Twitter's website is absolute garbage, the feed is unresponsive, tweets fail to load, it's just terrible. If you're modeling your frontend off of anything please do not pick Twitter because their web experience is awful.
I would think that your first visit loaded a resource/set a state somewhere which let's the second visit go through.
Much more likely to be a programming error then a purposeful decision by a PO as the end result is objectively terrible. I cannot imagine anyone getting convinced to register to a webpage that cannot even load. It's not that content is hidden behind a pop-up or anything, it just errors out (at least for me, but I'm not the parent comment author)
Some issues are fucking hard to detect and an initial page load breakage that only happens sometimes (likely a race condition that cannot be replicated in testing) is an example of that
It might go away if you log in. I wouldn't know.
It's unbelievable incompetence coupled with unbelievable not-giving-a-damn attitude.
Apropos of that: the MTA is using threads now, and I can't click 'show thread' without being redirected to a login screen. I don't want to create an account because I find twitter painfully shallow but also addictive, as well as generally untrustworthy (I don't want them to have my phone number).
I don't think that any part of the government should be using twitter the way the MTA is; but so long as they are, I think Twitter would do well to get out of the way of the distribution of vital info.
Any move to develop it further to add those sorts of things would have run into the same problem other protocols have for the last 20+ years, which has made them even more difficult to develop and promote than they already were: keeping huge numbers of users captive has become per se valuable, thanks to having as-complete-as-possible spying-generated profiles on people and massive datasets being important both to the money-spigot of the modern online ad market, and for developing machine-learning solutions to dominate other "verticals" (against any competitor who lacks the same massive, spying-obtained datasets, so is at a huge disadvantage). Application-level open protocols make tracking & spying harder, so companies chasing those datasets really don't want them around. The few that survive in common use do so under constant threat, protected only by luck, by already being "good enough", and by having already established a large presence before that sort of business model really took off.
[EDIT] incidentally, pioneering the inclusion of an open, cross-site social network as a first-class browser feature is one of my crazy ideas for how Firefox could save itself from irrelevance—but like most of the other ideas I have to save them, it's probably too late anyway.
It's not like they have 1000 of concurrent events like facebook, and their frontend should have nowhere the complexity of their backend. I don't understand how it's even happening.
The most absurd aspect of it is that how slow loading a single tweet is. The timeline could have the excuse of a slow and complicated backend call (but it doesn't, the backend is actually doing ok), but several seconds for one single tweet?
This is particularly frustrating, because my normal use case for this is that a bunch of new stuff loaded, which throws everything that's on the page already out of order, so then I can't see something that I know was there, that I was just looking at, and now searching for it doesn't work, either.
The biggest irony is they have the best designers in the world (i.e. the guys who created bootstrap) but god I hate the tech they have used to make twitter.com because of that "something went wrong" retry button error message .
It's insane how bad it works on a page that would have been perfectly fine to be simple and static.
That said, I fully agree - the website is literal garbage despite its insane simplicity, and there would be 0 reason to trust that their web tech works well because they have never demonstrated such a thing.
I have zero interest in retweets, replies, threads, hearts, etc.