Hacker News new | past | comments | ask | show | jobs | submit login

I've been running at home and at work with first-party isolation enabled for a few months now. Google login works fine for me, as does login with Google. It's broken a few internal tools, especially when people do things like hotlink across internal systems (which would have been broken at least some of the time for most people anyway, until they realise that to view _this_ page properly they need to log in over _there_). Also breaks PlayStation Network.

All in all, insufficiently broken for me to be bothered enough to turn it back off :). I do wish there was a "view from alternate origin" feature though, to let me load a site as if it were loaded in an iFrame -- that would let me work around the issues with my internal sites.

I’d appreciate a ‘New Temporary Non-Isolated Window’ for use with websites that don’t work with FPI.

Introducing another container doesn’t address compatibility problems for websites that absolutely must communicate cross-origin. You can’t disable FPI in just one window, tab, or container.

The best strategy is to not use Google however. It's not like you need to in 2018 anymore. There are better services for almost anything out there, although you may have to pay a few dollars for some of them.

Well worth it since they are superior to Google. For email, Fastmail is king.

I run my own mailserver and use DDG for all my searching.

And use GMail as an MUA and am the GSuite administrator for the school I support.

Using Google is a trade-off, and there are definitely services they provide where the trade-off is worth it for me. Your mileage appears to vary. First-party isolation definitely kicks the needle across towards Google only having data I'm happy for them to have.

I use “alternative search engines” daily. However, I have to crawl back to Google if I want to find things that were published in the last two weeks. Even Microsoft Bing can’t keep up with all the content that appears on the web every day.

If its just search engines, you dont exactly have to be logged in to use Google.

Also, there are Google search-proxies like startpage.com

Startpage and the like don’t get everything you find on Google. E.g. you can find today’s articles from even the most obscure blogs on Google. Searching for the same article won’t return that result on Startpage until at least a few weeks from now. I’m not sure if this is Google holding back on the good stuff, or if it’s Startpage not being willing to pay a higher fee.

HTTP-only? Nope, nope, nope.

site:news.ycombinator.com About 91,000 results (0.38 seconds)

site:news.ycombinator.com in google.com About 11,90,000 results (0.25 seconds)

And how many of the initial results in google will be filtered out (e.g takedown notices) when you hit the very last page? I'm doubtful about the usefulness of the results count in comparisons like this.

Right? I tried so hard to use duck duck go and bing, but at some point DDG seemed to start producing very bing-like results, and bing is just not as good as google for technical searches - apparently by design, they believe their results are better for general users.

Kind of frustrating :-/

DDG is powered by Bing and Yandex. DDG doesn't have their own index.

For many people (like me) this is not advice that can followed since many many companies use G Suite.

Use Google in a temporary container (Firefox plugin) and isolate Google completely if you need to use them.

Also you can use startpage.com for searches, it uses Google behind the scenes. Looks nicer too. :)

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact