Unexpected behavior for Facebook Sharing [duplicate] - facebook

Question

This question already has an answer here:
Facebook ignoring OG image on first share
(1 answer)
Closed 6 years ago.
First of all hi and thanks in advance to anyone who can help with this because I've been going crazy over this for weeks now.
So I've got a website which lists gif taken from my mobile application (which are then stored on AWS and my visitors ( I haven't found a use for me to have users) can share these gifs on facebook using the facebook sdk.
The problem appears when I try sharing an image for the first time
This is what the share dialog shows the first time I click on my sharing button:
http://i.stack.imgur.com/lNVNF.png
and then I close and reclick the same button and now it works:
http://i.stack.imgur.com/YsDUm.png
Now I've been trying to find a way to make this work on the first sharing attempt but to no avail.
I am using meteor in combination with biasport:facebook-sdk and Amazon S3 for the hosting of my files.
Edit here is the code used:
FRONT SIDE
HTML
<div class="facebook share">
<img src="/gallery/fb.png">
</div>
Javascript
Template.*templateName*.events({
'click .facebook': function(e){
e.preventDefault();
e.stopPropagation();
// this is in a modal so I store the data I need
// (events have photos which in turn contain a url to the gif
var url = Session.get('event').photos[Session.get("id")].url;
FB.ui({
method: 'share',
href: url
});
}
SERVER SIDE
JAVASCRIPT
if(Meteor.isClient) {
window.fbAsyncInit = function() {
FB.init({
appId : 'APP_ID',
status : true,
xfbml : true,
version : 'v2.5'
});
};
}
Edit: I found a manual solution using exec future and curl
so first I added a call to a meteor method on the share that updates the facebook crawler
JAVASCRIPT
Template.*templateName*.events({
'click .facebook': function(e){
e.preventDefault();
e.stopPropagation();
// this is in a modal so I store the data I need
// (events have photos which in turn contain a url to the gif
var url = Session.get('event').photos[Session.get("id")].url;
Meteor.call('updateCrawler', url, function(){
FB.ui({
method: 'share',
href: url
});
});
}
Then I defined my meteor method as such
JAVASCRIPT
Meteor.methods({
updateCrawler: function(url){
var future = new Future();
cmd = 'curl -X POST -F "id=' + url + '" -F "scrape=true" -F "access_token={my_access_token}" "https://graph.facebook.com"';
exec(cmd, function(error){
if (error){
console.log(error);
}
future.return();
});
future.wait();
}
});
it's ugly but since I'd have to wait for the crawler to update and it works I'll leave this here for future use for someone maybe
Edit2:
I did not use og tags at all since I was simply sharing a url to aws directly and not a url to my website

Answer 1

I worked around this problem by calling the Facebook API direct from the server to make it scrape the og data by requesting info on the page. First time round it doesn't have the image cached but second time it does so this workaround does the initial call before sharing.
Use an access token for your facebook app and call the below in an ajax call and await the response before opening share dialog. Replace Google address with your own uri encoded address https://graph.facebook.com/v2.5/?id=http%3A%2F%2Fwww.google.co.uk&access_token=xxxxx
EDIT:
As per comments, here is my server side method for calling this which I use when posts etc are inserted to make the initial call and prompt a scrape from fb:
var getTheOGInfo = function (link)
{
if (!link || link.slice(0, 4).toLowerCase() != "http"){
throw new Meteor.Error("og-info-bad-url", "Function requires an unencoded fully qualified url");
return false;
}
var url = "https://graph.facebook.com/v2.5/{{{{id}}}}?access_token={{{{token}}}}&fields=og_object{id,description,title,type,updated_time,url,image},id,share";
var token = Meteor.settings.private.fb.token;
if (!token){
throw new Meteor.Error("og-info-no-token", "Function requires a facebook token in Meteor.settings.private.fb.token");
return false;
}
var link_id = encodeURIComponent(link);
url = url.replace('{{{{token}}}}', token).replace('{{{{id}}}}', link_id);
var result = HTTP.get(url, {timeout:1000});
return result;
}
Or for your purposes you may not want anything that might be blocking so you could change the last two lines to be aynchronous:
var result = HTTP.get(url, {timeout:1000});
return result;
//Replace with non blocking
HTTP.get(url, {timeout:1000}, function(err, result){console.log('something asynchronous', err, result);});
return true;