Results of Second General Discussion

Written by Quintus on 2016-04-10.

Summary of the 2nd TSC General Discussion.

Yesterday, large parts of the TSC Development team met in IRC to conduct the Second General Discsussion (original schedule). In this article, I summarise what was discussed, what was decided, and what not was decided.

First, here’s the (anchored) link to the full chatlogs, so that you can read what exactly happened if you want.

Voting Rules Changes

The team voted on the following change proposal:

                                 Art. 1

§2(2) is changed to “The vote requires that two thirds of the team
members voting are in favour of adding the new person to the team.”.

§3(1)(Nr.3)(2nd sentence) is changed to “This requires a two thirds
majority of those voting, and the candidate for removal is not allowed
to participate in the vote.”

In §20 the words “of the team members” are replaced with “of all
existing team members”.

                                 Art. 2

A new §5a is introduced after §5 with the following text:

§5a IRC Voting Announcement

(1) For votings held in IRC — as an exception to § 5(2)(Nr.3) — it is
sufficient to announce all available voting options in IRC prior to
taking the votes.

(2) If at least three team members are available in IRC and are not just
idling, it is possible to waive the two-week preparation interval and
vote immediately. The waiver requires all non-idling team members in the
chatroom to agree on it.

(3) Any team member can challenge a voting made under the conditions
of (2) above until two weeks after the declaration of results have
passed by approaching the project lead accordingly. If the voting is
challenged, it is treated the same way as if it failed (§12).

                                 Art. 3

These changes enter into force immediately after the Declaration of
Results of the vote for accepting them.

Any votes running at the moment these changes enter into force are
transferred to the changed version of the rules and do not remain under
the old version.

Of 8 team members allowed to vote, 7 participated and voted in favour of it. Thus, the required 2/3 majority was reached, and the changes are in force now.

The newly compiled voting rules document is available online in the documents repository now.

Team Additions

Under the new rules, the team then continued with voting on adding kirbyfan and skarfester to the team. Both votes succeeded. As of now, the two just need to accept their membership on GitHub.

Bugtracker Software

After a fairly long discussion, all team members decided in consent to chose Redmine as our future bugtracking system as it has a lot of nice and useful features (like especially dependencies between tasks as metadata with useful summaries in parent tickets). In addition, it’s written in Ruby, which most of us find a nice language, allowing us to fix things in Redmine itself if we are uncontent with it.

Translation Software

This evergreen of discussion topic was meant to be cleared once and for all in the GD, but it didn’t happen. Albeit we have voted twice on the topic, no winning option could be determined – two options always received the same number of votes. As per §12 of the Voting Rules, this means the vote is failed and has no binding result, and the topic can be voted over again.

The failure of vote, which can spectacularly be seen in the chatlogs, was most likely a result of the low participation of only 62.5% of the team members, i.e. 5 team members, and one of them abstained. It was determined we need to repeat the voting on the mailinglist.

Audience, Story, et al.

Another topic we were unable to decide on was audience and story guidelines. It was noted that this discussion cannot be brought to conclusion without participation of Bugsbane, who was not available in the GD and could not be reached by email, but who is the project’s lead artist and thus must have a word to speak on this topic. Still, some interesting aspects were noted, which are worth to read in the chatlogs. This was also the topics several documents (including story suggestions) were submitted to before. Since there have been no decisions, these documents might be of value for further work, and are thus referenced here:

As a result, we postponed topics 6-10 on the schedule.

Life energy system

A more successful discussion arose around the topic of life energy and its involvement into game mechanics. The discussion benefitted from thoughts by kirbyfan and Luiji, who missed everything prior but managed to join the channel during this topic.

In consent, it was determined that the current one/two hit + lifes + saves + gameover system is to be replaced with a Rayman-style system, which means that the player is given some amount of Hit points (HP). In levels, there will be check points distributed properly (task for level designers), and when the player’s HP sink to 0, he is reset to the last check point he passed. In this system, there is neither manual saving, neither lifes, neither game over. It does not suffer from the problem of infinite points (this saves) by repeatedly jumping on Army and is logically correctly build, i.e. there is no rivalty between different ways to (re)store game progress.

It was further determined that powerups shall not be lost at the first hit a player receives, but there have not been further decisions on when exactly this shall happen then instead.

The Rayman-style progress system was found to be a good balance between not having deathly pits at all, and having too many of them as part of tricky jump sections which may cause frustration on the player’s side. This way, we are able to keep deathly pits, since they will simply have the effect of setting the player’s HP to 0, which means resetting him to the last check point. The discussion about level designers who need to place respawn points for abysses is thus obsolete now.

CEGUI vs. SFGUI

This was the last topic on the schedule, and with regard to the work Quintus has already placed into the CEGUI porting, it was decided we go with CEGUI. SFGUI was also mentioned to be still heavily developed, which was said to not be suitable for our game, which needs a stable API to develop against.

Conclusion

Once again, the GD managed to get the largest part of the TSC team into the IRC channel, allowing to talk to people one usually only sees on the tracker or the forum. The team grew by 2 members, and the decision was made to choose Redmine as our bugtracking system for the future. No decision was made on the story topic, but the life and energy system was finally defined to use a Rayman-like model without lifes, gameovers, and manual saving.