Halls of Mist plans

[Mikko Lehtinen]

As you may know, I'm planning to take everything apart yet again. My development version of Halls of Mist will eventually be very different from the current version. New combat system, new magic system, new stats, new everything really!

But I actually really like the current version, too. I've decided to keep developing it under the name Halls of Mist "Classic".

For now, the plan is to make the Classic version as bug-free and balanced as possible. The recent competition gave me lots of valuable feedback.

I should probably shelf the development version for some months and concentrate on fixing the Classic first...

[Mikko Lehtinen]

Can someone recommend a good tool to help me share the bugfixes and such for both versions of the game? I'm on Linux.

I don't know much about good coding practices, since I've learned everything I know about C from reading Angband source code.

[Magnate]

Can someone recommend a good tool to help me share the bugfixes and such for both versions of the game? I'm on Linux.

I don't know much about good coding practices, since I've learned everything I know about C from reading Angband source code.

Depends what you mean by share bugfixes. I find Trac absolutely brilliant as an issue tracker (i.e. a way to record and maintain the status of each bug and its fix), but not sure if that's what you meant. See trac.rephial.org if so.

For actual sharing of code you can't get any better than github. If Halls is not already there just get an account, upload it and share your bugs with the world. You'll be amazed how quickly the fixes come flooding in!

[Mikko Lehtinen]

I'm not really sure what I meant, either. Thanks for the good ideas, anyway!

I need to develop a strategy for updating the two versions of the game at the same time. Many of the fixes will be applied to both versions.

It seems a lot of hassle to first fix a bug in one version by changing (say) three files, and then doing the same changes in the other version. There's a big possibility that something goes wrong if I only keep track of the changes manually.

I'm working in Vim, and I know it has a good tool for showing differences between two files. That should help a bit.

[getter77]

Interesting plan to split ahead into 2 versions----I shall be watching it unfold closely!

Are you planning on any new technological or UI overhauls for the Next Edition to set it on different footing versus Classic alongside the various new gameplay systems?

[Mikko Lehtinen]

Are you planning on any new technological or UI overhauls for the Next Edition to set it on different footing versus Classic alongside the various new gameplay systems?

No technological overhauls. It would be reinventing the wheel, since I'm working on an ancient Angband codebase. But maybe some day I will change the codebase to Pyrel!

Some UI overhauls, yes. Less commands, for one thing, and a *much* quicker spellcasting system.

[Mikko Lehtinen]

One of the reasons for the split is that I'm actually removing *lots* of content in the next version. I'll only keep five classes, for example. As Halls of Mist is the only living variant that keeps the spirit of EyAngband alive, it seems proper to not just destroy everything...

Another reason for the split is that the current version has been in development for a long time, starting from FayAngband. Evolution has slowly removed all my pet peeves in Angband. Keeping the Classic version around allows me to relax and just enjoy the benefits of my work. It also takes away the stress and hurry from developing the next version.

The relationship between FAangband and Beleriand seems quite similar, although thankfully the scale of my games is much smaller.

[LostTemplar]

Can someone recommend a good tool to help me share the bugfixes and such for both versions of the game? I'm on Linux.

Just using GIT, having separate branches in the same repository for different versions and cherry-picking commits works more or less.

[Mikko Lehtinen]

Just using GIT, having separate branches in the same repository for different versions and cherry-picking commits works more or less.

Wonderful. I'll do some studying on Git. (I found an article on Using Git with Vim.)

Is Git or Subversion better for my simple needs? I know almost nothing about them. Maybe I should read some comparison articles.

[Derakon]

For a single developer, Git vs. Subversion is roughly equivalent. But if you ever want to work with multiple people (even if it's just so others can submit patches) then you want to be using Git or another distributed versioning system (like Mercurial, my personal preference).

Also, as a matter of course you should be setting up a repository for every coding project you do. There's practically no cost to you and the advantages are significant. You should never need to rely on the undo history of your text editor in case you need to back out a change, for example.

[Mikko Lehtinen]

In my work, I write and edit all kinds of text, and design books, magazines, and sometimes websites. How useful are these tools for version control in non-coding projects? I prefer to work with plain text for as long as possible.

I often work with other people, but mostly they are very non-technical folks. Just a tool to keep my own plain text files in order might be very useful...

[getter77]

Gotcha on the above, UI and less commands to muddle the waters are key in general!

As an aside, what is the current state of Pyrel in terms of "technological" and UI gains over the old as it stands? (My current hatred is font wrangling---I shake my head at thee, Sangband mini-list of commands window while the rest is playing somewhat nice at full screen or large window.._ There's probably feature begging/reckonings to be had~

Tangential aside: For your earlier crazy player manual/lore notions, and your above doings anyways: Why not go totally off the rails and do it up in fine Diskmag style ala Hugi or some such? A Diskmag Roguelike mega-manual would be oddly fitting I think, given the shared notions between the pair even if the former has fallen far from the limelight compared to the latter.

[Derakon]

As an aside, what is the current state of Pyrel in terms of "technological" and UI gains over the old as it stands?

Pyrel is currently behind in terms of UI, since we're still working on basic game mechanics. Proper UIs are going to be put off until either some dev makes them a priority or we run out of other things to work on. The problem with implementing UIs early is that when the game engine changes, the UI has to be adapted to suit, and if the engine changes often, then keeping the UI up to date can create a lot of extra work. So usually projects just stumble along with a minimal UI until the game engine reaches some level of stability.

This is not to say that we're going to turn into Dwarf Fortress here, where the dev never works on UI because he has an endless list of game-mechanic changes to make. Pyrel has a pretty fixed feature list, so if nothing else, once those features are done the devs will mostly turn to working on the UI.

[Mikko Lehtinen]

Tangential aside: For your earlier crazy player manual/lore notions, and your above doings anyways: Why not go totally off the rails and do it up in fine Diskmag style ala Hugi or some such? A Diskmag Roguelike mega-manual would be oddly fitting I think, given the shared notions between the pair even if the former has fallen far from the limelight compared to the latter.

The two versions will actually have very different player manuals.

Currently the various game mechanics are quite hard to explain to the player. I need different chapters for combat, magic, etc. Lots of long, awkward paragraphs explaining as awkward game mechanics.

In the development version I've redesigned the game so that almost everything the player needs to know is in just one section: The Skills. Once you go through the list of skills and their various uses, you should pretty much understand how the game works.

All the skills checks are simple 1d100 rolls under skill score, with no modifiers. On the character sheet (the 'C' command), skills are nicely listed under stats.

Perhaps my player manual could be even shorter than the excellent Sil manual...

I like PDFs. But if I write the root file of the manual in markdown, I should be able to transform the manual into many different versions, like a web page or in-game help files.

[Magnate]

I'm not really sure what I meant, either. Thanks for the good ideas, anyway!

I need to develop a strategy for updating the two versions of the game at the same time. Many of the fixes will be applied to both versions.

It seems a lot of hassle to first fix a bug in one version by changing (say) three files, and then doing the same changes in the other version. There's a big possibility that something goes wrong if I only keep track of the changes manually.

I'm working in Vim, and I know it has a good tool for showing differences between two files. That should help a bit.

This has github written all over it - it's exactly what we did when we forked v4 - we've applied a ton of Blubaron's UI fixes to both versions with minimal hassle.

Unfortunately github doesn't allow you to fork your own repositories, so you'll need to upload them separately and link them on your local machine. But it will then allow cherry-picking of commits between the two with ease.