@annaleen is absolutely right when she says this, and here's a demonstration. I mean, OCR of screecaps is literally an OS-provided service these days:
@cstross @annaleen Hmmm... isn't your screen cap a pull quote from a longer thread? A longer thread that I don't know how to get to because there's no link back to the original context?
@cstross @annaleen An advantage of a QT is it provides both recontextualization of an information source to an expanded audience *and* it provides a link back to the original context
@steven @cstross You can also include a link with a screenshot. It's not hard to cut-and-paste some text.
@annaleen @steven @cstross maybe on desktop. On mobile it’s a PITA. And then I have to get the alt-text too if I want it to be accessible. Computers are good at this, I am bad at it, let the computer do it for me
@annaleen @steven @cstross the computer OCR even on iOS is also still bad enough that I have to correct it substantially, it won’t be good enough as is for screen reader users
@kevinriggle @annaleen @steven @cstross I mean, even if you're no Android fanboy, you can still use Google Lens or other third party parsing.
@mrzaius @annaleen @steven @cstross that is also real bad. There’s a reason folks call the output craptions
"I've been a journalist for over twenty years I've written for venues ranging from tiny zines to the New York Times. And think users should have control over quote- posts. If a journo wants to report something that they can't quote-post, I believe that the ten-second friction required to cut-and-paste some text, or to screengrab it, is helpful to the journalistic process. Taking a beat to consider whether we really want to quote something, and how we want to frame it? Literally our job." -Lens
The two inconsequential punctuation errors I see are plenty close enough for alt text

@kevinriggle @annaleen @steven @cstross

One hack here: ctrl-+ zoom to enlarge before screencapping will provide better OCR accuracy.

@cavyherd @annaleen @steven @cstross oh god. It was sent as text, it needs to be consumed as text, it should stay text all the way through, and not get round tripped through some JPEG, please just let the computer solve this problem for me
@kevinriggle @cavyherd @annaleen @steven @cstross In the year 2000 Github will become curated photo albums of Hollerith cards.
@kevinriggle @annaleen @cavyherd @steven @cstross So leave it as text. Those of use who have been on the other site from the beginning spent years copying and pasting text to make a quote post, so I can say from experience it’s easy to do. You can still include a link to track back to the original.

@annaleen @cstross Agreed, adding a link would address provider a pointer back to original context. And, it's great that Mastodon has an edit button so @cstross can go back and add the missing link.

It sounds like we are in agreement as to what a good QT looks like... so I guess the question then is: should it require a multi-step error-prone manual process or should it be implemented more directly as a standard client feature to ensure all those elements are there in every QT?

@steven @cstross I would prefer the multi-step process, the friction, to prevent easy drive-by pileons. errors aren't an issue, as you already pointed out, because we can edit.
@annaleen @steven @cstross In terms of consent to be QT'd, I don't see any practical difference between a screenshot+link and a QT. Any barrier involved can be solved by automation and people who want to QT will end up using clients/scripts that automate the process.