I haven't noticed any missing (but that doesn't guarantee it hasn't happened).
Is there some sort of placeholder, but no JPEG? Or does it just leave one less entry on a row; e.g, totally gone? Either way it might suggest some sort of PBase database corruption. Perhaps even a "minor repair" replaced a chunk of it with an out of date backup. IF (ha ha hahahaha) the site runs on some sort of serious commercial DB platform, mayhaps there are tools to do checks and repairs etc. (Oh sure, let me dream!!!
) I do have vague recollections that after one of the previous -- uh -- "disturbances," the management solicited reports of similar situations. That might imply there is some repair process.
Another thought if you have multiple instances of such disappearances, is there any time and date related info common to the files? Perhaps some process misbehaved way back when and was later fixed but not all data totally recovered. That could be a hint for where the mechanics should look in their logs. (LOGS?
)
"links-to-links-to-links-to-links-to-... "Like you, (after some ancient system crashes), I have a directory hierarchy here at home that matches my PBase galleries. Besides image files I have HTML files to incorporate layout, sequencing, and captions, plus the gallery heading blather. The image to image navigation is just back up and go forward to the next, but enough is there to rebuild the PBase stuff from scratch -- but all manual -- 'twould be a helluva job though! Since fear of crashes led me to that, I now create the backup here for new galleries before upload. That's vaguely useful for checking the sequencing, etc. before firing it up to PBase. (Of course the additional amount of work may account for why chunks of several major excursions seem to have escaped completion.)
So I just figure to not worry too much unless someone reports it or I trip over it myself. It's true though if an entry is rebuilt by the owner, the rebuild would lose the comments, but whatever!