I never use general-purpose search engines if I already know which website the thing I'm looking for is on (assuming that website has it's own search functionality.) If I am looking for a wikipedia article, I go to wikipedia and search for it there. If I am looking for a tweet, I go to twitter and search for it there. If I'm searching for something on HN, then it's algolia.

Search keywords in firefox make this easy. I can type "wiki train" into my address bar and it will take me directly to https://en.wikipedia.org/w/index.php?search=train From a usability standpoint it's just as streamlined as typing the same into the address bar and expecting Google to find it for you, except you don't need Google.

Googling everything by default is a bad habit, but one Google obviously likes to encourage. They want everybody to rely on them for everything.

It’s heavily dependent on the quality of the search offered. For example — Reddit’s search functionality is terrible. I’ve found that Google is much better for searching Reddit.

Google's propensity to return reddit results is a negative to me. Reddit almost never has the information I'm looking for, particularly ever since their website redesign. Even if the search result takes me to a reddit comment section that might have the result, reddit will only show me the top N% of comments in that discussion and then when I scroll down, I see irrelevant top-level posts on reddit about people's cat pictures or something. Yes I know I can install an extension to redirect me to old.reddit.com, but fuck that noise. Reddit answers are virtually never worth that effort anyway.

Consumer advice on reddit is worthless, the subreddits about products are all run by shill-mods paid by the companies making the products. Technical information on reddit is often confidently wrong. And for general inquiries, reddit is worse than yahoo answers ever was. Most of the "I'm a [profession] and I have [experiences and opinions]" posts on reddit seem to be "creative writing" exercises, aka liars motivated by the karma system to make shit up.

OP was saying that the search quality of sites vary and used the commonly cited example that Reddit’s search bar with “<search query>” is not as good as doing a Google search with “<search query> site:reddit.com”

You are describing the fact you don’t like to look at Reddit which is a bit different as OP was just using Reddit as an example for a general concept of using google instead of a site’s internal search functionality.

Incidentally you can solve your issue by typing “<search query> -site:reddit.com” to remove any results coming from Reddit instead of having to scroll past results you don’t want to see.

Don't think they were talking about google returning reddit results for generic searches, which doesn't happen that much and certainly isn't a highlight feature like it is for Bing and DuckDuckGo. When it comes to generic searches, google's biggest problem is SEO-farming blogs that are substantially lower quality than even the most inane reddit comment from someone pretending to be an expert.

Rather, when they are specifically looking for something on reddit, using google is much easier than actually using reddit. Counter to the above person's advice in using the site's specific search over general search engines.

Which makes sense, there are a lot of great posts on reddit when you know what you're looking for. Not so great for general advice.

Interesting. My experience is the exact opposite and I lot of my review searches are start by adding “Reddit” to the end.

For opinions on non-tech topics and products Reddit is my #1 source, and I haven’t got burned even once.

But I’ll admit it’s not my only source, and I always cross reference my findings.

I have setup a custom search under search engines in chrome/settings which basically maps !r to google+reddit e.g. : SE: GReddit Shortcut: !r https://google.com/search?q=site:reddit.com+%s

I have a similar one for github too:

gh : order by last updated, with > 5 stars, and has been pushed to since 1/1/2022.

If you log into your reddit account you don't need an extension you can opt out of the redesign and automatically and always see the old site design.

Or Amazon. Amazon's search sucks.

Results are often completely irrelevant, or dismissive of the search terms.

You search for a refrigerator filter from a known refrigerator manufacturer and the first results you get are the toxic offbrand refrigerator filters from companies with randomly generated company names that poison your entire family.

That combined with Amazon's aggressive category filtering is annoying. Search a generic term like rings, it forces you into the clothing category without the ability to switch. You have to then slap on more keywords in hopes it navigates to the right category, because manually changing the category while including generic searches does not work. It does not matter if you first went to the category you believe your search to be in.

Somehow Amazon has managed to make the bloated Aliexpress experience seem more straight forward and easy to navigate.

Amazon, despite having a high bar for hiring, has fallen victim to one of the most trivial forms of SEO.

Companies create multiple subsidiaries with randomly generated names, so they can spam search results and faceted results with their products and exclude other vendors. Search for "flying bird cat toy" and you'll see what I mean.

Much of that could be solved using a filter by country of origin. Because the usual suspects are always from the same territory.

Why is this necessary? because shell companies sell products loaded with lead, cadmium, phthalates, formaldehyde, VOCs, and other highly toxic stuff.

You may get better results searching the Pushshift archive, e.g. via https://camas.unddit.com. It will search deleted and removed items too.

Oh, man. My issue is that so many in-app searches are just.. bad.

Reddit is the easy example. I learned ages ago to never use reddit search, always use a proper search engine bound to the reddit domain (or even a subreddit).

With that said i don't use Google. I generally agree that it's an adversarial relationship but at the same time Search Engines often have focused tech that is just way, way better than some random search of some content site.

Reddit search is somehow worse than a simple `WHERE markdown LIKE '%query%' ORDER BY upvotes DESC` so you can understand where that sentiment comes from.

It's not what you're saying but it feels a bit unrelated in that it doesn't mean that it is not useful for Google to index tweets. More likely than one wanting to find a specific tweet, is one wanting an answer or pages related to a query where one does not care about where the result is hosted, and I would rather not search across several sites separately.

Not if they have a better search engine than most sites internal one...

Many websites just redirect to `https://google.com/search?q=site%3Awebsite.com+`.

Some other interesting things I've noticed is counters are often wrong, sometimes even with insane values such as -1. Replies missing from retweet views, missing search results, people have reported that replies sometimes don't work or never appear.

There are now always exactly 35 new tweets to load... on web anyway. Seems to be an increasingly buggy mess.

Isn't it $1bn in interest payments Twitter owes, on what regularity is that paid, every 3 months?

Yes, they made a $300m payment about a week ago.

https://www.bloomberg.com/news/articles/2023-01-30/twitter-s...

Re: the counter issue; I've had "2" tweets for at least 5 years. I really have 0 tweets, but they mention if you ever delete tweets in bulk, the counter can be off.

They confirm this can happen on their site: https://help.twitter.com/en/using-twitter/missing-tweets

Makes you wonder if anything security related has glitches like this.

> counters are often wrong,

I heard about "optimistic rendering"

-1 does seem bad. Reddit has fluctuating numbers I thought it was to fight bots or maybe using the eventually correct type db.

I never really saw how making the counters random helped fight bots.

Anyone who wants to know if their bot armies upvotes are counted can just choose 2000 articles, upvote a random half, then see if the half they upvoted have higher vote tallies than the half they didn't.

No amount of delay, quantizing, or adding noise will defeat that tactic. So why try at all?

I don’t think the technique would be meant to counter blind voting, but rather strategic voting—i.e. counter-voting some particular post down/up with a random member of your bot army every time it’s voted up/down, to keep its count high/low/at some constant value.

I say this, not because it seems like an especially common style of bot to be running, but rather because 1. it would be a rather heavy backend write load on the site if two bots doing this in opposite directions did exist, and ever “clashed” on the same post—and, much more problematically, a never-ending load, as the bots would never be satisfied; and 2. such bots do depend on the exact vote count passing some threshold, so fuzzing votes is a simple way to make such bots confused—not all the time, but probabilistically, enough of the time to make any such “clash” loops eventually quiesce rather than going on forever.

You can't efficiently tell which accounts work and which don't, so

a) you'll quickly accumulate more invalid accounts than valid ones, but have the burden of maintaining all of them.

b) you'll continue to provide more signal to reddit from your invalid accounts that can be used to burn your valid accounts.

I would expect it's a database consistency issue across some distributed NoSQL store.

I would imagine that's the cause too. But there are ways to prevent that. It is eventual consistency after all, right? As I mentioned in my other comment, I've had the wrong count for multiple years.

How long until they have a major outage?

google is, however, thoroughly indexing mastodon posts, if you have your server and account configured to allow crawling. enjoy

Twitter indexing was basically real-time - you could see a tweet within a few minutes of it being made

Mastodon seems to take many days to appear, if at all. I suspect Google just scrapes the webui occasionally and doesn't use any push-based mechanism for notifications. That probably means it misses some too.

HN is indexed in effectively real time as well. I've googled several phrases I've read here, hoping to find more information, where the HN comment of interest is the sole result.

Googled "HN is indexed in effectively real time as well" and your comment is the top result

I just had the opposite happen, was looking for a post I know ive found through google before and had to go through bing and then yandex to find it. Google's historical archive of hn appears wonky, or at least did that day.

Delta of 7 minutes. I thought about putting in a nonsense phrase to test the hypothesis, but apparently that wasn't even necessary. Spooky.

Odd the same search from my end brings up 0 results.

This spooked me the other day. Made a comment about a phenomenon I knew of and the name I knew it by. Got curious whether other people knew it by that name too, googled the name, and the top result was my own comment made just a few minutes ago.

Any server recommended for general purpose quiet lurkers?

I am finding that this is not the case when I'm searching.

Yesterday I read this tweet (below) that talks about some military history relevant to current events. I wanted to read more so I googled `"1957" china balloon` and the first hit was the very same tweet. Testing now, searching for a substring from the tweet in quotes works too.

https://twitter.com/NavalInstitute/status/162163973828837376...

Next week, Twitter is starting to charge for their API https://arstechnica.com/tech-policy/2023/02/cash-strapped-tw...

Could be related? Even if they’re just crawling the public pages I could certainly imagine Elon trying to extract money from Google for the privilege.

Just a few days ago I wrote a comment, that Elon's decisions are pushing Twitter towards becoming a ghost town. But I'm unsure whether he could be so irrational to cut the biggest search engine off.

One thing I find important to remember, is that Jack Dorsey gave Elon Musk one bil for the Twitter purchase so maybe there's some kind of Grand Plan that I'm too stupid to see. But would it start with Twitter's bankruptcy?

I'm not really a conspiracy guy, but it could be a 4d chess move.

He can get publicity whenever he wants it, how he wants it. Twitter is the #1 platform for news, and political ideologies to express themselves. If he levels that, it throws a huge wrench since there's really no 'backup'.

Everyone has put so much into tweeting on twitter, now that it could go belly up, how does an entire network pick a good replacement? Doesn't even look like they can because of all the polarization.

If he had some evil designs, tanking Twitter could be worth losing 45 billion dollars if it gives him more power to do whatever he wants to do. Not saying he is, I'm not an Elon fan, but even I don't think he's this stupid (he's not donald trump) - but I guess he could be - if he isn't doing this on purpose he'd have to be.

My understanding was that Google was paying for its firehose access already. Though I suppose they could be forcing a contract renegotiation.

> Though I suppose they could be forcing a contract renegotiation.

This was my attempt at an explanation. I wonder who will blink first and if the public will ever learn the details.

Yeah, also when searching for a personality, the second result would always be a "Tweets by" carousel. Interesting.

You can add site:twitter.com to the google search to restrict the results.

"1957" china balloon site:twitter.com gives 677 results so that might explain why you don't get the result you want with a general search. without a _site_ filter its going to have to pick only one or two tweets

My recollection was that the deal to index and surface tweets on the SERP ended years ago. Was that the deal you recalled?

Now that I look, yeah the deal I was thinking of dates back to 2015. Did not know that expired. But is the Google(Twitter) search experience the same for you right now as it was in previous months?

Google has become a major force of censorship. Twitter (under Elon Musk) is now opposed to that. That should explain what you experience.

Censorship seems alive and well on Twitter under Musk.

https://www.motherjones.com/media/2022/12/elon-musk-twitter-...

https://www.nytimes.com/2022/12/16/business/media/elon-musk-...

https://www.msnbc.com/opinion/msnbc-opinion/elon-musk-twitte...

One good change that happened on Twitter recently is that the login nagscreen now has a close-button.

Part of my job involves investigating stuff that's often behind a paywall, but the paywall is leaky enough that Google Search displays snippets of what I'm looking for in the results. So I'm often grateful to see these paywalled results come back because it means I've hit paydirt, and if Google weren't indexing the snippets, I would have no visibility into this stuff, and it'd be a huge hindrance to me.

Under the same standards, HN (and many other sites) wouldn't be indexed either.

HN doesn't have a login wall to read anything.

But tweets are on the public web without login.

It changes over time, sometimes there will be a dismissable popup, for a while it couldn't be dismissed. It does seem to work now for non-nsfw tweets, but nsfw login-walled ones are indexed by Google too I believe.

Specific tweets, yes, but if you want to scroll a person's profile, you'll need an account after about a page of scrolling.

no, just close that popup? if it even appears... when is the last time you've used twitter

Logged out - about two weeks ago, couldn't close the pop-up, no X button. Perhaps it was changed recently but doing it on mobile was simply not an option for a long time.

Not yet.