I’m seeing this fairly consistently and wanted to report it, and see if others are seeing it.
I’m consistently seeing an empty Original Post if that post is the first one in the forum listings and it has no replies. Once the thread gets bumped from the top, the OP is visible. I’ve seen this twice today.
I just noticed it in Jack Batty’s Pit thread. No OP visible at first, but when I went back to the forum menu and reopened the thread, it was there in all its glory. Weird.
If it happens again, can someone do a “file->save” while it is happening, and e-mail the saved html file to me? I may be able to help the hardworking staff at the SDMB find out what is going on.
This is a complete WAG. I know nothing of the way that VBS works.
Could it be that, as it seems to be taking a very long time to submit posts, that people are trying to view threads that have not been completely submitted yet? I mean you hit the Submit button, and it seems to take a long time for the submittion to go thru. perhaps Vbullitin has created the page referance, but hasn’t added the content yet?
Thank you very much, I also received one from beatle too.
I can see no obvious errors or corruptions in the code returned. It just seems like the message block for the OP simply does not exist at that time, and the table row isn’t being created. Looking at how the page is generated, it would seem that the query of the OP is just returning “NULL” from the database, which unfortunately means that once again I am no help at all, without being in front of the server proper.
It could be a timeout, or loading issue, but it seems like there would be more problems in the page.
The database redesign to increase performance had an unintended consequence that can cause threads to appear before their OP has been committed to the database. This has always been the case but the time interval between the two events has always been relatively small. Now this time interval can be in the seconds depending on the posting/searching load on the database. We’ll never be able to eliminate this possibility but by reordering the code we can reduce it’s likelihood of being noticeable. Of course, we could always just go back to the days of terrible performance…