Roedy said:
Why? because I had done experiments and I had results. I presumed
others had not done any experiments, or surely they would have got the
same results. Others seemed far more interested in putting me down
than doing any experiments. I just assumed you were a stupid jerk who
was flatly refusing to test this out for yourself and was simply
pontificating that I was wrong. Once an idea like that gets locked in
my head, it takes something quite major to shake it.
"Why?" Presumption, assumption and stubborness... there was quite enough
evidence that was fairly presented in contradiction to your claims.
I have yet to sort out how I could have got different results earlier.
I did the experiments repeatedly because I was so surprised by the
results.
Which is why we (I) asked for a fully specified experiment. That would have
included file name, paths, versions, etc. Just specifying the exact steps
and conditions would probably have resolved this far more quickly in your
mind.
My theories now are that the my strange early results were caused by:
1. I was actually editing files outside eclipse and copying them in
rather than editing eclipse's files in place. Perhaps the files I was
replacing were somehow dated older than the ones Eclipse last had so
it did not notice the changes.
They aren't "Eclipse's files". Eclipse opens the file from wherever the
Project path points.
2. Perhaps I actually had eclipse running, but supposedly not using
the files in question.
3. Perhaps I actually had eclipse running, but the files it was
supposedly not using were just offscreen.
4. Perhaps you have to do the experiment very quickly or slowly to get
the anomalous results.
5. I was in an alternate reality where Eclipse is a true SCID.
6. I was copying my changed files to some bit bucket rather than the
eclipse working directory. At this point, this seems the most probable
explanation.
Sorry for leading you all on this wild goose chase.
If you really care about what may have caused your anamolous results, fully
document each experiment. I suspect you won't have it happen again unless
and until you discover that the process was flawed.