What does it mean to say that x is true for Sally but not for Jim? What sort of statements could possibly satisfy this schema?
Suppose we admit that knowledge implies truth, i.e., that "Sally knows that x" implies that "x is true".
Does it follow that if x is true, Jim must know it? Obviously not. It might very well be true that Sally is on the train to Tokyo, but Jim doesn't necessarily know it. Indeed, Jim might not even know that Sally exists, let alone her present whereabouts. Nevertheless, there are truths about Sally which can be asserted.
So at least at first glance it seems that, while knowledge describes a subjective condition, truth does not. What's true is true, whether or not anyone knows it to be true.
Next, when we say for example that x was true yesterday but not today, aren't we usually describing a contingent state of affairs? "It's raining" was true yesterday, but not today. But the statement "It's raining" has an implicit time index = "now". If we reformulate the statement to make the time index explicit, say, "It's raining at time t" then there is no problem seeing how the truth-value might change with different values of t.
Similarly, if all that is meant is that there are truths about Sally that are not true about Jim -- for example, that Sally believes that x, while Jim does not -- then there is really nothing surprising here (or worth arguing about.) Logically, it's like the fact that this ball is red while that one is blue.
The same goes for the fact that a hundred years ago most people believed that x, but now most people don't.
To say that x is always true is to say that for all times t, x is true at t. There is nothing inherently problematic about this schema, as far as I can see, although determining whether it applies to any given statement will often require some investigation and/or thought.
And again (using the the usual meaning of the words "know" and "true") what is always true is true whether or not anyone ever knows that it is.
So it might be an interesting question for historical research whether people at time t believed that x, and it might be interesting to explore the reasons they gave for their beliefs, but otherwise it has no direct bearing on the truth of x. (Dittos for what people have always believed.)
If you want to assert something different from this, I think you owe us an explanation of how you are using words like "knowledge" and "truth".
When people say x is true for Sally but not for Jim, perhaps what they're intending to say is that x doesn't have the same meaning for Sally that it does for Jim?
But if Sally and Jim aren't talking about the same thing, there really shouldn't be any puzzle about the fact that they assign different truth-values to x.
What this situation calls for is for Sally and Jim to explain what they mean by x. Once they do so, the apparent conflict between them will usually be resolved.
Many disagreements (but not all!) are of this kind. We must always be sensitive to the possibility that others are using words differently than we are. (I wish I followed this advice more often myself.)
(reposted from a thread at the Hermitary, with some revisions)