Most active commenters
  • insane_dreamer(3)

←back to thread

418 points thepuppet33r | 45 comments | | HN request time: 1.263s | source | bottom
1. thepuppet33r ◴[] No.42175024[source]
Yes, Google deserves to be distrusted and avoided as a whole, but Google Scholar is a genuinely net good for humanity.
replies(3): >>42175704 #>>42180078 #>>42181021 #
2. dumpHero2 ◴[] No.42175704[source]
I have similar feeing for Gmail (it's effective anti spam engine), google maps and google docs (which pioneered shared docs. It feels outdated on many fronts now, but it was a pioneer).
replies(8): >>42175773 #>>42175878 #>>42176170 #>>42177151 #>>42177404 #>>42179179 #>>42186118 #>>42187586 #
3. roflmaostc ◴[] No.42175773[source]
anti-spam is only an issue if people dump their email anywhere. I usually register my mail on webpages as first.last+webpage@mail.com and once they would spam this mail, it gets blacklisted.

I literally get only 1-3 real spam mails per month without any filter.

replies(3): >>42175853 #>>42176172 #>>42186581 #
4. dripton ◴[] No.42175853{3}[source]
Words great, until a page rejects email with a '+' in it.
replies(3): >>42175970 #>>42176067 #>>42176089 #
5. coderintherye ◴[] No.42175878[source]
Good for users of Gmail, but is it a net good? Gmail spam prevention is great for the Google Apps orgs I manage. However, for the other inboxes the vast majority of spam they receive comes from @gmail.com
replies(1): >>42177789 #
6. 6510 ◴[] No.42175970{4}[source]
dots are ignored, can filter by john.doe@gmail.com

not sure about capital letters

7. hks0 ◴[] No.42176067{4}[source]
Not everyone's cup of tea, but quite nice if one can afford it: I have my personal domain and a catch-all inbox. So if I want to register at acme-co.xyz I will just use acmecoxyz@my-domain.tld

Maybe I should start using random words though? Wonder if someone will go bananas seeing their brand's name on my domain.

replies(1): >>42177840 #
8. AshamedCaptain ◴[] No.42176089{4}[source]
Or just knows about this Gmail trick (it's been 20 years already) and sends spam to your real mailbox.

Actually, I am surprised _any_ spammy website these days would even honor the part after the +, and not just directly send to the real mailbox name.

replies(3): >>42177255 #>>42177362 #>>42182174 #
9. whiplash451 ◴[] No.42176170[source]
Try MS OneDrive before calling google docs outdated

Google spanks everyone else on robustness and responsiveness

replies(2): >>42176281 #>>42179002 #
10. janalsncm ◴[] No.42176172{3}[source]
I see this recommendation everywhere and I am genuinely surprised that it works. Any spammer can find out your real address since there is an obvious mapping from + addresses to your real address. An actual solution would hide this mapping.
replies(1): >>42176304 #
11. rty32 ◴[] No.42176281{3}[source]
Yes until it fails

https://www.theverge.com/2023/11/27/23978591/google-drive-de...

replies(1): >>42176556 #
12. bachmeier ◴[] No.42176304{4}[source]
Yeah. Fastmail masked addresses are random. The best you can do is guess that an address might be masked, due to it not being johnsmith@fastmail.com, but it provides no information about your real email address.
13. whiplash451 ◴[] No.42176556{4}[source]
That issue got resolved in a few days [1] -- and for each and every one of these extremely rare events at Google, you'll find similar ones at MS.

I am referring to robustness at scale and every day: Google released auto-save years before MS. MS pales in comparison in the UX.

Note: I have no vested interest in Google, not ex-googler, etc.

[1] https://support.google.com/drive/thread/245861992/drive-for-...

14. globular-toast ◴[] No.42177151[source]
Google maps would only be a net good if the data was available under a free licence. As it is they take data from people that should have gone to a public project like OpenStreetMap.
replies(2): >>42177478 #>>42178530 #
15. thechao ◴[] No.42177255{5}[source]
I used to require a "+..." on all emails. Any email that didn't have the "+..." was sent to Spam automagically. My family were whitelisted. I gave up, because too many websites (early on) refused to take the "+..." marker, so I ended up losing too much to Spam. It's easier to just let Google sort it out.
16. gnopgnip ◴[] No.42177362{5}[source]
It's part of RFC 5233 Sieve Email Filtering: Subaddress Extension
17. gray_-_wolf ◴[] No.42177404[source]
Most of the spam I get is from gmail. Maybe they should apply their so effective spam engine to outgoing mail as well...
replies(1): >>42177542 #
18. arccy ◴[] No.42177478{3}[source]
"take", these people would never have produced any data if gmaps wasn't there...
replies(1): >>42177599 #
19. crazygringo ◴[] No.42177542{3}[source]
It's probably not. You can put any domain you want on the "from" address. Just because it says it was from Gmail doesn't mean it actually was, unless it's signed with DKIM etc.

I had a domain for a while that people got spam "from" all the time. It had nothing to do with me and there was nothing I could do about it.

replies(3): >>42177685 #>>42177697 #>>42179265 #
20. hatthew ◴[] No.42177599{4}[source]
At one point I contributed quite a bit to google maps, because it was the primary map system I was using at the time. Had I been using an OSM-based system, I would have made contributions there instead.
replies(1): >>42177773 #
21. dpifke ◴[] No.42177685{4}[source]
I run mail servers for myself, a couple of side projects, and some friends and family. A double-digit percentage of all spam caught by my filters is from Google's mail servers, not just forged @gmail.com addresses.

Of the "too big to block outright" spam senders, behind Twilio Sendgrid and Weebly, Google is currently #3. Amazon is a close #4. None of the top four currently have useful abuse reporting mechanisms... Sendgrid used to be OK, but they no longer seem to take any action. Google doesn't even accept abuse reports, which is ironic because "does not accept or act upon abuse reports" is criteria for being blocked by Google.

Most spam from Google is fake invoices and 419 scams. This is trivially filtered on my end, which makes it perplexing Google doesn't choose to do so. I can guarantee that exactly 0% of Gmail users sending out renewal invoices for "N0rton Anti-Virus" are legitimate.

22. gray_-_wolf ◴[] No.42177697{4}[source]
I would hope google has DKIM and SPF set.
23. arccy ◴[] No.42177773{5}[source]
indeed, osm can't paint itself like a victim, it needs good end products to bring in contributors.
24. thaumasiotes ◴[] No.42177789{3}[source]
> Gmail spam prevention is great for the Google Apps orgs I manage.

Gmail is unlikely to let spam through.

But that doesn't make its spam filter great; it's also very prone to blocking personal communication on the grounds that it must actually have been spam. The principle of gmail's spam filter is just "don't let anything through".

It would be much better to get more spam and also not have my actual communications disappear.

25. kroltan ◴[] No.42177840{5}[source]
Yeah, I've had to explain that a couple times already, usually when dealing with customer support or in-person registrations.

And a "malicious" actor can get away with pretending to be another company by spoofing the username if they know your domain works like that. I don't think this has reached spammers' repertoire yet, but I wouldn't be surprised.

Eventually I'd like to have a way of generating random email addresses that accept mail on demand, and put everything else in quaraintine automatically.

26. wbl ◴[] No.42178530{3}[source]
I ran into trouble because Open Topo does not report a stream the 7.5" series does. There's serious data quality issues that can make it not work for some applications.
27. Fogest ◴[] No.42179002{3}[source]
As much as I try to "de-google" myself and try to avoid being trapped in the Google eco system, I'd definitely choose it over MS Office. I am stuck in the MS Office eco system at work. Some of their products are starting to improve in MS Office, but you can still tell it's a lot of hacks ontop of old systems. Especially when it comes to the whole teams/onedrive/sharepoint side of things.

One of my biggest gripes right now is that we heavily rely on Microsoft Teams. A lot of our work laptops still are stuck on 8gb of ram. I find Microsoft Teams can easily suck back a full gig or more or ram, especially when in a video call. From my understanding, Teams is running essentially like an Electron app (except using an Edge browser packaged).

I have no problem with web based apps, but man, some optimization is called for.

replies(1): >>42179145 #
28. nextos ◴[] No.42179145{4}[source]
I use a decade-old NUC with plenty of RAM as a daily driver. It doesn't struggle with anything except MS Teams. It can churn through Zoom or Meet calls while compiling code. Teams is a bloated mess that makes the fans spin at max RPM.

It's crazy I can boot a kernel, with an entire graphics and network stack, X and a terminal in less than 200 MB but then the Teams webapp uses a massive amount of resources and grinds everything else to a halt.

Word 365 also becomes incredibly laggy on long documents with tons of comments, whereas Google Docs is just fine. But, apparently, this is also a thing on modern hardware. I guess these days Microsoft has little attention to detail.

replies(1): >>42179632 #
29. AlienRobot ◴[] No.42179179[source]
"Google is evil, except for all the Google products Google produced"

Honestly, if we compare Google to Amazon, Microsoft, Apple, and Meta, isn't Google the least evil one?

replies(1): >>42180086 #
30. csomar ◴[] No.42179265{4}[source]
The Spam I get from "gmail" and ends up in my spam folder is spoofed. The Spam I get from gmail and ends up in my inbox is from gmail. Spammers will mass-create accounts and mass-sell them to spammers.
31. Fogest ◴[] No.42179632{5}[source]
It's funny because sometimes Teams uses more resources than the Edge browser. Despite Teams being Edge based for their application.

I think overall many companies have gotten lazy/sloppy when it comes to optimization. Game dev is even worse for this. I like how Microsoft products integrate with each other, but often the whole thing feels sloppy and unoptimized.

32. insane_dreamer ◴[] No.42180078[source]
Google Maps is a net positive as well
replies(1): >>42180135 #
33. insane_dreamer ◴[] No.42180086{3}[source]
No, I’d put them in this descending order of evilness: Meta, Amazon, Microsoft, Google, Apple.
34. robertlagrant ◴[] No.42180135[source]
Some more:

- Google Search

- YouTube (more debateable, but I think it's a marvel)

- Google Books

- ChromeBooks

- Android

- Google Calendar

- Google Earth

- Google Drive

- Google Docs

- Waze

- Android Auto

- Google Pay

- Kubernetes

- Go

- VP8 / VP9

I'd rather take all those products than leave them.

replies(1): >>42186501 #
35. codeflo ◴[] No.42181021[source]
I'll reserve judgement on its net effect until the moment they kill it.
36. aorth ◴[] No.42182174{5}[source]
Good resource on this trick from 2010. It's not Gmail specific.

https://people.cs.rutgers.edu/~watrous/plus-signs-in-email-a...

37. asdff ◴[] No.42186118[source]
Do people use shared docs often in the workplace? I only used it on like two group projects in school and it probably made things more clunky than if we just wrote our portions and compiled them after. Maybe it works for some workflows but having multiple people editing the same document is chaotic, unless you delegate who does what, at which point there's no point in having it be a shared doc when the responsibilities are delegated.
replies(1): >>42187223 #
38. insane_dreamer ◴[] No.42186501{3}[source]
ok, but Search aside (which is Google's primary product; we were talking about side project), many of these are also-rans; they didn't really change the landscape the way Google Maps (and of course Search). OK, maybe Android, but that wasn't developed by Google. Neither was YouTube (groundbreaking), or Waze (not groundbreaking).

The only one I would take from your list would be Kubernetes and Google Earth, and Kubernetes being more of a dev tool would really count as far as impact and usefulness to society (Go would fit there).

Google Books _could_ have been great, but Google didn't take care of it. Same with Google Reader.

replies(2): >>42186807 #>>42192032 #
39. JW_00000 ◴[] No.42186581{3}[source]
Too late for most people.
40. hfsh ◴[] No.42186807{4}[source]
>which is Google's primary product

Which used to be Google's primary product, way waaaay back when. Their primary product now is advertising, and has been for a very long time.

41. mwest217 ◴[] No.42187223{3}[source]
All the time, it is incredibly useful to send a doc for comments, which can be attached to the relevant piece of text. I use shared editing less often, but I find it's especially useful in incident response where there may be multiple investigation workstreams, and the incident commander needs to be able to see all of them.
replies(1): >>42200736 #
42. guappa ◴[] No.42187586[source]
Nah google maps shows/hides things with very obscure logic.

Like you can ask to find a restaurant and it won't point you to the closer one but to one that is few km away instead.

replies(1): >>42192312 #
43. robertlagrant ◴[] No.42192032{4}[source]
I don't see why NIH affects anything. They're google products now. And as a further point: do you think Android and YouTube have the same codebases as they did when they were acquired? Of course not.
44. thepuppet33r ◴[] No.42192312{3}[source]
I think that's all based on advertising dollars.

One of the nice things about Openstreetmap is that it doesn't do that weird behind the scenes manipulation.

45. asdff ◴[] No.42200736{4}[source]
I’m shocked people are managing incidence response over a google dock… what happened to emailing things to people?