4 ms·
I strongly disagree that not allowing scraping protection on social network is a good thing (reasons mostly from this thread https://news.ycombinator.com/item?i
by mic47 7y ago
I strongly disagree that not allowing scraping protection on social network is a good thing (reasons mostly from this thread https://news.ycombinator.com/item?id=22182144 https://news.ycombinator.com/item?id=22182144 )
What I would do on Linkedin side is:
1. Split public setting into 2 settings: public for everyone including scraping, and second choice that make it public for Linkedin users (aka banning scraping [1])
2. Everyone on Linkedin who have current setting public would go though dialog to choose which one, with default setting to ban scraping (which is hard to criticize, since that option is protecting privacy more).
3. If you allow scraping, the form should warn you about consequences, and make you to acknowledge them (which is again a good practice, if you want to expose your information to EVERYONE on the internet).
4. Now you have users consent to protect their data from scraping.
Sounds like dark pattern? Kind of yes, but for good thing and it's hard to argue against it, since it increases privacy.
[1] Idea here is that I can allow anyone on Linkedin to see my profile, whether it is potential employer, or someone who met me at the conference (i.e. to improve my Linkedin experience). But I do not want that data to be harvested, by third party for any reason (specially, because they would then use it to to send me spam / advertising / prefill my profile on different website / ...).
edit: formatting
- buboard 7y agoScrapers can just create a user account then. They still seem to be covered by this injunction
- mic47 7y agoLegal: It will be hard to argue in court that you have right to create fake accounts. That is definitely against TOS, and at that point Linkedin can sue the company instead, so I believe it would scraper would not be successful at court. Technical: This is actually what you want, as limiting scraping per user is easier than per IP, since IP can be shared by users, so your rate limits have to be higher for IP address than for individual user. Additionally, creating fake account is more work which will in the end making scraping more expensive.
- buboard 7y agoa real account could scrape tons of stuff given time