On Wed, Aug 23, 2017 at 7:07 PM, Poison BL. <poiso...@gmail.com> wrote:
> On Wed, Aug 23, 2017 at 6:29 PM, R0b0t1 <r03...@gmail.com> wrote: > >> As an example, I am interested in characterizing the power consumption >> of rendering a PDF document. I would hopefully only need to run the >> renderer once. > > > <snip>...</snip> That's potentially useful for profiling during > optimization of a very specific workflow in controlled conditions, but it's > really of questionable usefulness beyond that. > > -- > Joshua M. Murphy > Also, I meant to note, the same issues that apply to power profiling of rendering PDFs really also apply to nearly everything we tend to use desktop systems for on a regular basis. Our datasets change in complexity too much, many of the processes working with them are heavily impacted by user interactivity, and background tasks and other variable factors further complicate things so much that, aside from the extreme anomalies (like the pidgin issue Alan noted), it's rare that a coherent power usage can be accurately attributed to any one process. -- Joshua M. Murphy
