there’s been advisories from the govt and security agencies telling people to be careful with google - the ads are blatant malware a lot, google isn’t even really trying to do anything about it.
^^^
1 Like
fwiw 99% of my interactions with microsoft are azure (where they are a major underdog to aws) and github (which is sort of independent though I think we’re starting to see some tightening).
2 Likes
first time working with github a lot and I hate it. You can’t even partial match when searching, so if you want to find something like “host” it will only match that exact string, but not for instance if it’s surrounded by quotes in the code. It’s so useless I wonder why it’s even there.
lol this is totally unnecessary bundling exactly like MS got in trouble with the feds for 25 years ago, except I guess now no one gives a shit about that stuff?
Imagine Google being like “you can only use Google search on Chrome”, maybe right now that would be a misstep but several years ago that would have (likely illegally) crushed every other browser out of existence.
google paid $20billion for google to be default search engine on ios lol
1 Like
the funny thing is, unless google pivots hard, and I have read some good analyses that suggests they can’t, I think bing has potential to take a big chunk out of them and google’s definitely a bit shook.
Microsoft’s AI-Boosted Bing Can Run Rings Around Google Search - CNET
to me it definitely seems better even right off the bat using it. I didn’t realize how many dumb/bad google results I just mindlessly scroll through when I use it.
or googling something like “X Y Z” and every result is like:
does not include: X Y Z | must include: X Y Z
As documented here
Trumpbot is currently busted.
One thing that’s puzzling for me is that even when routing through a proxy, I see different results on my home computer versus on AWS. Like I can run
curl --proxy proxy-server.com --proxy-user myaccountinfo https://truthsocial.com/...
on my laptop and it will (often, depending on the rotating proxy) work, whereas the same command on an AWS box will never work. I would think the use of the proxy should make the source of the request irrelevant, because Truth should see it the same & coming from the proxy in both cases, but clearly that assumption is wrong; anyone know why/how that’s the case? The answer might help resurrect Trumpbot.
1 Like
try that proxy with https://httpbin.org/ip on both and see what comes back to verify
could be using some kind of device fingerprinting or looking at your headers too? if so, maybe a different vm would work?
1 Like
fine letting it die
2 Likes
On some days I might understand this, but not today!
Adding to my confusion is that, on my laptop where Truth works, I’m now getting errors running what should be the exact HTTP request depending on whether I use curl (fine) or Go (not fine). Like I would think this command
curl -H 'User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:109.0) Gecko/20100101 Firefox/110.0' 'https://truthsocial.com/api/v1/accounts/107780257626128497/statuses?exclude_replies=true&min_id=110113808184666539&with_muted=true'
should have exactly the same output as this program
package main
import "io/ioutil"
import "log"
import "net/http"
func main() {
rawURL := `https://truthsocial.com/api/v1/accounts/107780257626128497/statuses?exclude_replies=true&min_id=110113808184666539&with_muted=true`
req, err := http.NewRequest("GET", rawURL, nil)
if err != nil {
log.Fatal(err)
}
req.Header.Set("User-Agent", "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:109.0) Gecko/20100101 Firefox/110.0")
resp, err := http.DefaultClient.Do(req)
if err != nil {
log.Fatal(err)
}
defer resp.Body.Close()
data, err := ioutil.ReadAll(resp.Body)
if err != nil {
log.Fatal(err)
}
log.Println("output:", string(data))
}
but they do not. The curl command works for me, the program says (this is Cloudflare blocking):
2023/03/30 15:47:15 output: error code: 1020
I got Trumpbot working locally by literally adding a exec.Command("curl", ...) inside the Go program lol, so dumb
i’m sorry man. go ahead and ‘rm -rf /*’ and reinstall.
I just do the javascript
might give some hints as to why. I’ll go look at my cloudflare stuff tonight and ask my hacker friend a few q’s. no way we get outsmarted by truth, lol
Yeah I looked this up too and came across similar results (if not this exact one) suggesting the main difference would be the handling of redirects, but I don’t think that’s in play here. I tried one of the solutions for disabling redirect following in Go anyway and it didn’t change anything.
@TheDudeAbides had some good advice offline to try messing with TLS stuff - it seems like forcing TLS 1.2 (instead of 1.3) fixes both of these problems:
- requests thru proxy behaving differently on AWS versus locally, and
- curl and Go having different results locally
I guess TLS 1.3 makes it easier for Cloudflare to block stuff? idk 
1 Like
1 Like
just realized i saved the trump bot though 
1 Like
1 Like
In TLS 1.2 and earlier versions, the use of ciphers with cryptographic weaknesses had posed potential security vulnerabilities. TLS 1.3 includes support only for algorithms that currently have no known vulnerabilities, including any that do not support Perfect Forward Secrecy (PFS) .