about using winboard

[Ovyron]

And btw, the open source nature of the project, as someone already mentioned, means you can address whatever you want.

No you can't. Unless you're a programmer. 99% of users aren't programmers, it's no wonder they'd choose some other GUI to do their chess related stuff, even if that GUI does cost money.

I have a couple custom changes in the version I run. Nothing major, I just don't like how fat the target square circles are, so I shrunk them. Took 5 minutes.

My point is that, because of the nature of the project, Winboard could be the Stockfish of chess GUIs, and everyone would stop using Chessbase/Fritz, Shredder Classic, Chess Openings Wizard/Bookup, Aquarium and the other Commercial GUIs, because Winboard fulfills all their wishes, this is not like using Houdini and Komodo because they have interesting ideas of their own.

Instead, Winboard is like the Crafty of Chess GUIs, and HGM claiming it's the best is like claiming Crafty is the best chess engine, with the only difference that we have plenty of chess engines and not and many chess GUIs, but basically you can find anything Winboard does, and find it in some other GUI, done better.

[hgm]

While I'm thinking if I ever tried it before, let me try to click on this strange thing "File": omg!! A strange thing that wasn't there before appeared! How can I cope with that? They must be insane! How can they expect me to click on something that wasn't visible before!

Yes, remarkable, isn't it? That for pull-down menus it doesn't seem a problem, but that when you do it with a board with pieces it is disastrous. I have no real explanation for that. (Two dimensional?) I would not have believed it if I wouldn't have experienced myself. You obviously haven't.

Anyway let me see if I understood your various statements correctly:
- ChessBase design is "pretty awful" (and they placed 12 piece-types on the right side of the board).

Correct. I guess my main gripe is that click-click moving does not work, while this is how I usually move pieces around.

- "having an external palette is truly problematic when 72 (colored) piece types would have to be displayed next to the board."

Correct. The average distance you would have to move to select a new one becomes rather large, and it becomes difficult to find the one you are looking for if they are just arbitrarily crammed into a square table (like 12x6 or 8x9)

- "Displaying two board side by side, one empty, the other in the initial position, could be a possible solution."

Indeed, by organizing the exteral palette like the initial position, people would know where to look for the piece they need, even if there are a lot. They would not have to examine all the images until they see the right one, but would know its locationn. Basically the difference between a O(N) linear search and O(1) hashing.

- Displaying the piece-types on request is "absolutely the most awful experience I ever head. I still get the shivers" because "having to go click something that just appeared is at least an order of magnitude more inconvenient than clicking something that has already been visible before you decided you wanted to click it"

Correct. If we would rate WinBoard's currently implemented method as 8, ChessBase as 5, I would say this one was -2. Surprising, but true.

- "Perhaps it would be useful to have a partial board clear, which does not touch a 4x4 area in the center of the board."

That specifically applies to a design using an internal palette. An alternative would be to have a 'Select all', and then a way to deselect a few pieces (e.g. by Shift + click) followed by 'Delete', to just keep a selected few. This is the more conventional method (e.g. for deleting files in a folder).

- "The thing I am not sure of is how to best treat an attempt to drop on an occupied square". The intention of the user when dragging a piece over is to achieve an "automatic termination of the multi-drop series, and select the clicked piece for moving"

I don't recall having exactly said that about "intention". The issue is whether moving a piece should then automatically select that type for dropping as well. After some experimenting things appear to work most smoothly if you don't, (just stick to the type you were dropping before, even when things get moved), and just leave the drop-type selection for a (destructive) right-click, i.e. a 'delete piece' rather than a 'move piece' action. Provided that you only allow drops on empty squares.

So in short, I get the most pleasant user experience from
* right-click on piece = delete it and select that type for dropping
* left-click on empty (when nothing selected for moving) = drop the selected type
* left-click on piece = select that piece for moving
* left-click when piece is selected for moving = move piece there (and deselect)
The latter two are "business as usual"; moving of pieces always works that way. The first two would be additions specific to Edit Position. The first one is the on-board alternative for selecting by a click on an off-board icon, and the second is how you then populate the board with it. In other modes clicking empty squares as 'from-click' is ignored. (Nothing there to grab.) Both empty and occupied squares are valid as 'to-clicks' for a move, depending on whether it is a capture or not. So the idea is that moving has priority over dropping, working exactly as usual, and that actions that would be ignored for ordinary moving would now be used for dropping.

[Evert]

Instead, Winboard is like the Crafty of Chess GUIs, and HGM claiming it's the best is like claiming Crafty is the best chess engine, with the only difference that we have plenty of chess engines and not and many chess GUIs, but basically you can find anything Winboard does, and find it in some other GUI, done better.

Chess variants.
There are a few others that do a few, but none that have the extend of variant support in XBoard.

[Evert]

You know what I would find useful?
The ability to select a group of pieces and copy/move those as a group. Grab part of a pawn structure and move it over to another file or rank. Repeat a particular formation (perhaps mirrored or colour-flipped).

Sure, it's a niche application, and you can get by ok by just moving one piece at a time, but this would be awesome for tinkering with problems. Even more so if you could do live analysis of the position you're currently editing, but that might be a bit much to ask for since the position might be illegal.

[Ras]

Well, I am not so sure. It is just a 'collision' between two behaviors, the universal one that clicking something is selecting it for moving

Ah I see, yes, for moving a piece that is already there, yes.

On a more general note concerning GUI philosophy, the most important aspect is "don't force the user to think". Everytime you do that, you destroy the flow, and every time you destroy the flow, frustration results. That leads to the death of a 1000 cuts and the overall feeling that the product sucks, without quite being able to nail it down.

Ideally, there should only be one way to use a product, and that must be the right and obvious one. Only then the user interface allows user immersion, which means that the user doesn't really notice the GUI as such because it feels just as natural as his hand. He doesn't think when he uses his hand, after all.

[Ras]

No you can't. Unless you're a programmer.

I am a programmer, and I even like C. But I'd still not do that because getting into a new code base requires a lot of time, and I can't do that with every project that I think is broken.

even if that GUI does cost money.

Yeah so what, if I get something I really like, no problem. I paid for Shredder because it also offers a very nice GUI, and the copy protection is fair. No "always online" or "installation counter" because such customer hostile garbage rather works as buying protection with me.

Instead, Winboard is like the Crafty of Chess GUIs

That's not fair to Crafty. Crafty was already open at a time when most chess programmers jealously guarded their secrets, and Bob was always willing to discuss ideas. With Crafty, Bob has done a whole lot for chess programming, for everyone.

Xboard, on the other hand, has the merit of showing a chess related aspect as to why desktop Linux has never taken off.

[Ovyron]

No you can't. Unless you're a programmer.

I am a programmer, and I even like C. But I'd still not do that because getting into a new code base requires a lot of time, and I can't do that with every project that I think is broken.

I mean, you can, in theory. Unlike with the Shredder GUI, where, if I want to mix Exclude Moves+MultiPV feature, I just can't, the GUI goes back to including all the excluded moves, or back to SinglePV, without me telling it so, and it has been easier to just hack open source engines to be able to mix the features with an "MultiPV always on" to get what I want (because... I love the Shredder GUI so much that this path has been better to follow than switching to some other GUI that doesn't do this but isn't as user friendly.)

Yeah so what, if I get something I really like, no problem. I paid for Shredder because it also offers a very nice GUI, and the copy protection is fair. No "always online" or "installation counter" because such customer hostile garbage rather works as buying protection with me.

Yeah, same boat here. It's being claimed by some on this thread, that we shouldn't complain about any of Winboard's problems, because it was given to us for free.

Yeah? What about paying me for my time testing the thing? Time I could have used doing something else? Nothing is free, you have to spend time using it.

I remember once I tried to make HGM to fix Winboard, my evil plan was to get Shredder GUI out of business by making Winboard do everything it did but better, and if I recall correctly we got stuck because Winboard shows analysis bottom-top, and Shredder GUI does it top-bottom, but HGM refused to implement a top-bottom option, even though using the engine in the Command Prompt shows it top-bottom!

Instead, Winboard is like the Crafty of Chess GUIs

That's not fair to Crafty.

I meant, for the user. Sure, Crafty might be useful for people wanting to learn how to program their chess engine, but this applies to any other open source engine. And, anyway, with Stockfish being much stronger, and doing everything better than Crafty, an user may be doing themselves a disservice by checking Crafty's source and learning how to do things wrong.

At least Winboard might be the best at chess variant support quantity, according to Evert. What is Crafty best at?

[hgm]

OK, forget about a clear function with a 'safe haven', there is a much better way to do this: "palette invert". Suppose I want to set up a position with very few pieces, say KNNKP. Starting from an internal palette is now cumbersome, because I would have to delete PBRQnbrq, which, even though a single right-click suffices to delete, would still require 8 clicks. And in Chu Shogi, with a palette of 21 (supposing it does not show promoted types), setting up KQKTT from an internal palette would be a true disaster.

The problem is that you can indicate what to delete (and this then immediately happens), but not wat to keep. But if there was an "invert palette" function, you would just delete KNkp, and then 'invert' the palette, meaning that every piece in its palette position will be removed, and palette pieces will only appear on empty squares.

As you virtually always need a King, you can exempt the King from this inversion. So to set up a KQKTT position in Chu, you can start from the palette position, right-click white Queen and black Tiger (so they disappear), apply the pallette invert, to be left with a 'mini-palette' of KQkt, and move those to the desired location, duplicating the Tiger. (For which there are now three methods: move with Ctrl pressed, move with double-click, or lift with right-click and drop twice.)

As it doesn't make much sense to lift the King to remove it or drop it in multiple copies, a right-click on the King could be used as a trigger for inverting the palette, the GUI at that point realizing "oh, this guy was clicking for keeping the pieces, rather than deleting them". In the odd variant where you can have more than one King you can always use Ctrl + move to get a second one. (Or drag it off-board when you don't need one.)

[hgm]

I remember once I tried to make HGM to fix Winboard, my evil plan was to get Shredder GUI out of business by making Winboard do everything it did but better, and if I recall correctly we got stuck because Winboard shows analysis bottom-top, and Shredder GUI does it top-bottom, but HGM refused to implement a top-bottom option, even though using the engine in the Command Prompt shows it top-bottom!

What engines do in the command prompt is not really relevant. Raw engine output is not meant for human reading, and the whole format is awful, both in UCI and WB.

I consider bottom-up better, because the most recent PV would always appear in the same location, so you know where to look. With top-bottom the most-recent PV would start at the top, and would then move towards the bottom.

That I have more important things to do than add optional inferior ways of doing things that are already supported doesn't mean no one else would be willing to do it. Pushing Shredder out of business isn't really a strong motivation for me; SMK is a nice guy, even though he can be blamed for inventing UCI.

There was a discussion about half a year ago of developing the 'perfect GUI', and I suggested that the fastest path to that would be to modify the front-end of WinBoard or XBoard. Then you can make the interface to invoke the existing functionality exactly as you want. But that doesn't mean that I am interested to do that.