. This will make it easier for people to run -- they can just
  121. install it from Urbit, and they don't have to have it running separately
  122. on their laptop.
  123.  
  124. ---
  125.  
  126. ^ yosoyubik 5 hours ago [-]
  127. v
  128. Nice going! This mirrors my experience.
  129.  
  130. ^ Retric 5 hours ago [-]
  131. v
  132. Dude, you could have just spun up a postgres database and a Kubernetes
  133. cluster, added "Sign-in with Google" and reCaptcha, designed a little
  134. notification system and chat and groups and whatnot. I've been
  135. running my own mail server for years, and yeah it's a bit isolating
  136. that nobody receives my messages, but the technology works just fine.
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement