I have no clue why you'd build an app on Twitter's API anymore. I get it from their perspective (they want to control the ecosystem completely), however they've made it clear they don't care about anyone developing on them. Which is sad, since they only exist thanks to their early community of developers.
I can't believe they killed off their MacOS Twitter app, told people to use third-party apps, and now are effectively killing them off too.
(All this being said... Twitter is ripe for abuse, and locking down their API is probably one of the best ways to fix this)
They killed off the Windows Twitter app a few weeks ago too, replaced it with an iFrame to www.twitter.com
I really don't understand why a social media site so large and influential is spending so little on their own product. Twitter won't let anyone develop any clients, but Twitter itself won't develop any clients either. Twitter acts like they wished no one would log in at all, they act like they're just waiting to be acquired and shut down.
If anything you'd think they would want native apps to make adblocking harder. If you use tweetbot you dont get ads at all. Kinda why facebook pushes you to their app when you go to the mobile site.
Tweetdeck is still offered as a desktop client. Though these days its is ony available as a webapp via https://tweetdeck.twitter.com or as a Chrome app.
There was a point in time, around 5 years ago more or less, when third party twitter clients were driving a fair amount of innovation in desktop UI, particularly on OSX. As a mentor I'd even ask if my mentee would want to learn to code by pulling some stuff out of Twitter and putting it in their own design on their own webpage. There was a lot of hope and promise back then, and Twitter was giving a voice to people in harsh regimes as their other modes of contact were being shut down (like with Arab Spring). They looked like they were on the up and up.
I don't think moves like this solve Twitter's existential abuse problem, which I believe has far more to do with their 140 characters and unsolicited heckling than it does anything else. It turns your whole reality into gossip and this gossip overpowers any semblance of rational or reasoned discourse. It's intensely toxic and shutting off an API and some client apps changes nothing about that.
A follow on from this question is why you’d ever build a company based entirely on analyzing data from another company. Best and worst case scenarios:
* you do something amazing, and start making money. So your source company turns off your access to the data and makes there own version
* you never make enough money for them to care/notice you. They remove the source of data for non-competition related reasons/accidentally break the api and don’t care enough to fix it (if you’re not really making money presumably they’re not making money off you)
Shrug - maybe it’s simply a relatively easy way to get to the point where you can say enough buzzwords convincgly to get VC money?
The article explicitly said it was "bootstrapped" - ie not dependent on outside investors.
If I had a great idea for software that I could do on my own with the only investments being my time + some AWS resources where hopefully I could make enough money to pay myself a decent wage and pay for the operating expenses, why not ride the wave, make some money and when the bottom falls out, shut down the AWS resources, close shop, throw it on your resume and get job.
'you make money and then the company takes away your income' is not a 'best case scenario'. You can trivially come up with real scenarios that are much better than that.
Sorry an option is that your source company seeing you making money, and then decides to let you continue to do that rather than making that money themselves.
I deleted my twitter account a while ago but this is a new level of forgetting history. If I remember correctly Twitterrific/Iconfactory were the first folks to actually use the verb 'tweet', and absolutely provided the first iteration of the interface to twitter that we all expect today. It took acquiring Tweetie, which literally invented 'pull to refresh', before they had their own native app 3 years after Twitterrific had been released.
Twitter's success is directly attributable to a fantastic native app ecosystem on mobile platforms. If there's a lesson to be learned here, I guess it's "don't make apps for platforms you don't control because they'll drop you like it's hot the moment investors want an exit"...
As an aside, Tweetie/Twitter for iPad's column based interface was one of the absolute best interfaces I'd ever seen and used on a tablet, and one I have yet to see reimplemented in a use-friendly way to this day.
You're absolutely correct re: the history. Twitterrific not only came up with the word "tweet," they were the first one with a blue bird logo!
Twitter basically encouraged a third-party ecosystem to blossom, then decided that rather than finding ways to monetize that ecosystem, they should start slowly strangling it to death. I can't imagine that this is happening because investors are insisting on it; yes, they want a positive return on their investment, but who's really saying, "If only you guys knife Tweetbot, the stock price will double!"? There's a very small minority of users who use third-party clients, but those users are disproportionately the highly engaged "super tweeters," if you will. Is supporting them -- ideally better than they do now -- really so expensive that it's easier to just kill them? And if the answer is yes, is it really better to keep pretending that you're not doing that, and instead just make the experience of third-party developers and dedicated Twitter users steadily worse over time?
They faced the prospect of aggregation of the clients via takeovers, this would have allowed the owner of the majority of the clients to change out the platform. I had a conversation with a private equity guy in 2009 who wanted to do this, what stopped them was the idea that twitter would ratelimit and control the interface. So there was a pretty rational fear driving Twitter to do this.
The thing I was talking about was that around the time of Twitter's IPO (I don't remember the exact timeline) Twitter announced changes and limitations to 3rd party clients, including a 100k user maximum via a hard token limit[0].
Investors didn't say "knife the indies" but they certainly said "monetize, monetize, monetize" and Twitter decided the way to "monetize" was to force users into channels that had ads.
Sure, nobody ever told Twitter to go after Tweetbot, but it's not like anybody on the board said "wait a minute..." when this was the direction they announced in 2012. I'm lamenting the tail end of a strategy that started 6 years ago.
> I just find it strange how developer-hostile twitter is, considering the community it once fostered.
Think about the money and it's not so strange. Everyone has a price.
I'm not saying it's right, but it's not a mystery. The got big off the backs of others and they've been screwing them over ever since. People building stuff on top of Twitter are crazy. They've been developer hostile for years and years now.
I’ve been meaning to start my own Pleroma instance due do the excessive hardware requirements of the Mastodon client. How would you compare the two projects since you’ve used both?
So currently, choosing Mastodon vs Pleroma boils down to what you care more about: performance/less complexity, or more features/maturity.
Mastodon is really nice, but it's fairly intensive. Precompiling the rails/webpack assets in particular made my tiny VPS choke, even with 4GB of memory I believe. It also requires Sidekiq and other services [1].
Pleroma is newer and written in Elixir/Phoenix purely as an API. It only has itself as a dependency, and it's lean enough to run on a Raspberry Pi. It offers two frontends, pleroma-fe and mastofe, the latter is a port of Mastodon's UX.
---
Features wise, Pleroma doesn't really have a lot beyond the basic functionality of subscribing to people, posting, etc. There aren't real moderation tools (yet), since it's mostly a weekend project for the core devs. Blocking instances requires an IP block still, I think.
Mastodon wins features-wise, hands down. But some things aren't really easy to reconfigure- since the frontend is baked in, changing things like the post character limit (500 chars) is tedious. In pleroma it's a simple config file change.
---
I like Pleroma personally but it's not for everyone (yet). I plan on contributing more soon to help change that.
I also was only running single user instances so I can't really discuss scaling. @technowix@niu.moe [2] would be better to ask about that, considering their instance has ~3k users [3].
Mastodon and Pleroma are both part of the Fediverse [2]. They work with any software that works with OStatus and the newer ActivityPub. This includes things like GNUSocial, etc.
Mastodon doesn't explain any of this well, or at all, to my dismay. (As far as I'm aware, on the joinmastodon page).
I use the web app version exclusively -- based on that I don't agree.
Both the Twitter and Facebook Android/iOS web applications constantly push you to download the native client. They even get demanding sometimes. They want to be able to send you push notifications.
They're not dead, they just can't offer push notifications, or automatically update your timeline when there's a new tweet. Nothing else in Tweetbot will have to change.
It doesn't. There's no replacement for push notifications besides using the official app, but you can still manually refresh the timeline (via pull-to-refresh).
I switched away from the official app to get away from the "algos" that keep surfacing content that's either old or I'm not following. No I didn't miss those tweets, and I don't care what someone else liked.
Stop turning Twitter into Facebook! I'm sure it increases your engagement metrics, but at some point I'll feel overengaged, realize how much time I'm wasting and quit the app completely.
Same here. Algos surfacing stuff I don't care about at all (and can't turn off!), and I've run across enough instances of the official app/page not showing me some tweets (checked exhaustively for a couple days of tweets around the ones it missed, couldn't find) that I've utterly dropped the official clients.
Talon supports the use-case I care about: reading all the tweets from people I follow because I care about them. Twitter and Facebook both make this nearly impossible, so I don't use them directly.
Bye Favstar. I followed you for years. They shut down my https://2fb.me app so many times, I no longer keep track of how bad they f'd it up. The latest blow was to my @shareU account. Yes, about 3 months ago I owned twitter.com/shareu and it was alive and sharing people's tweets to Facebook. Many people loved it, and it was a great promotional tool. All it's following was 100% organic, and it followed no other users. 17% of it's followers were from verified accounts, yet it was shut down for doing what it was set out to do, share tweets to Facebook.
but the original “like” on Twitter is back, http://2fb.me for the time being, let's see how it goes.
I don't understand what Twitter is doing, I think they arguably have one of the most valuable social tools on the planet in terms of real time information and news, more than Facebook, more than Google...but they don't seem to understand what they have, how useful it can be in rich API forms, etc. It's baffling. They are hostile to their developers and to their userbase all in some weird chase to be Facebook. The one and only good move Twitter has made recently is teaming up with Bloomberg...but even that has been a botched a bit.
IMO Jack has to go, he seems to understand Square, but he seems very very out of touch with Twitter.
I started to use third-party twitter clients exclusively because of the annoying timeline manipulation. Now this.
Is there any other way around other than quitting twitter ?
Instead of following people, add them to a list. You get a sequential feed of tweets from everybody in the list, and no ads. Works across all official Twitter clients, no need to worry about API restrictions or token limits.
They should have waited a little more before shutting down on June 19th. Twitter announced they will postpone the death of real time streams after Tweetbot and some other apps complained about it weeks ago. It was told on the Twitter API account.
All these years and I haven't grasped what favstar is for. The article also really doesn't clear it up. I only see it congratulating people on milestones of 500/1000/etc. Favs.
Favstar was useful back in the day when you didn't get notifications about favs, etc. Not so much anymore. I also used it to read the most liked tweets of certain users, but these greedy fucks want you to pay to read the top one (which you can bypass with Inspect Element) and to read more than the top 10 or so. So, good riddance.
Not associated with Favstar. But, if you want to call names for someone who wants to make a living on the product they built, then you have a wrong understanding of software.
I respectfully disagree. I could regurgitate the content on the website but that would hardly be any more substantive. Often these threads lead to curious people asking questions about Mastodon, and I'm happy to answer them and engage in more discussion.
I can't believe they killed off their MacOS Twitter app, told people to use third-party apps, and now are effectively killing them off too.
(All this being said... Twitter is ripe for abuse, and locking down their API is probably one of the best ways to fix this)