Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

https://en.wikipedia.org/wiki/Real-time_text

Back in the day, to chat with a coworker you and him would connect into the same unix box and do `talk <username>`[0]. You didn't have to press Enter to send a message, the other person could see what you were typing on every keystroke. Like in Google Docs.

I probably only like this idea because it's different. The wikipedia article mentions that older versions of the ICQ messenger used to have this feature. I have no idea why they removed it but I'm guessing they it's because it was a useless source of complexity that no one was using. Another counterargument is that Google Docs actually stores all the information you need to play back the creation of a google doc[1] and no one really seems to use it besides teachers checking for plagiarism. This could just be because no one knows about it.

But it's still surprising to me that none of today's chat clients do this or even have it as an experimental option. You could use peer-to-peer UDP for real-time text and only provide delivery guarantees using your centralized service after the user hits enter.

It's difficult to describe, but I noticed chatting with people in this way leads to more abstract conversations, so you'd probably have a hard time adopting this with people that don't already trust each other. It's also frustrating to see how slow people type and everyone types slow because you're comparing it to speech.

It could be useful in situations where you need answers as fast as possible (think first responders), but it would be such a tiny improvement.

You could generalized this idea to any piece of human generated text. It would be cool if instead of just the finished text of a book we also had a list of every key stroke the author and editor typed and their associated timestamps. It would be like open sourcing a small part of their creative process. Such a fine grained revision history might let you clear up some ambiguities in the finished text. Beyond that, it's probably useless for creative writing since composing the book would become an artistic performance, which might take a toll on creativity. I still think about it from time to time. Storage is cheap, why not collect the data now and see if it becomes useful later.

We could have a Hacker News where you could watch me composing this comment, like in a Google Doc. And play it back after I had posted it. It wouldn't provide any value, but still, why not? We have the storage space. Why not have it as an option? For those who want to be a bit more "transparent" or just want to watch themselves type after they're done. I'd use it. Surely there's no drawback (besides performance/scalability and space) for simply giving people the option to not destroy information...

If you had this information for all the text you've ever written, you probably would want to keep it for at least a few files.

[0] https://en.wikipedia.org/wiki/Talk_(software) just do `apt install talk`. If you're on macOS, you have it installed already I think.

[1] http://draftback.com I don't know if it still works. I also don't know if Google stores absolute timestamps or just a history of edits.



Not only could you chat in real time, you could also share parts of each others console. When I was in school, this was very useful as most of my friends were older than me. A couple of them effectively tutored me over ytalk (this was before we had learned what screen could do, then we switched to that). ytalk was great for ad hoc chatting and sharing, though.




Consider applying for YC's Fall 2026 batch! Applications are open till July 27.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: