« first day (1657 days earlier)      last day (3287 days later) » 

12:45 AM
I'm agree with that, but in cases like the one above exceptions are the unique way to prevent race conditions.
 
how can a post be deleted during that extremely short time? we are talking about fragments of a millisecond.
 
 
10 hours later…
10:28 AM
On a system I developed (the client was someone that frequent this site), there is a CPT used to track user votes from frontend. The website had always an average of 500+ user connected at same time. If two user clicked in same time a button (and it's not so impossible) is not possible that two votes are processed in the exact same moment, one is processed firts, even a fraction of millisecond first. I know that because it happened.
I didn't see that happen, but I saw the results, and after few hours I come to the conclusion that the only possibility was that.
 
The bug here is that a CPT was used to track votes. :)
 
 
3 hours later…
1:43 PM
race conditions!
 
 
6 hours later…
7:20 PM
Anybody that keeps debug.logs on - notice any errors with undefined index HTTP_HOST ?
My canonical.php is flooding my logs with it recently.
 
7:49 PM
backtrace?
 
@toscho has this marketpress.com/support/topic/… moved anywhere? what's the current strategy to filtering translated taxonomies and post archive slugs
i think there was just that simple text url filter, essentially would have to build my own mappings
 
add_filter( 'mlp_linked_element_link', function( $url, $site_id ) {

    if ( 1 === (int) $site_id )
        return str_replace( '/projekt/', '/project/', $url );

    if ( 2 === (int) $site_id )
        return str_replace( '/project/', '/projekt/', $url );

    return $url;

}, 10, 2 );
 
right, yes, that
 

« first day (1657 days earlier)      last day (3287 days later) »