File talk:20201210 Antarctica ice mass variation - NASA GRACE-en.svg

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Inkscape and revisions

[edit]

User:RCraig09. I'm quite new to Inkscape, so I hope this is all fine. I just wanted to replace the hyphens with minus signs, but decided to modernise the graph too (which involves taking away graph elements that don't convey information, so that data is emphasised). If you prefer for the future that I just jot down a list of requests, that fine too. Femkemilene (talk) 11:39, 22 January 2022 (UTC)[reply]

@Femkemilene: I started to learn Inkscape with a ~23-lesson series on YouTube: https://www.youtube.com/watch?v=8f011wdiW7g&list=PLqazFFzUAPc5lOQwDoZ4Dw2YSXtO7lWNv (YouTube account: TJ FREE)
Inkscape makes complex curves much easier, but it leaves a HUGE amount of wasted specifications in the file (your version is 4x the byte size as my file, and defines some structures in a very roundabout way, etc). That's why I use my spreadsheets to create lean SVG that's easy for humans to edit if needed.
I think your changes are OK but I want to explain:
— At one time, I had (and I think it was Efbrazil who had) trouble with "transparent" backgrounds appearing *black* in the PNG view, which made axis labels impossible to see. Since I'm not seeing this problem now, I won't change anything here, but I strongly prefer opaque backgrounds. (If these graphics were for printing on t-shirts, I might think differently!)
— In particular, I prefer white plot areas (to emphasize the data) with a light neutral surround. The light neutral (#f8f8f8) surround functions like a picture frame, which helps to focus the view's attention. (Using #f8f8f8 rather than a pleasant colored surround was my compromise a year or two ago.)
— Our audience is non-techy, so I think it's important to give images a variety so they are not so "sterile". For variety of charts commonly found, see: https://www.google.com/search?q=%22climate+change%22+(chart+or+graph)&source=lnms&tbm=isch&sa=X&ved=2ahUKEwjcl7Gnpcb1AhXgkYkEHVlBAIYQ_AUoAXoECAEQAw&biw=1345&bih=757&dpr=1
— I chose to label every five years, not every two years, to reduce visual clutter.
— Incidentally, I have a few small changes to make to my spreadsheets; nothing big.
It might be easier if you were to make requests on Talk Pages here. RCraig09 (talk) 21:32, 22 January 2022 (UTC)[reply]
Will ask on the talk page in the future, may have been carried away a bit here.
  • The transparent background was a mistake. I didn't think it mattered too much, so I didn't spend too much time trying to figure out how that worked
  • We disagree. My thinking is partially based on this paper (tip number 8)
  • I agree the figure is sterile. It's just difficult to make it interesting as it's just a line.. Confidence interval shading is always a nice filler, but are tiny here. Also, we could mimic NASA, and make the title and explanation colourful (https://climate.nasa.gov/vital-signs/ice-sheets/)
  • The reason I chose every four year was to allow a minor vertical raster line in between, so that we'd get squares rather than more obtrusive rectangles. Femkemilene (talk) 21:47, 22 January 2022 (UTC)[reply]

9 Feb 2022 edit (thicker lines, text-to-path)

[edit]

@Amitchell125: The climate change community at en.wp has been working to make CC-related diagrams more standardized. The above section on this Talk Page is part of that process.

  1. Can you explain, or maybe reconsider, your 9 Feb changes to line thicknesses?
  2. Also, when text is changed to paths, the text can't readily be adapted to other languages using the <switch> designation (see File talk:20200324 Global average temperature - NASA-GISS HadCrut NOAA Japan BerkeleyE.svg#How to preserve the translations for how this can be accomplished with a simple text editor or translation tool. (The "path" file is also 12 times larger than the "text" file.)

Please consider. Thanks. RCraig09 (talk) 16:47, 10 February 2022 (UTC)[reply]

@RCraig09: , hello - and thanks for all the work you do on diagrams. I'm not concerned about file size or translation possibilities more than I am concerned that the text quality is correct, and the lines are actually visible, which they weren't before. If it helps, I'll make my own version of files with these issues, and link them to the original, to avoid creating the problems you mention. Amitchell125 (talk) 18:57, 10 February 2022 (UTC)[reply]
@Amitchell125: I think it's a waste to duplicate the same essential chart, and, undesirably, editors would have to choose which of two competing charts to use. Especially in a chart about Antarctica, an international (non-English speaking) audience is important, and translation is practical only with native text and not paths. The text quality, being in SVG, is always correct, and is not crowded in this diagram. That leaves line thickness: maybe a compromise can be made. @Femkemilene: sorry to bother you, but do you have thoughts? RCraig09 (talk) 20:11, 10 February 2022 (UTC)[reply]
Funny that we have such discussion for the simplest of line graphs. I think we should get closer to the original layout: no minor raster lines, bit of colour in the title, making sure that the raster line in the year 2000 is either absent or consistently bolded.
I've never understood those problems with text quality, but I hope that SVG rendering will improve as part of the meta:Community Wishlist Survey 2022/Tracking (SVG improvement now on 5th/6th place).
I don't think file size should be taken into account. These SVG files are significantly smaller than jpg files anyway. Femke (talk) 20:19, 10 February 2022 (UTC)[reply]
The problem with Wikimedia's SVG font renderings is that thumbnail views can vary in size for different values of upright=_._ (see File:20200522 SVG font rendering at different thumbnail sizes.gif). But that's not a problem in this particular graphic. SVG text itself is correct because it's scalable vector graphics. I'll wait for comments, but in the next few days I'll update the prior version to arrive at something everyone accepts. RCraig09 (talk) 21:00, 10 February 2022 (UTC)[reply]

I've uploaded Version 8. To compromise between User:Amitchell125's bolder gridlines and User:Femkemilene's request for no minor raster/gridlines, I have made the minor vertical gridlines purposely be invisible in thumbnail view, but visible enough in full-screen view when enlarged to see ~repetitive annual patterns. Major vertical gridlines are stronger. RCraig09 (talk) 18:32, 11 February 2022 (UTC)[reply]