Hi all, I see there was a discussion about the topic yesterday. I'll consider the ideas mentioned in the meeting minutes.
In the meantime, I was working on bringing the Writer grid to the same level of visibility as the grid in Draw / Impress. I've already changed the default grid colour in Writer to match the grid colour in Draw / Impress. I think the same reasoning applies to Writer, regarding the colour contrast. https://github.com/LibreOffice/core/commit/592df0a2e5e838567f1c8a1e0983e9ffd9886502 Now, I'm finalising a code change which introduces a 3x3 pixel cross for the main grid points in Writer (keeping 1 pixel dots for subdivision grid points), having the same visual appearance of the grid as in Draw / Impress. This code change also affects Calc, because Writer and Calc share the rendering code, and I did not see any reason to define different behaviour for Calc and to increase the code complexity while achieving that. I think this change is in line with what was discussed earlier under tdf#120897 bug ticket, and also consistent with some suggestions I read in the meeting minutes: https://www.mail-archive.com/[email protected]/msg10064.html This change improves the visibility significantly in Writer, improving the situation for users who have a problem with using the grid (see tdf#89544, tdf#135517), without making the grid too distracting. Best Regards, Tamás Eyal Rozenberg <[email protected]> ezt írta (időpont: 2025. nov. 4., K, 10:41): > Hello everyone, > > in this message: > > 1. A meta-note and a semi-apology. > 2. The drawing grid - aspects of how we might indicate it > 3. Partial history of how we've dealt with this > 4. Bottom line / tldr > > > > 1. A meta-note and a semi-apology. > --------------------------------------------------- > > I'll start with a word to Tamas. I am sorry for putting you in a > situation where you've started working on a change, thinking it was all > good and accepted, then someone suddenly tries to "put the breaks", > claiming otherwise. > > However - it's better that objections and reservations come up at this > stage, rather than at an even later stage, like a release. Of course > it's best if they come up even earlier, when the idea is first discussed > and before coding work has been invested... but that can't always be > guaranteed. On a personal level - if I had been aware of this issue > earlier, I would have said something. > > Anyway, the upside of these situations is that we get more eyes and > minds on an aspect of our UI (and UX) and this often leads to a better > eventual result, even if the road taken is longer. > > > 2. The drawing grid - aspects of how we might indicate it > --------------------------------------------------- > > Now, to the question at hand: The grid, in Draw and Impress. I would > list the following important questions or dilemmata about it: > > Q1: How visible/prominent should the grid be? > Q2: What graphical will make up the grid indications? Dots, Plus-signs, > diamonds, lines (dashed or otherwise)? > Q3: Should the grid indicators at grid line intersections be the same as > along grid lines? > Q4: Should the grid become more prominent as we zoom in - and how? > > > Before getting into these questions, it should be said that even though > bug 120897's title is about Q2, its (factually incorrect) opening > comment indicates it is also about Q3 and Q4. And such changes are > likely to have an effect on Q1 as well. > > > About Q1: > > While some want the grid to very very prominent and noticeable, so that > they know exactly where things are on the grid - others prefer it to be > subtle, faded, so that you notice it only with a little mental effort > and it doesn't otherwise distract you, and clash with the actual drawing > elements you put on the page. > > The "pro prominent grid" folks may say: "If the grid bothers you, you > can just hide it"; but the "pro subtle grid" folks would retort: "If you > need better visibility for a grid line, just place a few rulers". So, > both sides can mitigate the extent to which the opposite choice bothers > them. > > I personally would not like it if the grid were much more prominent. I > work with the grid on - and I like that its elements are at their > current level of subtlety (at least at zoom 100%). I believe this > perspective was not taken into consideration so far, with some assuming > that everyone wanted a more prominent grid. > > About Q2: > > I won't survey the space of all possibilities here, but: > > * We currently (26.2) have dots on gridlines, and +'es (crosses/pluses) > at intersections of gridlines. > * Bug 120897 proposes either all-crosses, including at > non-intersections, or making the grid more dense so that that there is > no need for indicators at non-intersections (see the screenshots at bug > 120897). > * Powerpoint has diamonds at zoom 300%, and diamonds/dots at zoom 100%. > * Sketch has full grid lines, not periodic indicators. > * WPS copies MS Office > > Personally, I think that using +'s at non-intersections is jarring; and > I would not appreciate making the gridlines denser at 100%. > > About Q3: > > There are two-and-a-half options here: > > Different indicators at intersections: > > +---+---+ > | | | > +---+---+ > | | | > +---+---+ > > Same indicators always: > > +++++++++ > + + + > +++++++++ > + + + > +++++++++ > > and - no indicators except at intersections: > > + + + + + + + + > + + + + + > + + + or + + + + + > + + + + + > + + + + + + + + > > > when choosing the the third option, we either make the grid lines denser > (to catch what were previously non-intersections, as intersections), or > significantly reduce the number of indicators. > > And whichever of the options we choose, this affects > prominence/visibility of the grid. It should also be said that the > specific choice of indicator effects whether or not it makes sense to > differentiate intersections from non-intersections: Pluses literally > tell us that two lines intersect; and dashes as indicators tell us that > we're looking at a single line; but if we go with dots or diamonds, for > example, it's a different story. In that case, a differentiation could > be slightly-bigger-diamond vs slightly-smaller diamond. I believe MSO, > which uses diamonds, does not differentiate intersections. > > > About Q4: > > Currently, we just draw the same grid indicators, for our grid lines, at > each zoom level. We could, in principle do one of two things (at least): > > * We could make the indicators larger at higher zoom levels; and this > enlargement could either be just like drawing objects, or sub-linear, so > that the indicators at still small, just a bit larger to indicate we are > zoomed in. > > * We could increase the density of indicators along existing grid lines > > * We could draw indicators at more lines, perhaps less-prominently than > the main indicators, at half-grid-resolution, or at grid subdivisions - > rather than at full grid resolution. > > Note that the first option, once you zoom in enough, means that we still > might not see grid indicators whatsoever, or very few of them. > > > I personally think we should add indicators along lines - and see if > that's good enough for us. If people then also think we should take > additional measures at higher zoom levels - I wouldn't oppose, but I > would also not ask for that myself. > > > 3. Partial history of how we've dealt with this > --------------------------------------------------- > > The matter of grid prominence, and comparison with other apps, came up 7 > years ago, in this bug: > https://bugs.documentfoundation.org/show_bug.cgi?id=117348 > > the bug poster concluded that a necessary (and sufficient?) measure to > take would be to make the grid indicators darker. > > The matter was discussed in one of our design meetings: > > https://wiki.documentfoundation.org/Design/Meetings/2018-10-24 > > the results then were: > > * Accept the dark-grid-indicators proposal > * File bugs opened regarding a bunch of other possible changes: Changing > the grid color to blue; have dot indicators "depending on screen > resolution"; and using "zoomable pluses" (so, a combination of switching > to + indicators and increasing grid indicator sizes with zoom). > > It was not decided that making the grid gray was insufficient; the > additional bugs were about additional options to be considered. When the > bug regarding "zoomable pluses" was filed, the opening comment states > that "This idea is supported by the design team." - a much stronger > statement than the design meeting minutes. > > In that discussion, nobody seems to mention the interest of _avoiding_ > and overly-prominent / distracting grid. Perhaps this was because, at > that time, the grid was even less prominent than it is today - with > lighter gray color and not many indicators - so it was not on people's > minds. > > > 4. Bottom line / tldr > --------------------------------------------------- > > I think that: > > * The discussion 7 years ago was sufficient for the bug they had at hand > - but not sufficient for making more fundamental changes to the grid, > like "zoomable pluses". Particularly, it did not take the countervailing > interest/need in keeping the grid subtle, especially at 100% zoom, into > account. > > * The 'zoomable pluses' proposal, as reflected in Tamas' "POC 2" > screenshots - are an undesirable change, in my opinion, especially in > 100% zoom. See: > https://bug-attachments.documentfoundation.org/attachment.cgi?id=203694 > The grid becomes too prominent, too distracting, and I find the pluses > at non-intersections aesthetically displeasing. > > * I suspect we may be able to find consensus if we opt for an increased > prominence at higher zoom levels and same-or-similar prominence at 100% > zoom level. Or we may not - I don't really know how everyone on this > list feels about the grid. > > * If we don't reach consensus - we should probably give users some level > of choice between a more-prominent or less-prominent grid indication. We > already offer many grid options (under Tools > Options > LO Draw > > Grid), but those are about the grid itself and its effect on behavior, > not about how it's drawn. At the very least we could add a toggle or > two-value radio group, "subtle" vs "prominent" or whatever we'll label > it; or we could allow control of other aspects of grid drawing, like > shapes, like density of grid lines drawn (relative to the grid > subdivision), line/strike width etc. > -- To unsubscribe e-mail to: [email protected] Problems? https://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/ Posting guidelines + more: https://wiki.documentfoundation.org/Netiquette List archive: https://listarchives.libreoffice.org/global/design/ Privacy Policy: https://www.documentfoundation.org/privacy
