All tests look good but DKIM still doesn't sign

I had to transfer my mail to my facilities due to restrictions with payments made with credit cards (thanks to socialism, I live in Argentina). I have removed several blacklists and solved many problems that mxtoolbox remembers.

What happens here is that I can't manage DKIM to sign my emails

I tried my domain against Mail Tester, MxToolBox, and everything looks good, but when I try it against It simply says

"This message does not contain a DKIM signature"

What am i missing? I have a DNS noip in the middle, do I have to load my DKIM records there too? Because I have my domain registered in Globat and I loaded all my DKIM records there. Maybe that is the problem?

My domain is and the DKIM selector I'm testing is called Luke

Thanks for your attention,

moderation – exploratory tests – Exchange of user experience stack

I am doing some interviews with similar clients this week. It is not really a usability test, there are no tasks, but I will show customers an initial phase design and tell them about their needs. So, it's more of a focus group interview, except one by one.

I was skeptical about this type of user research, but it has given us some good ideas in the past.

Even if you don't have a prototype to show the customer, you can point to a similar user interface on another site and ask about it. Let them explore the new interface, then ask:

  1. What's it for? That makes?
  2. If you take an X action, what do you expect to happen next?
  3. Is this thing / flow / UI something you would use? Why or why not?

It is not quantitative, but it can help guide new developments.

lo.logic: examples of tests that use induction or recursion in a large recursive ordinal

There are many tests of induction or recursion use in $ omega $, or in an arbitrary ordinal (may be uncountable). Are there any good examples of tests that use a large but computable ordinal?

The original proof of Ramsey's theorem and Hales-Jewett's theorem uses induction in $ omega ^ 2 $, but the use is not essential, because Erdos and Shelah have given better limits when using induction only in $ omega $. Plus $ omega ^ 2 $ It should not be considered large.

A typical use of large ordinal induction is to demonstrate the consistency of axiom systems, for example, using $ varepsilon_0 $-induction to test the consistency of PA. This is a kind of examples.

The existence of the Goodstein function uses induction in $ varepsilon_0 $, and I think it's just a direct explanation of how recursion works in ordinal.

Are there more examples?

Ab tests: what version of A / B should I show if you get different versions in two instances?

Let's say I am doing A / B tests, and this particular user is randomly assigned to group B, for example. I have the option to save it in your account in my database or in your browser. Then, tomorrow, that same person visits the website from another device, randomly assigns group A and sees version A of the website. And then, from this new device where they were watching version A until now, they now log into their account.

Would it be correct to now load all B, which has potentially been stored in your account? Or would it be correct to persist in the current visit and continue to display A in this local browser? Or, go so far with persisting in the new visit to actually store this new version A in your account for any other page processing from your first device.

All this assumes that the different versions of the site differ visually and significantly, therefore, what the user expects is, therefore, quite important.

Need help with Treejack tests

I need help to organize this website that new employees and contractors will use. I am testing to see if new users can find what they need on this site. Take this quick test in Optimal Workshop, it only takes a few minutes. Any comments and comments will be appreciated.

Thank you.

usability tests: what are the disadvantages of asking your current users to participate in a study (for example, for new functionality)?

What I mean is that you rarely see companies that place some type of banner on their homepage that says:

"Hey, we are developing some new things and we would like to know the opinions of our customers. Take a quick test!"

They usually do those tests behind closed veils and with specifically filtered testers (for example, in remote tests).

So my question is, what is the main reason for that?

These thoughts come to mind as arguments against that:

  • Only specific users would participate and skew the perspective. The average user, who is the main buyer, will ignore it; while the power or frequent users will take the opportunity to complain or inject their very specific desires.
  • Its users are also accustomed to the site and will likely see any new design changes as "bad" and will vote to keep what they already know.

On the other hand, these are the things that could be said in favor her:

  • You can get information about the problems that only long-term users may face, while new users don't even know they exist.
  • You generate more loyalty due to the confidence you show for your opinion.

I guess I answered my own question, since it generally depends on the context. Use new foreign users to try things that are supposed to attract new paying customers, while you can ask existing users when you want to improve the deeper and more complex functionality.

But it seems that you never see the second case, but almost always the first.

Does anyone have more points to add to the two lists or other thoughts related to that?

software – Tests can detect the presence of error and not the absence of error, why?

I listen and see this statement in almost all academic books related to software engineering.

The tests can detect the presence of error and not the absence.

But I do not understand clearly. What can be your possible explanation?

Does it mean that it does not tell us what are the characteristics of the software that remain to be added?

How to form equivalence classes for equivalence class tests for the following

they are only the equivalence classes that include the age ranges mentioned in the image with the invalid ranges that are the age<16 and age>100 good enough or anything else comes into consideration when forming equivalence classes
enter the description of the image here

Do not run unit tests in visual studio 2017

in the visual studio output window shows

"No evidence matches the test case filter
  provided "

not even creating by the wizard the class that you want to debug / test.

What could be happening?

html5: integrated YouTube videos make my site fail Google mobile optimization tests

I have a website with a responsive theme optimized for mobile devices.

Despite that, many pages do not pass Google’s “mobile optimized test” (

Here is an example page affected by this problem:

Here is the result of the validation attempt in the Google tool:

enter the description of the image here
I discovered that all pages on my site that cannot pass this test contain an embedded YouTube video.

If I delete the video, each of these pages passes the test.

The video is added with the following code:

My question is: is there a way for my mobile page to include a YouTube video?

I would also like to know if it is normal for Googlebot not to analyze CSS and JS, and show the fragment "your browser does not support this function" instead of the video.

Thank you!