I don't mean to tread on any toes. I generally go by WP:BOLD. But the minimum width (the table scales to fit the page above that) is actually almost exactly the same in the horizontal direction (in my summary, I meant the width was reduced from my second edit), and while it is admittedly longer in the vertical direction, it shows both X-sections, and maintains symmetry with the balanced-unbalanced section both having a full-span ladder cell.
I understand about the colours being too loud, but I think it is is clearer if the title bar and the "balanced/unbalanced" cells are somehow emphasized slightly with respect to the cell holding the content, and if the title is slightly larger it helps to show it is actually a title. Of course, the colours can be changed, if that is the major sticking point - they were just the best I could come up with at the time!
Nice work on the filter articles by the way - that information is that kind of useful but hard-to-come-by stuff that's just the thing to get it though an undergrad's thick skull! Inductiveload (talk) 03:50, 25 January 2009 (UTC)[reply]
Yeah, ok, I'm being a bit WP:OWNerish about this, usually nobody takes much interest in this set of articles, but I really didn't like the style, the original was much more in keeping with the general Wikipedia page style. Also, you lost the rounded-corner frames which I quite liked. If we are going to do anything bold with this template, I think it should be radically shrunk in size, it already takes up too much vertical space for some of the articles it is in and your version just made that worse. The main reason for that (as I am sure you realise) is because you have unpacked the dual of the lattice filter (by the way, that should be described as mid-T-section derived, just as its dual is mid-Π-section derived). The reason it was originally squeezed into the end in that rather Heath Robinson way was just so the template did not get any longer. But you are right, it would be better to unpack it, but the whole template should then be shrunk to something more manageable, even at the expense of readability, the user can always click on an image of interest to expand.
Its great that you are going through some of my older diagrams which badly needed redoing (some were done in MSWord), especially the plots - your plots are MUCH better. On the circuit diagrams though, I have a few issues with the style you are using, I hope you don't mind me taking the opportunity to list them here;
I don't like the components being coloured in, this could be confusing. The originals had plain components on a coloured background chosen to match the Wikipedia page style. The coloured background serves just as well to highlight the diagram and make it interesting as colouring the components.
I really don't like your inductor symbol, it just doesn't seem to fit properly.
The symbols I am currently using are illustrated in this sample;
I know this is a bit retro, but so are the articles. Also, a lot of American readers are having problems with the European standard of using a box for resistors. It's very ironic to my mind that the new symbols were introduced to make life easier for draughtmen at the very moment in history when CAD and other design aids made it entirely unecessary.
I understand that your special interest in this - if I'd done so much work, I would also be a bit possessive of it! However, I am primarily a commons editor, and as such my goal is to make all diagrams as useful as possible for as many people as possible.
This generally involves not having a borders in the image - in fact, we have a whole category devoted to removing unneccessary ones, as this restricts what they can be used for (for example - an image with a big rounded border than looks out of place with all images in Wikipedia "frame" borders, or with images without, and it is impossible to give all diagrams the same kind of borders.) There is also an option to use wikimarkup to give an image a border (not sure about rounded though).
Personally speaking, I like colouring in the components, as this highlights the components in the diagram rather than the whole diagram. When you get to very complex diagrams, this is much easier on the eye that a mass of black lines on a constant background.
As for the background, I never put backgrounds on my diagrams, because it draws too much atttention to the image, and unweights the balance of the page - a diagram with black lines and maybe gentle colour sits much more happily IMO next to a piece of black text with blue links. If a person really wants a background on an image, there is Wikimarkup for that (right). Besides which, the image "frame" tag is designed to draw the attention of the reader as it is.
The inductor symbol is modelled after the modern symbol for a general inductor - it doesn't neccessarily mean air-core, like an open-loop symbol (possibly - this use, along with symbol, is deprecated AFAIK) does. Im my mind, there's no sense using an out-of date symbol when the new one will do. Most of these images are uploaded by me in the hope that one day I will use some of them on Wikibooks - and there's no sense trying to teach an old symbol set.
That said, I do use the old fashioned symbol for resistor (the zigzag) frequently as this draws a line between a "generic impedance" or "admittance", for which I use the European symbol, and a resistance, which gets the 'Mercan symbol. I don't stick religiously to the rule (I've though long and hard about whether to use the zigzag at all, but it seems standard practice, even in Europe, to do this). Interestingly, the box symbol was introduced in order to make CAD drawing easier - back in the day computers found it easier to draw boxes more clearly that zigzags, especially with additions like thermistor bars. This is the same reason that the "jump" over nonconnecting wires was/is being phased out of formal use - the computer could not draw the curved line easily or well. I've always found it harder to hand draw the draw the boxes in my real-life work than a zigzag.
While you say that the articles are retro, by that token so is integration, but it doesn't mean it needs to have period illustrations in 17th century cursive.
As for the table, how about one of those nifty javascript collapsible ones? That would solve all our problems!
What do you think of this for the table? I've toned the colors down (the last one was too dark and stormy, I agree), and made it collapsible, so it no longer takes up space. Also the "balanced" "unblanced" headers are now above their repective tables, which I think is clearer.
Textbooks and design drawings usually show the unbalanced implementations, but in telecoms it is often required to convert the design to the balanced implementation when used with balanced lines.
Yeah that's great, just what's needed. On the borders, I agree images should generally not be drawn with borders, but these ones were drawn specifically for this table so its a special case. I can live without it though, so not pressing that point. I think the colours are much better, but still a little clashy. Can you get all three levels of heading on the same hue but different levels of saturation?
On the symbols, even though I like the old ones, its pointless to fight the tide, and as I say, I am not getting into a lame argument over it. Even so, I think your symbols just do not look quite right. The inductor line weight is too heavy and the capacitors look too long compared to typical textbook representations. On the resistor zig-zag, that is still standard in the USA, or if it is not, there are a lot of American engineers who think that it is and get very confused when a box is used so I think you are right to continue using it. The usual problem on Wikipedia is USA-centrism but Euro-centrism can be just as exclusionist. SpinningSpark22:12, 25 January 2009 (UTC)[reply]
.svg)
.svg)
.svg)