(In reply to Sahil Gautam from comment #0)
> 6. flat, sharp cornered icons.

It hits me as if you are describing not "why it looks bad", but "why it looks different to Linux".

FYI, the current icon set is created specifically to follow the modern Windows interface. Compare to "flat, sharp cornered icons" of e.g. that same Excel. Please don't use your personal preference as the "how the modern UI should look" measure. Thank you.

(In reply to Mike Kaganski from comment #1)
> (In reply to Sahil Gautam from comment #0)
> > 6. flat, sharp cornered icons.
> 
> It hits me as if you are describing not "why it looks bad", but "why it
> looks different to Linux".

i didn't consider linux at all, the platform is windows 11 and if you open any application, settings, file manager, store, you would find fluent icons with rounded corners. i will attach two icons below, there's a day and night difference between them, the bottom right gives off "modern windows 11 vibes" and the top left one gives off old windows 10 vibes.

> FYI, the current icon set is created specifically to follow the modern
> Windows interface.  Compare to "flat, sharp cornered icons" of e.g. that same
> Excel. 

i went over all the icon themes, and none of them felt like a match for windows 11 icons. i might be comparing apples to oranges here (system icons with application icons) but we can't say the comparison is baseless without actually testing things.

Created attachment 201498 [details]
wifi icons screenshot from icons8.com

flat wifi icon (top left) vs fluent wifi icon (bottom right)

(In reply to Sahil Gautam from comment #2)
> i might be comparing apples to oranges here (system icons
> with application icons)

Yes you do :)

> but we can't say the comparison is baseless without actually testing things.

That's why we need to compare Calc to, say, Excel. And you are going in a wrong direction. Please focus on things that are less subjective. You created a rather long list; don't put there just anything you can think about: such lists should focus on things that won't create this king of objections from let go.

(In reply to Mike Kaganski from comment #4)
> (In reply to Sahil Gautam from comment #2)
> That's why we need to compare Calc to, say, Excel. And you are going in a
> wrong direction. Please focus on things that are less subjective. You
> created a rather long list; don't put there just anything you can think
> about: such lists should focus on things that won't create this king of
> objections from let go.

i quickly replaced some colibre icons with some fluent icons and took some screenshots, pasting them here to get done with this apples to oranges comparison here :). won't talk about it after this maybe after the important things are done.

Created attachment 201499 [details]
single toolbar

single toolbar with some fluent icons

Created attachment 201500 [details]
regular toolbar

regular toolbar with some fluent icons

Created attachment 201501 [details]
tabbed ui

tabbed ui with some icons

Created attachment 201502 [details]
groupbar

groupbar with some fluent icons

although just changing icons isn't enough, Mike it seems like the apple has to become an orange :), the label font also needs to match the windows system font if this apple-orange comparison is to make sense.

closing my case here, thinking of the font change, the difference became more apparent... that's why excel doesn't use these fluent design icons.

@Sahil, *

Does this really make sense to try to replicate WinUI3/Fluent appearance [1] when the underlaying framework remains VCL based?

Can't avoid it, so maybe better to select a different UI Widget library since win32 is deficient, and refactoring native win32 for WinUI3 and XAML is going to be too expensive for LibreOffice (to refactor, thought maintaining it might be less troublesome if well done). 

So just move the UI fully onto that non-MS motif? That seems more in line with a cross platform project.

E.g. why not Qt Widget [2], and reuse some of the kf6 implementation on win. Is it really an issue that the icon theming does not match Metro/MDL or whatever WinUI3/Fluent suggests [1]? 

I don't think so and folks really don't expect visual mimicry on non-native cross-platform implementations. Just solid functional control implementation and reasonable UI cohesiveness--not fully implemented os/DE UI integration.

=-ref-=
[1] https://en.wikipedia.org/wiki/Fluent_Design_System
[2] https://doc.qt.io/qt-6/gallery.html

(In reply to V Stuart Foote from comment #11)
> E.g. why not Qt Widget [2], and reuse some of the kf6 implementation on win.

As I've mentioned in the past, that would be the first thing I'd suggest to try as well, since Qt aims to integrate well with the platform and this would allow to reuse the existing qt6 VCL plugin without having to rewrite + maintain everything for yet another widget toolkit, and would also eventually give us UIA support on Windows (though it will require changes in assistive technology as well if they currently rely on IAccessible2). Qt provides OS-specific default styles, including one for Windows 11.
(There's even an older WIP branch by jmux, `feature/qt5-win+mac` experimenting with using the Qt-based qt5 VCL plugin on Windows and macOS as well, but further work would surely be needed to make this "production ready".)

@Sahil: Your tdf#164656 comment 8 says
> I would be lying if I say "I don't want to go Qt", but people say Qt is bad, really bad :).

, but you haven't yet answered who says that (and why), so I think it's hard to take that as a serious "last word" disqualifying Qt from the start.
(Of course, there may be reasons to not use Qt, but knowing which ones those are in practice would in my opinion be helpful in that case.)

What icons to use is likely a separate question at least for the ones that LO provides. Depending on where this discussion is going, I think it might make sense to split single aspects into separate tickets.

(In reply to Michael Weghorn from comment #12)
> As I've mentioned in the past, that would be the first thing I'd suggest to
> try as well, since Qt aims to integrate well with the platform and this
> would allow to reuse the existing qt6 VCL plugin without having to rewrite +
> maintain everything for yet another widget toolkit, and would also
> eventually give us UIA support on Windows (though it will require changes in
> assistive technology as well if they currently rely on IAccessible2).

FWIW, GTK 4 also supports Windows, but I don't know how well it integrates with the theming,... there.

(In reply to Sahil Gautam from comment #0)
> the problem is that windows doesn't have  a "modern" and stable widget
> toolkit, winui3 has performance issues and is xaml based. i don't think an
> implementation standing on workarounds for workarounds would last (i am
> willing to give win32 another try will see how far it goes this time) but
> even if that succeeds, libreoffice won't look "modern" because that
> "modernness" belongs to winui3 not win32.
> 
> assuming that winui3 won't be abandoned by microsoft, it would take me
> atleast 6 months to get something meaningful if possible at all. i already
> spent ~2.5 months trying to make win32 work, but since it's not object
> oriented like gtk and qt, it didn't work and from the experience i can say
> that it won't work without clever tricks.

Adding this here as well, WinUI3 has apparently been disappointing for developers: https://github.com/microsoft/microsoft-ui-xaml/discussions/9417

Quotes:

"Our stack is mainly WPF with a few newer WinUI apps. Our clients have expressed concern at several long standing issues so we are now looking at moving backwards (if you can believe that). [...] It feels like this project is completely abandoned now, but I hope I am wrong."

"WinUI is definitely not the native stack for Windows. Of widely used programs, only Watsapp, I think, uses WinUI (or some of its predecessors). The native stack for Windows is Winforms and (to a lesser degree), WPF."

WPF is C#.

Did some more digging: https://old.reddit.com/r/dotnet/comments/1kaj91k/why_are_there_not_more_winui3_applications/mpn4np3/

"I have written a relatively complex application in WinUI3. There are a number of issues. First off, lack of pre built components. Major major issue. Second, the MVVM pattern is difficult for lots of newer devs to grasp. Third, MS has effectively abandoned it, moving to a web stack for their in house apps."

:) i was trying to avoid the "native widgets" question in this thread, not because i don't want to answer it, but because it's not worth answering at the moment.

if you ask me what would improve libreoffice's appearance on windows, native widgets or these 14 action items, i would bet on the later first. the users don't care about the "nativeness", they say "it looks old/ it looks different" and we need to fix that first.

windows doesn't have a good widget toolkit at the moment. once all these 14 issues are fixed, we can talk about which toolkit to choose once again, on https://bugs.documentfoundation.org/show_bug.cgi?id=164656 not here.

i don't think the toolkit situation would get any better on windows even after an year or so, then using qt seems like the next best option as it supports fluent ui widgets.

but let's focus on these action items for now.

(In reply to Sahil Gautam from comment #15)
> :) i was trying to avoid the "native widgets" question in this thread, not
> because i don't want to answer it, but because it's not worth answering at
> the moment.
> 
> if you ask me what would improve libreoffice's appearance on windows, native
> widgets or these 14 action items, i would bet on the later first. the users
> don't care about the "nativeness", they say "it looks old/ it looks
> different" and we need to fix that first.
> 
> [...]
> 
> but let's focus on these action items for now.

Fine for me.

Unless "change XYZ (e.g. switching to an existing toolkit that takes care of it) will fix multiple of them at once", this seemed a bit like a "random" collection of different aspects not necessarily directly related to each other (so maybe useful to keep track of separately and have this here as a meta ticket), but if you plan to take care of all of that yourself, I'm fine with it if that's the way that helps you (and the UX team) organize/track that.

Is this supposed to be a bug, or a meta-bug? It seems somewhere in-between.

Anyway, as a sort-of-a-meta bug, I'm marking two dependencies which basically encompass items (7.) and (10.): 

* Bug 166350 - Support framed groups with corner expansion button, on Tabbed notebookbar tabs
* Bug 166641 - Support group-title bottom label in tabbed UI

Screenshots:

Sahil you've added some screenshots, but their names don't clarify how they relate to the bug.

Anyway, I would like to ask that you post some screenshots which illustrate some/most/all problems on your list, hopefully in comparison to other Windows GUI apps (not necessarily MSO).

For now, I'll ask some questions.

> 1. no proper toolbar background color, or separators for distinction between the toolbar region and the document region.

I see a different BG color and a separator line, of sorts, both for standard and tabbed UI on Win11 with a recent LO 26.2 nightly (and Light theme). Are we talking about feedback for earlier versions of LO? Or is there something I'm missing?

> 2. crowdedness in the tabbed ui

Please explain. Too many buttons? Not enough large buttons? Not enough buttons with label? Not enough space between buttons? Something else?

> 3. boxy old school tabs

Is this different from the "sharp corners" point?

> 4. calc fitting way more cells at 100% compared to excel. comparison with
> excel is even more important here as that's what the users use as the
> baseline while comparing.

Do you mean the default zoom setting, or defaults for column widths? Row heights?

> 5. sharp corners in buttons, menus, menu highlights and various widgets like
> dropdowns..
> 6. flat, sharp cornered icons.

You mean the icons, not the buttons with icons on them, right? Ok, so can you give an example of the "flatness" of an icon, and of icons with "sharp corners"?

> 9. lack of smooth cell focus rectangle animations in calc

You mean, an animation of moving the focus from one cell to another?

> 12. sharp corners on treeview boxes, searchboxes and edit boxes
> 13. missing preview expander widget for the tabbed ui.

What is a "preview expander widget"? And where exactly in the tabbed UI is it missing?

> 14. missing animation controls/previews from the toolbar, they are on the
> sidebar but most of the users migrating from mso are not familiar with it,
> we need those on the notebookbar as well.

Do you believe a more prominent position (and size) of the tabbed bar button which expands the animation sidebar, would help, if only as a stop-gap measure?

I want to point to high-level perspective.

The actual problem that this is intended to resolve looks greatly exaggerated to me. It seems like a situation, when someone deeply interested in a specific area (UI) starts to see disproportionate amount of complaints around this area (not because they are actually that frequent, but because the person pays attention to each of them).

As someone who regularly reads our Ask site, but myself is not very involved into UI, I can tell my perception, that there was a really influential change, that was done some time ago, which made a huge impact of *real frequency* of "your UI is outdated" complaints. It was the introduction of Colibre icon set. Nothing else that I can recall made such a huge difference before vs. after; and I can say, that current flow of "outdated" complaints is really not high, very normal for a "you can't please everyone" situation.

(In reply to Eyal Rozenberg from comment #18)
> Screenshots:
> 
> Sahil you've added some screenshots, but their names don't clarify how they
> relate to the bug.

these screenshots show various user interfaces with fluent design icons (i replaced some of the colibre icons with some fluent design icons). you can find the fluent ones on the left side.

> Anyway, I would like to ask that you post some screenshots which illustrate
> some/most/all problems on your list, hopefully in comparison to other
> Windows GUI apps (not necessarily MSO).

will do soon.

> 
> > 1. no proper toolbar background color, or separators for distinction between the toolbar region and the document region.
> 
> I see a different BG color and a separator line, of sorts, both for standard
> and tabbed UI on Win11 with a recent LO 26.2 nightly (and Light theme). Are
> we talking about feedback for earlier versions of LO? Or is there something
> I'm missing?

let me check that once again, i checked it in dark mode, maybe on 25.2..

> > 2. crowdedness in the tabbed ui
> 
> Please explain. Too many buttons? Not enough large buttons? Not enough
> buttons with label? Not enough space between buttons? Something else?

no, it's mostly about lack of layout planning. the ribbon follows a proper scheme throughout, large buttons with label at the bottom and taking, or columns  of small buttons, or columns of checkboxes.

> > 3. boxy old school tabs
> 
> Is this different from the "sharp corners" point?

yes, it's a separate action item as tabs are drawn differently than buttons, buttons are just rectangles.
 
> > 4. calc fitting way more cells at 100% compared to excel. comparison with
> > excel is even more important here as that's what the users use as the
> > baseline while comparing.
> 
> Do you mean the default zoom setting, or defaults for column widths? Row
> heights?

yes at 100% zoom and same window size we fit 28 rows whereas excel fits 21 rows, and the cell width in excel is also less than ours.


> > 5. sharp corners in buttons, menus, menu highlights and various widgets like
> > dropdowns..
> > 6. flat, sharp cornered icons.
> 
> You mean the icons, not the buttons with icons on them, right? Ok, so can
> you give an example of the "flatness" of an icon, and of icons with "sharp
> corners"?

all the screenshots shared till this comment demonstrate just that, there's a wifi icon example. windows 11's icons are way different from windows 10's, they have a "3D-ish" look.

> > 9. lack of smooth cell focus rectangle animations in calc
> 
> You mean, an animation of moving the focus from one cell to another?

yes


> > 12. sharp corners on treeview boxes, searchboxes and edit boxes
> > 13. missing preview expander widget for the tabbed ui.
> 
> What is a "preview expander widget"? And where exactly in the tabbed UI is
> it missing?

if you open writer with tabbed ui, you would see the styles preview on the home tab. note that mso also has similar widgets all over and they have a dropdown button to the right of this preview which on click creates a dropdown with a lot   more options.
 
> > 14. missing animation controls/previews from the toolbar, they are on the
> > sidebar but most of the users migrating from mso are not familiar with it,
> > we need those on the notebookbar as well.
> 
> Do you believe a more prominent position (and size) of the tabbed bar button
> which expands the animation sidebar, would help, if only as a stop-gap
> measure?

expands the animation sidebar? i am talking about a new tab called "Slide Show" or "Animations", which provides buttons and controls and dropdowns with icons (which the sidebar currently has) like the "Animations" tab in Power Point https://support.microsoft.com/images/en-us/b92cdac9-05aa-424a-a448-b0e41f4f9479.

(In reply to Sahil Gautam from comment #20)
> these screenshots show various user interfaces with fluent design icons (i
> replaced some of the colibre icons with some fluent design icons). you can
> find the fluent ones on the left side.

... which is why I suggest you rename them. Never mind.

> no, it's mostly about lack of layout planning. the ribbon follows a proper
> scheme throughout, large buttons with label at the bottom and taking, or
> columns  of small buttons, or columns of checkboxes.

The OnlyO ribbon is also reasonably well arranged.

Perhaps we should try to set up some brainstorming session for re-partitioning / sectioning the commands on the tabs in the tabbed UI, to see if we can't get a better result?

> > > 3. boxy old school tabs
> yes, it's a separate action item as tabs are drawn differently than buttons,
> buttons are just rectangles.

but the complaint is the same?

> yes at 100% zoom and same window size we fit 28 rows whereas excel fits 21
> rows, and the cell width in excel is also less than ours.

So, we have cells which are vertically-shorter but horizontally-longer.

TBH, I somewhat doubt this is much of an issue given that one can use zoom or resize; but there's good reason to change the default, I wouldn't be against it.


> all the screenshots shared till this comment demonstrate just that, there's
> a wifi icon example. windows 11's icons are way different from windows 10's,
> they have a "3D-ish" look.

I'm looking at the WiFi icons screenshot and I'm not really following.

> > > 12. sharp corners on treeview boxes, searchboxes and edit boxes
> > > 13. missing preview expander widget for the tabbed ui.
> > 
> > What is a "preview expander widget"? And where exactly in the tabbed UI is
> > it missing?
> 
> if you open writer with tabbed ui, you would see the styles preview on the
> home tab. note that mso also has similar widgets all over and they have a
> dropdown button to the right of this preview which on click creates a
> dropdown with a lot   more options.

Well, a "preview expander" sounds like bug 145038; but that would still be a button, while

> i am talking about a new tab called "Slide
> Show" or "Animations", which provides buttons and controls and dropdowns
> with icons (which the sidebar currently has) like the "Animations" tab in
> Power Point
> https://support.microsoft.com/images/en-us/b92cdac9-05aa-424a-a448-
> b0e41f4f9479.

... is another idea, also relevant.

Bottom line: I suggest you split this bug up into the constituent specific issues. Some of them might benefit from VCL changes/improvements/toolkit adoptions, other not as much.

(In reply to Mike Kaganski from comment #19)

It's possible that you're correct, but in my not-so-humble experience, it is difficult to assess "user annoyance" at things. People who specialize and have a deep interest in something may bring to the surface something the for many people is an annoyance, but not one they would give much thought to or take the time to complain about.

Plus - a combination of a large number of small issues may result in an overall negative impression. If we (well, you guys reall) were to resolve most of this individually-minor issues, it's possible that less people would be complaining about how our UI is not "modern" enough. (Or they won't, I can't really say.)

(In reply to Eyal Rozenberg from comment #22)

Please take a look at my comment 19 in the light of comment 1. What you say about upsides is true; but the very first list had things that were IMO not "problems", but a matter of personal taste. And item 6 attacked **the** most successful change in this area. So I have a vary scary impression, that it may be based on actual problem much less than it seems (and has a potential of making things worse).

I do not mean to say "do not work on this". I only try to say "please try to be less subjective". See: what you wrote about "an annoyance, but not one they would give much thought to or take the time to complain about" is totally irrelevant for us - given our size, every small annoyance will get its report; and we can really evaluate the impact, looking at the available feedback. And *that* gave me a chance to *see* the difference. When the mentioned icon set was in works, I was really skeptical - I didn't think that yet another icon set could change anything; I thought that we need huge changes to improve the perception a little. I was proven wrong (and I realize that it means, by the way, that one or several of the items discussed here may have similar tectonic effect). Just don't break what was already improved, just because of personal opinion.