Synchronous XMLHttpRequest on the main thread is deprecated because of its detrimental effects to the end user's experience.

[Blake Niemyjski]

I get the following error in chrome when loadSource is called:

Synchronous XMLHttpRequest on the main thread is deprecated because of its detrimental effects to the end user's experience. For more help, check http://xhr.spec.whatwg.org/.

function loadSource(url) {
        if (typeof url !== 'string') {
          return [];
        }

        if (!TraceKit.remoteFetching) { //Only attempt request if remoteFetching is on.
            return '';
        }
        try {
            var getXHR = function() {
                try {
                    return new window.XMLHttpRequest();
                } catch (e) {
                    // explicitly bubble up the exception if not found
                    return new window.ActiveXObject('Microsoft.XMLHTTP');
                }
            };

            var request = getXHR();
            request.open('GET', url, false);
            request.send('');
            return request.responseText;
        } catch (e) {
            return '';
        }
    }

[Blake Niemyjski]

I think this code may also have issues on IE 8/9... I'd like to abstract this code so we could inject node specific / web specific logic to get the source.

[Blake Niemyjski]

I've battle tested some code here: https://github.com/exceptionless/Exceptionless.JavaScript/blob/master/src/submission/DefaultSubmissionClient.ts#L56-L161 We should take a look and slim this down to fit this use case.

[Yavor Georgiev]

Do you recommend we leave remoteFetching off in production until this is addressed?

[Blake Niemyjski]

Yeah, I think we should make this the default, as it could cause extra requests to be made

[Blake Niemyjski]

Can you submit a pull request for this.

[Adi]

Why do you fetch with a sync request anyway? :)

[Blake Niemyjski]

@adi518 I didn't write this logic but I assume it's because it has to work in ie6, we try and have it work as expected on all browsers.

[Vivek Marakana]

Also, we need URL validation before making call. Sometimes error stack is not valid and that leads the call to eval at globalEval (http://cms.pharmeasy.in/assets/a929fb88/jquery.js:2:3123), <anonymous> for context

[Blake Niemyjski]

@vivekmarakana would you mind looking into this?

[Vivek Marakana]

sure

[Victor Musienko]

Hey, any updates on this? Maybe drop support of ie6 (and maybe 7,8,9)?

[Blake Niemyjski]

I never heard back, would you mind looking into it? I don't see us dropping support for IE anytime soon. (https://www.netmarketshare.com/browser-market-share.aspx?qprid=2&qpcustomd=0) there are probably ~billion?? at least in the millions users still running those crappy versions of IE and we need to detect errors still for them :(.

[Marcelo Ogeda]

I recently received this as a warning level in Chrome version 63.0.3239.x

[Blake Niemyjski]

@marceloogeda Would you mind submitting a pull request for a newer implementation that falls back based on support?

[Marcelo Ogeda]

@niemyjski sure, I'll make the required changes and submit it.

[Blake Niemyjski]

Might be able to take the code from here: https://github.com/exceptionless/Exceptionless.JavaScript/blob/master/src/submission/DefaultSubmissionAdapter.ts but we need to be sure we try catch any errors.

[Blake Niemyjski]

Can anyone submit a pr for this.