[ANNOUNCE] Final NPPAngband 0.5.4

[nppangband]

NPP 0.5.4 has been released. The changes from NPP 0.5.3 revolve primarily around improved mouse support (based on the mouse support found in FAAngband 1.6), enhanced graphics support (inlcuding a complete NPPAngband 32x32 tileset based on the DVG tileset found in Angband), and code changes to help NPP compile on Linux. Also many minor features were added and bugs fixed.

Souce Code:

http://download.nppangband.org/npp-0.5.4-src.zip

Windows Binary:

http://download.nppangband.org/npp-0.5.4-win.zip

Github respository:

https://github.com/nppangband/NPPAngband

The complete changelist can be found here:

http://forum.nppangband.org/viewtopic.php?pid=3929#p3929

[getter77]

Congrats on hitting final, now things get particularly interesting given your early stated aims for .6.x!

I'd also recommend now being as good a time as any to throw up to the Roguebasin for the game, as it is currently being freshly populated with this years 7DRL entrants and thus a good venue to let the release be seen by perhaps folks utterly new to it.

http://roguebasin.roguelikedevelopment.org/index.php?title=Main_Page

[Mikko Lehtinen]

code changes to help NPP compile on Linux

This is awesome. It guarantees that I will play this.

[nppangband]

Congrats on hitting final, now things get particularly interesting given your early stated aims for .6.x!

I'd also recommend now being as good a time as any to throw up to the Roguebasin for the game, as it is currently being freshly populated with this years 7DRL entrants and thus a good venue to let the release be seen by perhaps folks utterly new to it.

http://roguebasin.roguelikedevelopme...itle=Main_Page

Great idea! Thank you.

I also posted to rec.games.roguelike.announce

[getter77]

Nicely done---all of the various variants and the base Angband itself are well served by popping up in the various places releases are mentioned and discussed considering much of what is and has been brewing in these parts and tangential ones otherwise is going very much under the radar of all manner of folks.

[nppangband]

That's not a bad idea. I also checked out temple of the roguelikes (http://www.roguetemple.com/forums/index.php). I only a couple threads on Angband, and not too much discussion at all. I am sure most of that traffic comes here instead, but it doesn't seem like Angband questions on that site get much interest (compared with other roguelikes).

[getter77]

Part of that comes from the old trait as discussed here in the other thread or so about "Angband" being one of the rather older and longer running projects such that for a good number of folks over there the "Angband" from however many years on back now is pretty much crystallized in their mind and the impact of seeing a new release announcement is kind of muted outside of ones that strike out rather different, and differently named, like Sil.

I guess you could say "Angband" is seen as "done" by lots of folks since a long time ago, somewhat similar to Nethack's fate of late---but Nethack has the Mystery Dev Intrigue to keep it topped off on the notoriety front alongside the various variants it has, with I guess the latest dramatic one roughly comparable to Sil here being Wazhack. So challenging the perception that such an "old" project has nothing more to do than fix bugs and rearrange the deck chairs, so to say, is pretty much the thing to do given there's quite a good bit simmering away in here even if there's not currently a flurry of activity as it was back at what I'd imagine the last period people harken back to when Steamband and the other now abandoned variants were part of the vast multitude.

[ghengiz]

minor usability issue for newbies in the GCU port

The announce on roguebasin, which can potentially attract new players, is a good thing, but alas there are still (at least) two or three things that can confuse them in the GCU port.

In the screen when you start the game after a character dies, where it asks if you want to use the previous character, the question is missing.
I know I can just use Y/N to answer, but a new player may be puzzled.
I don't know if this applies to the other ports as well, so this could not be worth a bugfix release.

Slightly related, for it's a glitch too, the whole color related thing I reported in
this other thread is not yet finalized, but it's less severe for a newbie than the former report, so IMO it can be fixed later...till the revamp foreseen in npp 0.6 perhaps?

One last suggestion. In the thread I linked above, you said that the mouse support is not supposed to work with the gcu. So, if possible, it can be sensible to change the defaults in the console version to *not* show the mouse buttons status line

edit: the post about the mouse is this one, while the color issue I was thinking of is the cursor keys not working in the edit colors menu, as reported here.
The colors are still a bit off (especially the red granite walls...), but it's not that bad. I'm grateful for the partial fix you did on the original issue (I mean the orange text, see first post in the thread linked above).

[nppangband]

In the screen when you start the game after a character dies, where it asks if you want to use the previous character, the question is missing.
I know I can just use Y/N to answer, but a new player may be puzzled.
I don't know if this applies to the other ports as well, so this could not be worth a bugfix release.

I will look for this. NPP is currently based on Angband 3.1.2v2, so maybe they fixed it since then.

Slightly related, for it's a glitch too, the whole color related thing I reported in
this other thread is not yet finalized, but it's less severe for a newbie than the former report, so IMO it can be fixed later...till the revamp foreseen in npp 0.6 perhaps?

I gave it my best shot. My problem is when I compile GCU, I get a different color scheme. There are two different color schemes in GCU. You are able to compile with the 256 color scheme, and for some reason I get the basic 16 color scheme. I took my best guess for the color scheme you play with. I don't know why there are red walls. For some strange reason, if you try to load a tileset, it fails but then the wall colors display normal. Probably something in a .prf file. I will try to find it. But in the meantime if you have better color values to suggest for the gcu port, I will be happy to add them.

If you have better suggestions for color values for the gcu port, I will be happy to add them to 0.5.4

[ghengiz]

I did note that the variable COLORS used in main-gcu.c is not initiated anywhere (you may see it with a simple grep -i colors src/*)
I did run some test defining it and various colour related vars (can_fix_color, can_use_color, max_colors, defined in defines.h and main-gcu.c) to every sensible combination of true, false, 16, 128, 256, AND defining the env variable TERM to xterm, xterm-color, xterm-256color.
Some results were better than others*, but, in the end, the default setting you chose is one of the nicer ones, if not the best (at least on my terminal), so I'll stick with it

As for better color values, I did look at the webpage you refer to for the color codes. As I said, the ones you chose (but the walls that become red for mysterious reasons, oh well...) look good enough, so I don't really feel compelled to propose a change.

BTW, does *anyone* have any idea about how to force main-gcu, or whatever is responsible for deciding which colors go to the terminal, to use just the 16 base colors? I hope there is something to set somewhere...my tests were inconcludent


*the worst one had yellow text and violet walls...ugh!

[nppangband]

The colors are still a bit off (especially the red granite walls...), but it's not that bad. I'm grateful for the partial fix you did on the original issue (I mean the orange text, see first post in the thread linked above).

The problem for the red granite walls are in the prf files called font-sdl.prf, font-gtk.prf, and font-gcu.prf. I just uploaded what are (hopefully) fixes to github.

[ghengiz]

The problem for the red granite walls are in the prf files called font-sdl.prf, font-gtk.prf, and font-gcu.prf. I just uploaded what are (hopefully) fixes to github.

[GCU] It's *much* better now, in both TERM=xterm and TERM=xterm-256color, thanks!
The dungeon walls are now white. The problem remains in town, tough...BUT i then tried the change visuals menu, and I think i found the issue:
the terrain codes from 49 to 60 are still somehow changed from the default.
The codes 49-56 are just set to different characters and that's it (mainly the # are replaced by solid walls), but the ones in the range 57-60 are also set to red.
BTW, why is the default for char 55, that is "granite wall (outer)" set to "?" ?

edit:
I changed the current (terrain) codes to the defaults and dumped them, now the colors are right everywhere.

So, the problem seems definitely related to some pref file. Your edits to font-gcu-prf partially fixed it, but alas still not completely out-of-the-box...good job anyway, it's really appreciated

edit 2:
Ah yes, the cursor keys in the change colors menu are still not working, nor is the mouse (but this is the GCU port, so it's expected).
It's not of paramount importance, especially now that the visuals have been fixed elsewhere, but it may be worth to add a few lines to the readme or in the FAQ explaining that the terrain colors may be changed in the appropriate submenu, so there is no need to panic if they can't be changed here

[nppangband]

I think the problem is the terrain numbers are slightly different in NPP than in vanilla (for walls, magma veins, etc), so I need to make sure they match up in the pref files. Those pref files were for vanilla, but I guess I never converted them properly. {I only had a few minutes last night, and I was on my windows machine when I made those changes. I have a bit of a hectic week, and probably won't be able to get to my linux desktop to check everything out until this weekend.}

BTW: I have a way to navigate the color menu in the gcu port. At least with a typical keyboard found in the USA, using the numberpad as direction keys works, but the 4 specific direction keys do not. I don't think editing the colors will work, because the color values displayed in the menu are not used in the GCU port.

[ghengiz]

BTW: I have a way to navigate the color menu in the gcu port. At least with a typical keyboard found in the USA, using the numberpad as direction keys works, but the 4 specific direction keys do not. I don't think editing the colors will work, because the color values displayed in the menu are not used in the GCU port.

[Homer mode]
The numpad as an alternative to cursor keys in a menu? DOH!
[/Homer mode]

Yeah, it's as you say (USA keyboard and all)

[nppangband]

One last suggestion. In the thread I linked above, you said that the mouse support is not supposed to work with the gcu. So, if possible, it can be sensible to change the defaults in the console version to *not* show the mouse buttons status line

There is an option for this (mouse_buttons). The default on but turning it to off hides the buttons.