Something happened. A fire of some kind, which damages at least three of the containers in the same quadrant as the Roomba/Octograbber is garaged.
In the image below, far left edge, you can see the open "garage" door (pivots to horizontal), and all the containers in place behind/to the right of it:
Later, there was a photo, somewhere, showing the biggest container (the one to the left, elevated, with stairs leading to it) being removed by a crane. Can’t find that pic.
Pics since then have been overall views of OCISLY from a distance away, so in cropping to show just that area the quality is not great.
But in this one you can see that first container is gone but the two smaller containers on deck “behind” it are still there:
Then in this image, those two containers are gone:
Also of note, the vertical light blue object that is visible to the left in the above two photos, reaching down into the water. That is one of the thruster “Azipods”, which rotate and provide precision thrust levels to keep the ASDS on station, to within 1 meter accuracy. They are always raised up for the ASDS to be towed. On this trip back, it was not raised, it was dragging in the water. The hydraulic pumps and some of the control equipment for that azipod are in the same area as the removed containers, as well as a generator. So the fire knocked out at least part of that stuff, even if it might “just” have been burning the hydraulic and electrical lines that go to that azipod (more likely the pump system also got taken out and perhaps a generator too).
So, something definitely happened, a fire for some reason. Totally unknown what effect it may have on OCISLY being ready to go out to sea soon for the Koreasat-5 mission which is scheduled for October 30th, 11 days away (SpaceX has said nothing about this at all). Koreasat-5 is a heavy payload with a GTO orbital profile, so it’s going to be a “hot” re-entry, no boost back, with minimal re-entry burn, so ballistically the booster is going to coast LONG distance downrange. IIRC for such missions, OCISLY had left at least 3 if not 4 days before launch, to get into place.
So they’d need to have it fixed by end of next week, to be ready for an Oct 30th launch. If they needed a few days, I could see them delaying the launch for that reason. If they needed say 2 more weeks, they might really get into thinking about flying it as an expendable mission since the “hot” re-entry flights cause so much damage its not very practical to fly them again (I have learned recently that one of the FH side boosters will be one that flew a “hot” reentry…… although it had more of a re-entry burn than some others so it was not as “hot” as the rest have been. The other FH side booster also will be reused but it had a gentler RTLS re-entry profile).
They are so hard pressed to try (or by this point pretend?) to fly FH before the end of 2017, and needing 39A for yet another launch (“Zuma”, Nov 16th, just added!) before taking 39A offline to finally complete the FH pad modifications. So that is where the pressure in the launch schedule would mostly come from to possibly decide to do Koreasat-5 as expendable, if repairs to OCISLY may take too long.
I started saying last winter that I expected FH to fly the first quarter of 2018. And that still seems to be on track, even if OCISLY is fixed soon and everything launches as scheduled. Of course some of the most gullible of the “but Elon said” people still think FH will happen in November, ‘cuz he said so way back in late July. When he said that, that was when SLC-40 (destroyed in the AMOS-6 pad explosion a year ago) was supposed to be ready to launch Falcons by late September, so 39A would be available for FH mods to begin for 60 days. As of now, SLC-40 is slated to finally fly a Falcon, NASA’s CRS-13 resupply to ISS, NET December 4th.
By now even SpaceX says “Late December” for FH. IIRC, they said that before the surprise revelation of the November 16th “Zuma” mission (that's a code name, actual name not announced yet, presumably a secret military satellite). But they may have taken that Nov 16th launch into account when they said Late December. The track record of SpaceX launching even a normal Falcon in November/December without a delay of a few weeks, is not good. In many ways, scheduling-wise, it’s almost like “NovDecember”, one month of work crammed into two months. Though some of that is not SpaceX-unique - seems to happen with a lot of launches at the Cape that time of year, more than for other months (not counting weather delays, but actual rocket/pad/payload not ready for launch, or other non-weather issues). Indeed mostly a U.S. cultural thing that not as much gets done that time of year.