You can’t lurk on Instagram anymore unless you’re logged in (2019) | Hacker News Reader

You can’t lurk on Instagram anymore unless you’re logged in (2019)

mashable.com

31 pointsannadane14 comments

For the past several years, my experience of Instagram (as a non-user) has been that occasionally I'll click on a link that leads there, and I'm met with a page that asks me to attach a phone number to my account. There is no option to log out (it seems to be an evercookie) and no option to refuse. So I leave, not willing to give them a phone number and not wanting to bother with an incognito window. If it's important, I'll ask the person who posted the link to repost the content somewhere else.

There are mirrors: bibliogram.

Instagram just tells me to login, I can't even view individual posts.

It's really easy to scrape Instagram from outside the browser as a non-user. No API or Access Token needed. Two steps: 1. Get the profile ID, 2. submit query via GraphQL.

For example, if the username is "fish" and we want the first 36 entries.

   1. x=$(wget -O- -q https://www.instagram.com/fish/?__a=1|grep -o "profilePage_[^\"]*" |sed 's/profilePage_//');

   2. wget -O- -q "https://www.instagram.com/graphql/query/?query_hash=e769aa130647d2354c40ea6a439bfc08&variables={\"id\":\"$x\",\"first\":36,\"after\":\"\"}"

Do you enter these two commands sequentially?

Yes the profile ID is needed for the GraphQL query

While the API access isn't quite as limited as their regular website, don't they eventually block your IP after too many queries? At least that has been my experience when browsing via Imginn – eventually picture loading stalls and in the devtools you can see that in that case the GraphQL requests redirect to a login page instead of returning the desired data.

At least that's where having a dynamic IP comes in handy :-P, and with Imginn some minimal fiddling with the page inspector (unhiding the "Load more" button again) lets you retry the query, so your infinite scrolling progress isn't lost.

In their robots.txt, no crawl delays are given, someone seems to have overlooked that ?__a=2 works just as well as ?__a=1, and overall, whether they intended to or not, they appear to be saying they only want to allow certain bots. Perhaps try changing your user-agent string to match the ones in robots.txt.

Generally I do not use middleman sites like Imginn as I find I can perform for myself whatever "service", e.g., downloading resources from some other website, they are offering.

Alternatives like wget and mirrors are just not worth it. If it asks me to login, I will just abandon. Same with Pinterest. Neither is worth it.

I am getting very tired of Mark Zuckerberg buying up the internet.

I'm thinking of posting my photos to something like https://micro.blog instead. Or posting them on my Gatsby site which is just a little bit more work for me (since I'm a developer) but definitely lacks a little bit of convenience. The biggest thing that's hard to give up for me is the simplicity of Instagram's image editing features built right into the flow of making a new post. I don't use the predefined filters, but I've got a pretty good personal style that I've developed with the manual editing tools. Any tips on ways to reproduce this flow are welcome!

Pixelfed?

They don't even seem to have an app, so I'm not sure how that is a viable alternative.

I've seen this pattern repeated with many websites. Once they become popular enough to afford it, they lock all their content, disable their APIs, and start requiring their users to verify emails, share their phone numbers etc. Export options become hard to use, open protocols get squashed, API access becomes really difficult to get.

It seems like the end game of any large website is to hold data hostage.

Reply on news.ycombinator.com