Hacker News Reader

I'm not sure why you're getting downvoted. I went to download my photos today from Google Photos after reading about the end of the unlimited storage. The interface is incredibly annoying. It split them into 2GB zip files, which is fine, but then it takes multiple clicks to get each download to start. Oh, and then it logs you out every 5 minutes so you have to re-enter your password.

It's pretty clear they were not super concerned about making it a user-friendly process.

If you say something negative about Google you’ll reliably get downvotes from Android fanboys even if it’s completely accurate. (Apple has similar fans, although they seem to be less defensive now that the company is doing so well)

It’s not worth worrying over except as a reminder that rating systems need to handle bad-faith voting.

This doesn't mirror my experience at all, and I think blaming it on fanboys isn't fair. I used Google Takeout recently, had it set to split into 50GB tar.gz chunks, and it worked perfectly fine. I wasn't logged out once, and downloaded the archive with around 90 MB/s (720 mbps). It was a very smooth data export experience. I'm very much not a Google/Android fanboy and avoid Google's services wherever possible. The OP is being downvoted because their experience isn't representative.

EDIT: It does seem that there are more people than I expected for whom the experience isn't as good as the one I had, so maybe Google should test this with (and make it work better on) internet connections that aren't as good as their office lines.

I just finished the process last night for about 80GB of photos and GP's comment represents my experience very well. It took hours of tedious clicking and logging in dozens of times to get through it. Miserable.

I’ve also used it and the original poster’s experience rang truer than yours. Most people do not have gigabit internet connections and manually downloading the default smaller chunks is annoying.

Yeah, it's been a real hit-or-miss process for me, too. The process is way too manual, and increasing the chunk size from 2GB to 10GB causes a lot more failures of the individual downloads for me.

And I don't particularly want to hear "get a better network connection". My connection works just fine for everything else.

That said, if you're able to download everything, it's reasonably well-organized, although as stated elsewhere, there's a lot of duplication of data.

But can't you just download larger chunks over whatever connection you have? Or is it common for internet connects in (I assume) the US to be so bad that you can't download a 20GB or 50GB file without errors?

It’s really a question of how well it resumes: if you’re trickling in a 20GB file and get a couple lost packets at 10GB, was that earlier transfer wasted? Most people use wireless so it’s not hard to have a transient failure which will disrupt a TCP connection but won’t last that long in absolute terms.

> We can't download more than 5-8 files at a time.

I understand that big file transfers are hard, but that's the entire point of the service.

There are plenty of file transfer tools that can take a list of URLs, but this forces you to do everything in the browser.

You can takeout each service seperately...

not sure why you're getting downvoted, i just wrote about the same issue i had about a year ago, it's an absolute mess for photos exporting, i don't think people actually use this service here, or they use it without actually looking what's in there.

Before downvoting, just try to use the new interface and try for youself. I have used takeout about 2 years back and it was okay. Ofcourse I had just few gigs at that time.

I have tried to spin off an Amazon EC2 instance to download and copy to an S3 bucket. But it logs out every few minutes disrupting the downloads. It will not allow to down the same file multiple times. If one zip fails, the whole set is useless.

Get faster internet. Take your laptop to a google campus and download on their wifi. Set up a VPS and download and manipulate there.

Plenty of solutions.

Those aren’t good solutions, they’re workarounds - and if you think about them even slightly you’ll realize they’re not very good:

1. The interface requiring multiple downloads prevents automation or simply waiting out a large transfer, and not having a robust automated retry mechanism ensures wasted time and increases the odds of data loss.

2. Few people have a high-speed free WiFi network nearby. You’re not getting better results at Starbucks or the local library, and Google’s campus networks require logins even if you are one of the few people who lives near one.

3. Setting up a VPS and running downloads from a web app requires money and skills most people don’t have, especially if you care about not accidentally leaking your personal data. If you have enough data to matter, you’ll also hit many providers quota limits or bandwidth charges. If you navigate all of those challenges, you still haven’t solved the problem of getting it home - at best you can now use rsync to remove the manual component of the second transfer.

“Take your laptop to a google campus and download on their wifi.”

LoL

i think he's just trolling - can't be that stupid.

GoogleGuest is an unsecured SSID at every campus, if you have one nearby. It generally reaches the parking lots (which are conveniently empty right now)

https://about.google/intl/en_us/locations/ ... not as isolated as I thought, waiting patiently for the geo-data people to work out an approximate census of people who live within a 50-mile radius of a google office. :-)