I don't want to post many of these, but as I ended up making some charts of the evaluation of my work in Excel instead of in R (sorry R folks - you need to make more friendly graphing libraries), I needed to get them into latex somehow, and there is no export or save button in Excel. Google told me how to do it via scripting, but that is silly.
So: if you want to export or otherwise save your excel chart as a jpg, png, eps, or whatnot, all you must do is copy the picture (ctrl-c ctrl-v) into PowerPoint and right click on the image there to get the expected "Save as Picture" dialog. Hopefully Google will pick up on this tidbit for other aggrieved users out there.
On a more theoretical level, Mira Dontcheva wrote a paper on macro-learning awhile back and some UTA folks on better error reporting for black box components using a social scheme that would have been a useful way for MS developers to realize that this stupidity exists, assuming it was not intentional. Features being used, not even how, is valuable information for developers - internal Adobe teams seemed to be going ga-ga over availability of such data in certain contexts - and I think it does reveal another interesting question for achieving this in a lightweight manner: assuming you're working on a popular product, can you infer feature requests by mining google searches for how to do something? Sort of like the black box testing paper, but without the heavy instrumentation and sneaky, expensive tracking.
No comments:
Post a Comment