John Schettler - Anvil of Fate

“That’s entirely possible,” said Maeve. “Some sources claim the battle was fought October 10th.”

“But the consensus was that the battle was fought on October 25th, 732” said Nordhausen. “Even the Islamic sources seem to indicate it was fought in the year 114, the first day of Ramadan, yuam-as-sabt. That’s the 25th in the Julian calendar… You did use the Julian Calendar, correct? Because if you put in Gregorian numbers Paul would have arrived at the wrong time.” He gave Kelly a suspicious look.

“Yes, I used the Julian calendar,” said Kelly. “You said it was on a Saturday, near the end of the month. In the Gregorian calendar that would be either the 22nd or the 29th. In the Julian calendar it would be the 25th, the day you wanted. Want to look at the algorithm?” Kelly anticipated a battle with the professor, and he was stealing a march on him, choosing ground where he had a decided advantage.

The professor pursed his lips. “Then they got it wrong,” he said dejectedly, waving his hand dismissively at unseen historians the world over.

“Notice he didn’t say “I” got it wrong,” said Kelly, taking a little swipe at the professor for the static he had received earlier.

“Well I can only report what is known about this period,” said Robert in a huff. “And the sources are pretty thin.”

“Look, this is getting us nowhere,” said Paul. “What do we do about this? Do we select another date and try again, possibly a few days earlier or later?”

“You’re saying you didn’t see anything?” Maeve asked. “No sign that an army had marched through that area? We put you right near the old Roman road.”

“Correct,” Paul confirmed. “The ground would have been rutted with the wheels of their carts and wagons. All accounts are that the Arabs were heavily burdened with their plunder. But the whole area seemed completely undisturbed. No horses, carts, tents, soldiers, campfires, banners, and no sound of fighting. Believe me, estimates are that there were upwards of 30,000 men on each side of this battle. You put 60,000 men into that location and I would have certainly seen something. And the Arabs brought wives, family, slaves, and personal possessions as well. It would have looked like Woodstock after the concert if they had been through that area at all.”

“Yes, if we shifted you in too late then you would have seen signs of the battle, not to mention dead bodies all over the place, even if it was fought as early as the 10th as one source suggested.”

“Right,” said Paul. “So we were early then. The battle had to be fought later. It’s the only thing that could account for the unblemished condition of the ground I saw—unless we’re entirely wrong about the location of this battle, and I think that is unlikely.”

“Then we’ll have to take a look at the following Saturday,” said Nordhausen. “Make it November 1st, Kelly.” He folded his arms. “I guess I’m up then,” he said with some anticipation. It was clear that he was eager to get a look at the situation first hand.

“Paul?” Kelly looked at the project team leader for confirmation.

“If that's the case both armies would still be on the field at that time.” Paul was obviously not happy, but there was nothing else they could do at this point. “I suppose we’ll just have to try again as Robert suggests,” he said dejectedly. “What other option do we have at this point?”

“Alright,” said Kelly. “This shouldn’t take long to program a small variation in the temporal coordinates like that—twenty minutes at the outside. I’ll set it up and get the Golems to verify the numbers ASAP… Which reminds me…”

He suddenly remembered those lost sheep in the Golem history variance module. A small segment of his installed user base did not respond to the command to join the network cloud. He assumed they must be from systems that came on line after his command was sent to the active units. While the others continued discussing the situation, he focused his attention on the alert system, trying to get a fix on how many sheep were still outside the pen. Something caught his eye at once, and he frowned.

“Hello?” he said aloud. “Now what’s this all about?”

Paul looked over his shoulder at him. “A problem?”

“Well I had a few Golems that didn’t join the Network,” said Kelly. “And they’ve been continuing to feed data to the alert module all this time.”

“Anything serious?” Paul came over and leaned in to have a look. They were looking at the 8th century, by decades, and the monitor still showed the obvious demarcation from green to lighter shades of avocado yellow right around the first three cells.

“Just for yucks, let’s zoom way in to the year itself and get a more fine tuned look at the data,” said Kelly. He keyed in 732 for the desired year, and now the screen put up the twelve individual months on the horizontal line. They immediately saw that something was wrong. The color had changed to amber as early as January!

“Hold on,” said Paul. “That was green when we looked at this earlier. We didn’t start seeing color variance until late in the year, right around the date of this battle.”

“Well now the whole year is grade 2 yellow,” said Kelly, “and this is grade 1 amber here.” He pointed to October on the line.

“Scroll back,” said Paul. “Show me the previous year.”

Kelly scrolled the chart and the yellow remained all through the months of 731. “This is really odd,” he said. “Let’s zoom back out to individual years now.” When he refreshed the screen they could clearly see that the alert had migrated to the left of 732. There was yellow all through the 720s, slowly fading from chartreuse, to pear, then avocado, apple and shades of olive green.

“Can you fine tune this even more?” Paul asked. “I mean, can we see numbers to give us a sense of how much variance we’re seeing in these color shifts?”

“No problem,” said Kelly. “Here’s the data in decimal readout.” They scrolled back.

“Hell, we’re seeing variations much earlier now. When did it fall out of nominal ranges?”

Kelly ran his finger back. “Here,” he said definitively. “The year 705. It was holding well above 99% on all prior years. Then it starts to show variation in that year and it just degrades from there.”

“Why didn’t we see this earlier?”

“We’re lucky to see it at all,” said Kelly. “It looks like things have been changing all along. Tours was not the origin of the first major variance. At least according to my lost sheep,” said Kelly.

“How many? Is this based on one report or a stronger weight of opinion?”

“Let me see if I can call up some of the actual documents that no longer jive with our RAM Bank.”

“Maeve, Robert, you better come over here.” Paul waved at them where Maeve was fussing with Robert’s headdress across the room.

He explained what they had found, clearly concerned. “We have a problem,” he finished. “Apparently a few of Kelly’s stray Golems seem to think the variations begin here in this range now.” He pointed to the year 705 on the screen. “We don’t get a safe nominal green until August of that year. In September it begins shifting. The numbers confirm it.”

“Nothing really shifts out of yellow into the orange spectrum until the battle of Tours though,” said Kelly as he reviewed articles, “but we get precursor variations all through the years prior to the battle now… almost like foreshocks to the big one at Tours.”

“Robert?” Paul looked at the professor. “Could we have jumped the gun here and missed something? What was happening in the years before Tours?”