feat(core): Return array or string from getTraceMetaTags()

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201

[s1gr1d]

Changing the function signature for the convenience of returning a string or an array of strings.

[Lms24]

Let's maybe wait a bit with this PR as #13281 makes some changes to the API

Generally I'm not sure if we should move forward with this change for now for the reasons outlined in the PR description of #13201

[Lms24]

if we move forward with this, we can adjust the types so that TS knows that if asArray is true, it returns an array and otherwise a string

[s1gr1d]

The type checking should already check this as I created two function overloads above. But I am open for another approach (in the case this gets merged) :D

[Lms24]

sorry, I missed this. This is what I had in mind :)

[s1gr1d]

I'll close this for now as this is a legit reason not to return an array:

I don't want to hand them an API that suggest they could pick one. Also, returning the API would make the API a tad less convenient to use and there's precedence for the simple string approach.
-- from #13201