Proposal to revisit extending the character limit for posts
In 2022, a proposal to extend the post limit limit was opened, and while it yielded fruitful discussion, it never entered a formal voting stage. I frequently find myself running against this post limit, and many other instances that we federate with have a limit that is much higher.
To kick start discussion, here are my notes on the previous thread:
Pros:
Many servers already have this
Very little storage burden
Makes it easier to post longer things
Cons:
Additional maintenance burden as we would need to switch to a fork (which could have it's own pros/cons), or apply a minimal patch to the mastodon codebase to extend the limit (Basically just replacing a few `500`s with whatever `x` we decide)
If people don't like longer posts, the local timeline could be marginally worse (the federated timeline already experiences this).
I will open the floor for discussion and then begin a voting phase. For clarity, the two questions at hand are
Should social.coop extend the character limit?
If so, what should it become?
I like this as another option in the toolkit. Sometimes threading shorter posts makes sense, sometimes it doesn't.
1000 characters still feels 'micro-blog' to me.
It's classic open source culture to try to be everything to everyone, but I think that the creativity that drives twitter-like social media is found within constraints of character limits.
@Robert Guthrie not that this is a good point because twitter has gone to hell, but at this point their character limit is 25,000
One other thing I'd like to note is that I don't consider this permanent. We should try it out and then run another poll which is "do you want to revert back to 500 characters"
People should be able to express themselves how they see fit. I think we will need to do something to address collapsing posts in a better way, but that is not relevant to the question of 'should we increase the character count' - thanks to tech WG for looking into this. Longer thoughts here:
https://www.loomio.com/d/C4HKmTOa/proposal-to-revisit-extending-the-character-limit-for-posts/54
I originally said that I don't have feelings about this and was happy to let others decide, but having seen some of the technological "solutions" for making it easier to skip over longer posts, I'm switching this to disagree. 1000 words isn't long enough for a properly developed set of ideas, but it's long enough to make Twitter-style "threads" obnoxious to read, especially if there a technological fix is applied to make sure that the posts don't take up too much screen real estate.
I would much prefer a stable fork such as hometown or glitch that not only increases character count but crucially also collapses/truncates longer posts in the timelines.
I feel Mastodon is ill-suited for long-form writing, e.g., due to lack of formatting tools, and due to the feed becoming unwieldy when filled with long posts. Shorter posts serve as a useful way to "preview" whether I want to engage with a thread or blog post. IMO longer posts are better put on a federated blogging platform, cross-linked from Mastodon. Thus, I prefer the current setting, but I can live with a change if that's what the majority prefer.
I too am a fan of brevity but I frequently run into the 500char limit. 1000 isn't a blog, it's just a bit bigger. I'd like us to try this and watch it carefullyl
Agree. 1000 character limit would be a good length to make my back and forth conversations more coherent and readable, without reaching a length that becomes unwieldy and gets into novel-writing territory. I'm actually way under the character limit on most of my actual posts. It's more so replies which can start to grow a lot, and most frequently they need to be broken in two.
Agree with Abdi-Hakin Dirie and Erik's comments. Concision is a useful constraint. Longer posts approach "get your own blog," territory for me. Threading can provide organization to extended writing which is harder in a single post. Shorter is more readable, and one can choose to see a whole thread just like any other extended conversation. Thank you.
Disagree. I think 500 characters is fine. It encourages people to keep their posts concise, while threads allow for additional context if needed. Threads also provide a better reading experience in contrast to blocks of text. I would agree if the proposal was to extend by 250 characters and test the look/feel — but doubling the character limit seems excessive.
Every character count increase from 140 to 280 on T, and then 500 here has felt odd at first then a relief as I lose less time fiddling with punctuation and swapping words out to get things under the limit. I assume the same will happen with the next increase - yes there will be more verbose posts, but short and snappy posts will always be more popular imho, ie there's already an implicit penalty for writing mini-essays, there doesn't need to be a hard constraint.
I'm opposed because I find the limited size a significant feature of the system. I like having to edit myself down to fit within the current constraint. Shorter posts are easier to read when you scroll down your feed.
I think a fairly short format text is what this technology wants. They are "toots" not whole songs.
I don't have a strong opinion, though aesthetically I tend to agree with those who think 500 works well.
I like the current situation but if I have understood correctly, the change will be subject to evaluation after a certain period of time and I'm OK to test this change. It's important for me to get the feasibility feeback from the Tech WG and Community WG (assessment of the impact of this change on their work).
For pretty much the reasons everyone else has said: less need to use threads. I'm also on another instance that has a 2222 character limit and that hasn't led to people tooting long screeds, and there's a "Read More" button you have to click if the toot has over 1000 or so characters. We could do something similar here.
I often use my profile here to post about projects I'm working on, and 500 characters isn't enough to go into an amount of detail that properly expresses what I want to say, so it ends up being just "here's something cool I made", instead of "here's something cool I made, and here's why it's cool from a technical standpoint". I think the latter is much more valuable.
I'm only commenting on my own personal use-case for a longer character limit because I've genuinely no idea how it'll affect the community at large, or other people's experiences.
I unequivocally agree. I think there is no good reason not to.
Sometimes the 500 char limit seems needlessly restrictive. Doubling it would, I think, provide more breathing room without turning this instance into a full blogging service. We already federate with a bunch of instances that use longer limits so they are already in a lot of people's timelines.
I agree with the sentiment expressed by @Luis Villa. I might change my mind in the future if Mastodon, in general, moves this way as a whole. At the moment, I don't support moving more to a blog style as an exception to the standard of the default Mastodon settings.
I don't think that Mastodon is the right medium for longer posts, 500 is about right for what it is. Firefish / Misskey is more adapted for that "too long for toots and too short for a blog" space
Long posts don't fit the timeline-based, microblogging format.
I've used Mastodon (500 chars), Bluesky (300), Twitter (280). I already find Mastodon's timeline the hardest to scan. When I crosspost from Mastodon to Bluesky and have to abbreviate, the concision makes my posts better.
Microblogging means a low bar to posting & low bar to me seeing a post. This model doesn't work well with multiple paragraphs. Posts need to get to the point so I can quickly filter.
No strong feelings either way.
A larger limit means, to me, more freedom for our users; people who don't like it can probably ignore the feature with minimal disruption. Strong yes from me to this experiment!
I agree with others that doubling the character count makes Mastodon more blog-y and brevity provides a good forcing function. Would prefer to keep the shorter format but I don't mind if the community ultimately votes to increase it.
it would be cool to have a soft constraint, where you are warned, in the yellow or something, when over 500 characters. But agree that moving to 1,000 removes an occasionally frustrating blocker to a post or reply while keeping us in microblog territory— hitting 1,000 characters is a clearer indication it is time to get a real blog :-)
@benjamin melançon This would be sweet, noted!
Great job summarizing and sharing the outcome, Evan! Thank you for your work here shepherding this experiment forward.
Many of my comments on Mastodon are technical answers that are difficult to fit in the limit, and breaking it into multiple replies is less good. The posts I originate can usually be edited to fit but when replying to questions that require a lot of detail in the answer it's more difficult.
Social Stuff
I respect other people's desire to keep their own posts short, but I think that opposing increasing the character limit because they have grown used to the 500 character limit as some optimum of brevity veers pretty strongly into "you shouldn't want what you want" territory for me. Long posts clogging the feed is a UI problem that other forks have solved (and we should as well, see below) and is an entirely separate issue from whether we should raise the character limit. as others have noted in thread, reading is optional, so by longposting you aren't 'demanding other people's attention' any more than you are when you are regularposting. If techwg says they can maintain the fork, then i don't see any reason not to raise the character limit - let ppl express themselves how they want. It is deeply patronizing to me to say things like 'having a low character count is what makes creativity/culture/etc work' - don't impose your beliefs about what 'works' or not on me or anyone else, imposing those beliefs on others is contrary to the cooperative spirit to me. If you like brevity, cool. If you want to longpost, also cool. I think it's an especially pretty hard sell to suggest there is some qualitative culture-breaking threshold somewhere between 500 and 1000 characters.
Technical stuff
I run an instance with a 10k character limit that's a fork of glitch, and whenever I use social.coop I am both astonished that base masto a) doesn't expose this parameter more easily, but more importantly b) doesn't have a "collapse long posts" feature.
So, eg. here's what my profile with a pinned long post looks like on our glitch fork:
and here's what it looks like on social.coop
so there is a "read more" button that then takes you to the full post, which works for a post within a feed, but since masto uses the [same component in notifications](https://github.com/mastodon/mastodon/blob/616789454707be00b334070646491b5e028d3be6/app/javascript/mastodon/features/notifications/components/notification.jsx#L208) then any long post makes your notifications unusable. I see above that the collapse behavior has been noted, but it's a) not enough of a collapse, it would still totally break your notifs and b) the UI is awkward, taking you to another page rather than expand/collapse in-page.
like this is generally true of base masto (notifications are unusable), but is an especially important thing if we have longposts. I think most of the concerns above about long posts are mostly UI concerns - base masto handles them terribly, and so collapsing long posts is like a de-facto "long" c/w where one can see a preview and expand the longer post if they want to see it.
(imo we should be running glitch anyway, but just wanted to put this on ppls radar as another concern)
@jonny twg isn't opposed to glitch, just there hasn't been enough user initiative... and by that I mean we should run an alpha, have a proposal, ect. I think that at 1000 characters, the see more concern isn't particularly pressing — at least not more so than it already was. While I don't feel strongly either way on glitch, I'm willing to drive this once the character limit discussion concludes (and also I agree that "should we run a fork" is (or at least should be) a separate question to "should we change the limit".
When we have that conversation your opinions as an instance admin on glitch will be immensely valuable, stay tuned (or start the conversation yourself!)
But yeah, I agree that at least from a reading perspective not much will change unless you spend a lot of time in the local timeline and/or have a feed of mostly social.coop users, and I still don't think the change will be huge.
@Evan Boehs maintaining a glitch instance is almost identical to maintaining a base masto instance, and we maintain a fork of glitch which is similarly nbd. we merge upstream every few weeks and it is basically never an issue. having the freedom to hack on the instance software is extremely worth it since base masto is broken in a bunch of ways
@jonny
I actually think the social.coop screenshot looks a ton better.
@Dynamic It really doesn't when you see a hundred of those in a row in your notifications, trust me. like yes on its own it looks better but that wasn't the point of my post
@jonny
Do they automatically all expand when you try to read a thread? Because imagining trying to read a thread made of these things just brings on psychosomatic eye strain for me.
@Dynamic it's exactly the same React component in notifs as when displayed normally, so yes without having the narrow collapse it does render like that for every single interaction on a post. this is one of the first things you notice during live usage of masto with longer character counts that's not immediately obvious from the base UX.
it's also true for 'edge case' masto posts. so eg. here's a screenshot of what it looks like when i post a 1px wide, 100000px tall image and then like it from another account. :
so that actually goes on for an unbelievably long scroll distance.
point being that you need the narrow collapse as the default view to make masto sorta usable with longer posts. i would expect that most people in this thread who are talking about short posts being the soul of brevity and how longer posts get in the way/should be on a different medium have just never experienced an instance that handles long posts well.
@jonny
I don't know if you actually answered my question, but it sounds like my guess is correct that this "solution" would make it even harder to read long threads made up of long posts. I'm on the brink of changing my vote to "disagree", just on the basis that allowing longer posts might open us up to this kind of interface shift.
In short, I like things the way they are, including when I see long posts from instances that allow it. I think it's fine for different people to have different use cases with some making longer posts and some making shorter posts and people deciding who to follow accordingly, but I don't want to lose the current interface experience.
@Dynamic sorry if i'm being unclear - i'm saying that the notification problem is the much more impactful one with longer character counts (the proposal at hand) than the reading problem. expanding the posts in a thread is not a big deal at all - unlike in base masto, in glitch the expansions happens in place rather than going to a different page. this makes it actually more possible to read a thread, since otherwise it's impossible to get a sense of the branching structure with the current interface, abbreviating the posts allows you to get a sense of the whole thread much more easily. adding a 'toggle all' button would be trivial, we've never had someone ask for it bc the benefits are pretty obvious once you start using it in context instead of seeing it in a single screenshot.
in glitch this setting is optional and configurable - you can turn off collapsing posts altogether and set the threshold for collapsing. i really think the answer is 'working on the UI' rather than trying to freeze it in place. none of this is necessarily relevant to the question at hand though, i was just trying to raise the notification ui problem as a potentially unforeseen one for anyone who hasn't been on an instance with raised character counts.
since most 'no' votes are talking about how hard it is to read long posts when scrolling in one way or another, i also wanted to say that forks that deal with longer character counts have figured out ways of dealing with them, and the example of the long image shows that those features would benefit base masto even without expanding the character counts - aka that the problems are orthogonal, and difficulty of scrolling through long posts is not a reason to vote no.
@jonny
What I needed to hear was that collapse could be turned off, so thank you for that.
I'll try to keep my relief at that separate from my extreme irritation at what you wrote here,
this makes it actually more possible to read a thread, since otherwise it's impossible to get a sense of the branching structure with the current interface, abbreviating the posts allows you to get a sense of the whole thread much more easily. adding a 'toggle all' button would be trivial, we've never had someone ask for it bc the benefits are pretty obvious once you start using it in context instead of seeing it in a single screenshot.
I've been talking about slightly unpleasant experiences that I've had on Mastodon as it currently is, which I know would be more unpleasant if the longer posts got shortened further. Telling me that I will be able to read long threads better if the individual posts are abbreviated when my lived experience is that it doesn't do that at all is incredibly off-putting.
@jonny
"so by longposting you aren't 'demanding other people's attention' any more than you are when you are regular posting"
Hard agree. In fact I'd go further. When it comes to replying to people rather than being OP, I inwardly cringe every time I end up posting multiple times to get my point across. Putting one long reply into someone's notifications feels like way less of a demand than 2-3 short ones.
@dash82 This is very relevant to me.
I don't have strong feelings about where we set the word limit, but I do have strong feelings against efforts to "aesthetically" shorten longer posts in ways that break up the readability of long threads.
The only real concern I had with longer limits was that I've found it unpleasant trying to read long threads where the individual posts were long enough to trigger the "Read more" link, making it impossible to view the entire thread at once. The Glitch style shortener that Jonny posted a screenshot of looks even worse in my opinion.
As expressed elsewhere: a larger limit means, to me, more freedom for our users; people who don't like it can probably ignore the feature with minimal disruption. Strong yes from me to this experiment!
♡Dear @Evan Boehs
Thanks a lot for the clear explanation and results. :D
As a researcher I'd love to help with the evaluation (these 3 questions "(How) has this changed the way you interacted with Mastodon", "(How) has this changed the culture at social coop" and "If you noticed changes, are you happy with them?"), but this doesn't need to be done now. :)
Thanks @Evan Boehs, this was a great process overall!
I just spent 10 mins hacking sentences over and over to get my 700 chars for a toot into 500, and found myself wondering what the status of this was? It says above there should have been a review by June 13, but I didn't see any mention of it in the recent minutes.
This isn't intended to be pushy – I realise there's multiple steps to this and it's not urgent. But the discussion we all invested in here seemed like a promising area of community democracy, so was just curious to learn what, if anything, is holding things up.
@Randy Hall
I actually think there's even more to the problem with long Mastodon posts than @Scott Feeney notes here.
I just posted a thread about it on Mastodon: https://social.coop/@dynamic/112428200242353565
None of this is to say that "threads" (a term I really dislike but I don't know a better word for microblog monologues) are "good" and longform posts are "bad", but I do think that allowing longer posts moves Mastodon further away from its strengths.
It feels easy to say "well, people who don't like longform posts don't have to write that way or follow people who do", but I think the reality is we would all write differently without the 500-character cap, and that that change is not without tradeoffs.