Posts by Kursiv

    Rolling average COP with a suitably large time-constant is obviously a much better method of presentation

    I tend to disagree. Me thinks the only dependable key indicator would be total energy out / total energy in from the beginning of the testrun.

    You can then still discuss e.g. how much energy was used to heat up the reactor itself or "sneaked out" somewhere, but any moving average COP is subject to distortion due to all kind of artefacts during the test run.

    Edit: Henry just came up with the same argument just while I was typing.

    This is their scanner technology currently not applicable, data was collected on a computer not using a scanner and probably the graphs created in excel.

    I agree, but no one knows how the graphics was transferred from (presumably) Excel into the presentation and which compression algorithms the programs involved were using. When I do these things quite often I use a screencapture program (e.g. Snagit or Greenshot), the Windows cut-n-paste function, and display program like Ifranview, which compresses the image. On top the presentation went to Powerpoint, presumably saved, and later converted to pdf, the uploaded to Scribd(?). Nowhrere am sure about how the programs are handling data and compressions.

    Quite offen image processing programs are using compression algorithms availble in source code somewhere in the net, so the Xerox bug msy pop up somewhere else.

    On a side note, when I would want to "enhance" my data in the graph I would modify the source data in the excel table, not the graph itself. Simply because it's more easy and more straight forward to do,

    A bit off-topic, but this can possibly be an artefact form image processing/compression. See the attached image how Xerox Workcenters are altering scanned images. Column 1 is the original, the other columns show the scanned results on various WC models

    The bug is coming from the image compression algorithm which Xerox uses. It tries to find similiar aereas in the image, then storing the aerea only once in the compressed file.

    The bug hit the fan in 2013, when they found that scanned invoices display numbers different from the original paper, and some libraries which changed to archiving material using Xerox machines found that their archives are possibly corrupted.

    An engilsh writeup from David Kriesel who found the bug you can find on his blog:…ten_numbers_when_scanning

    For those who understand German you can find a talk from David given on 31C3 about the bug. Hilarious story, quite entertaining to watch: