6 ms·
Sometimes, things that look obvious and simple require a lot of work and research to build so that it will fit the product and won't damage the community and mo
by Illniyar 9y ago
Sometimes, things that look obvious and simple require a lot of work and research to build so that it will fit the product and won't damage the community and momentum.
This isn't one of those times. Seriously, what in the world took so long.
- partiallypro 9y agoI wish it would take longer, because frankly it's a dumb idea. It defeats the purpose of Twitter.
- daveFNbuck 9y agoYet Twitter is much more popular in Japan, where the higher information density of the written language effectively gave them this feature from the start.
- wpietri 9y agoTwitter is used very differently in Japan than elsewhere, and in ways that seem tightly bound to aspects of Japanese culture. I'd be very careful about assuming that higher word count is really the key difference.
- daveFNbuck 9y agoI was responding to the idea that a higher word count defeats the purpose of Twitter. Japan shows that Twitter can be successful under a higher limit.
- wpietri 9y agoIt shows that it can be successful if you use it like it's used in Japan. It does not show that it can be successful when it is used like English speakers use it. Twitter is unusual in that it is a pretty spare tool. They've added a number of small features in support of how people were seen to be using it. But what mainly defines it is the ways people use it, not the tool itself.
- daveFNbuck 9y agoMeanwhile Facebook has added tons of features that no one wanted based on pure speculation and they've done much better. People will change how they use Twitter as Twitter changes. If that weren't true, no one would care about this.
- wpietri 9y agoThey have introduced and removed many things because they have a culture of continuous experimentation, most of it very small. Twitter is much less good at that, and this is definitely not the sort of small, quiet experiment that has done so much good for Facebook. So instead of this being a step toward Facebook-style experimentation, I'd say it's a step away. I agree this will change how people use Twitter, so I'm not sure why you're saying that in a way that seems contradictory.
- daveFNbuck 9y agoMany of your comments have been about how 280 characters won't work well with the way English speakers currently use Twitter. As this won't be the way English speakers use Twitter going forward, none of those comments matter.
- wpietri 9y agoIf you're just saying, almost tautologically, that users will use Twitter before and after the change, and that use may evolve, sure I don't dispute that. However, what you started out saying was that Twitter was even more popular in Japan, a place where the word count is higher. Given the context, it seemed like you were implying that this could or should make Twitter even more popular. I'm saying that's poor reasoning, because Twitter's use in Japan is distinct not for reasons of word count, but because the culture of use is deeply different. There are deep sociological differences, likely driven by the substantially different Japanese culture, plus a different early history of use than in the Anglophone Twitter sphere. 280 may be better for English-speaking Twitter and it may be worse. But Japan proves nothing about English-speaking Twitter because actual use is too distinct.
- subroutine 9y agoIt’s also dumb when I have to read 3 tweets in a sequence by the same person trying to get a single point across, each one with a different set of comments, likes, and retweets. Im not sure this will fix it though, maybe it will help. I feel like the same could be accomplished with an expando at whatever limit they set. That way when I browse twitter the tweet boxes are still the same size as originally, and if i want to read more I can tap the expand box button.
- rollinDyno 9y agoI enjoy being able to choose a specific statement and spin off a whole debate from there. I often think Twitter’s main problem is their inability to play this to their advantage. What if they could organise threads where if the author changes their mind in an earlier tweet (the premise), then the later tweets would no longer follow and stop forming a valid argument.
- partiallypro 9y agoThey could fix this by fixing how they thread tweets. Currently they thread tweets, but it's grouped in a weird way and sometimes someone that isn't the person that threaded them is in the middle of the thread. Also, instead of making people manually put (1/x) they could just create a counter and put them in sequential order treating any time the author only replies to themselves (which is how you thread) as the thread, but any reply to someone not themselves as outside of the thread. I'm sure that's complex, but I believe it could be done with some logic.