Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Untelligent]

The first number is the same as before, the letter at the end of the old number corresponds to the number at the end of the new system. The other numbers from the old system can be ignored mostly.

In other words, 0.28.181.40d is roughly equivalent to 0.28.04 under the new system.

0.28.10, actually. Before 0.28.181.40a, we had six releases: 0.28.181.39a through 0.28.181.39f (a big release that added a lot of stuff to worldgen and five more minor updates). Then, Toady noticed that there was a minor feature  that was almost finished, and he finished it by adding Obsidian to the economic stone menu. It wasn't a core feature, but it had been planned for a while, so it made the last number go up instead of the first number or the letter. Under the current system, we would have had 0.28.01 through 0.28.10 instead of 0.28.181.39a through 0.28.181.40d.

Toady started using the new version number system because the old system was heavily tied to the old devpage, which included the vast scope of the entire game from closed alpha to v1 and beyond and was eventually deemed too much of a hassle to keep updated. Most if not all of it is still planned to go in eventually, though, so it's still worth a read (some of the powergoals are amusing).

In the old system, the version number included the number of core, req, and bloat features finished. In the new system, the first number is a percentage of core features finished, but there's about a hundred of those (a little more, I think), which is why it's roughly comparable.

[Alev]

The only release I couldn't pin-point a devtime for was the original release.

Dwarf Fortress started October 2002

Now how did I miss that?

Edit: Thanks for that. Now I have that as well.

In have to account for the fact that Toady used to work only pqart time in DF. I don't remember when he started working on DF full time, but it was quite a few months after the inital release.

Some time 2008, he quit his part time (?) math (?)job at a university (?).

[EmeraldWind]

I feel you should make a graph showing the progress of the version number over time.

Here's a simple scatter plot that I set up.

Each point is release. Flat lines are result of bug fix periods.

[Codyo]

I think tendency to have longer and longer periods between major releases is visible. If I was to bet, next major release will be longer than 323, but shorter than 763 days - but ANOTHER release after that will beat 763 days. Nice definition of "outlier" designed to downplay and deny this tendency noted.

I hope I am mistaken and my prediction will not happen.

Yeah I totally agree with you, this graph doesn't show whatever OP is trying to suggest. I see times between majors releases becoming longer and longer. I would say it looks like it's between 500 and 763 days for the next big release at this rate.

For this new graph, I don't really like the line since it implies something. We're not getting these versions handed to us every day over the timespan, we have to wait for them.

[miauw62]

For this new graph, I don't really like the line since it implies something. We're not getting these versions handed to us every day over the timespan, we have to wait for them.

Your point is invalid because of a reason.

[Untelligent]

Yeah I totally agree with you, this graph doesn't show whatever OP is trying to suggest. I see times between majors releases becoming longer and longer. I would say it looks like it's between 500 and 763 days for the next big release at this rate.

For this new graph, I don't really like the line since it implies something. We're not getting these versions handed to us every day over the timespan, we have to wait for them.

I admit that the graphs show a slight trend of the big releases getting longer, but it's worth noting that some trends are eventually self-correcting, and I suspect that DF's big-release-wait-time is one of them. The more framework Toady finishes, the less likely he is to have to add features in big, soupy wads in the future, and this release had quite a lot of framework.


And as for the second graph, the lines DO mean something -- how efficiently Toady core features into the game (strictly in a "how soon we get v1.0" sense). DF2012 and DF2014 both had noticeably steeper lines than DF2010, so there's arguably some improvement there. Wide lines are bad in the short term, but steep lines are good for the long term.


And perhaps we'll be getting some lines over the next couple years that are steep, but not wide. Maybe we won't. Let's wait n' see.