3 ms·
Is news.YCombinator an interesting test bed for software startup philosophies?
- juwo 20y agoObserving this newsgroup, I see at least a few things in this experiment: 1) Release Early, Release Often. I am certain the programmer who wrote it is a better, more intelligent, more competent programmer than I. IMHO there are so many irritations with this software because it was released prematurely. Even Paul Graham says it is bare-bones. Yet people keep coming back. In spite of it. It's because of the benefit of chatting at a high quality watering hole for entrepreneur elephants. 2) The UI is very different from the traditional newsgroup interface. If you are making people do things differently, then the benefit should be extraordinary (Weinberger?). I am realizing this for myself. (See http://juwo-works.blogspot.com/2007/03/i-need-your-advice-startup-failing.html) 3) Did they have to write this from scratch - was it the right strategy. Couldnt they have built on top of some open source newsgroup code?
- volida 20y agoi think the answer to (3) is that they wanted to test Arc...the web server is written in Arc also, if I am not mistaken...
- dfranke 20y agoI haven't read that the webserver is in Arc. But Paul said on Reddit that the current implementation of Arc translates it to PLT Scheme. PLT has its own Tomcat/Webrick-ish HTTP server, so they probably use that.
- juwo 20y ago4) Is the technology more important than the application? Isn't technology (here, Arc) merely a vehicle for the application?
- juwo 20y ago
- danielha 20y agoYou answered your questions for 1 and 2. volida is right with number 3 -- they were also looking for something useful to build with Arc.
- juwo 20y agoI didnt really have a question - I am merely observing the ongoing experiment! Any other principles you can observe?