Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

On Week 1/2012

edited 2012 Jan 12 in Developers

Last week was quite busy with Doomsday development as we were working full steam on the Candidate phase for the first time. My efforts were centered around fixing multiplayer regressions. Things are working almost well enough for me start the public servers again. There is one more thing I need to do before that: make sure that the client will not get confused about surface materials if (when) the server is using a different set of resources.

For the next couple of weeks we will be focusing on making the required fixes in order to get 1.9.7 into a state that we can with good conscience call "stable." We will not be able to fix everything in our bug backlog for this release; instead we are targeting the selected criticial issues listed on our 1.9.7 todo list. Our plan is to address the rest of the bugs gradually over the course of future stable releases, which will be arriving at intervals of 2-3 months after the stable 1.9.7 is out.

For the duration of the Candidate phase, new builds are available on Mondays, Wednesdays and Fridays. If you're interested in fixes for a particular bug, keep an eye on the list of changes in each build. You can find the builds in our Build Repository.

My plan is to continue working on the items in the 1.9.7 todo list, starting with the material dictionary.


Last week I was focused on resolving some of the more critical issues in the current Candidate builds. As mentioned in my last update, the most important of which being the regressions with addon loading.

Closer inspection revealed that the resource locator was returning all resources in all subdirectories descendent from a given search path. This was what caused the animation issues with the DHTP and Plutonia's textures being mixed in with DOOM's, for example. The issue has now been addressed for today's build 374, so any addons loaded from the console or the /auto data directory should now be fully working.

There however remains an issue with the handling of the -file argument, meaning that addons loaded from within Snowberry are not yet working as they should. This issue will be addressed for Wednesday's build.

In addition to resolving the addon loading issues I also spent some time optimizing the file system.

This week my plan is to resolve the remaining issues with addon loading and then focus on cleaning up some of the more minor issues, like object alignment in the game menu.


  • Nice to hear that from this point forward, releases will be quick. Keep up the good work. Also, a major release coming out every 2 to 3 months is excellent news. That is quick. It's nice that so much life has been breathed into this project recently and apparently won't cease anytime soon. ;)
  • Seconded. The progress is amazing!
  • gary wrote:
    Also, a major release coming out every 2 to 3 months is excellent news. That is quick.
    Just to clarify: not a "major" release, but a stable one. We'll be working just as fast as before, but the automated release mechanism allows us to make stable releases more often. Consequently, the number of changes per release will be smaller.

    In other words, we'll be able to release features as stable soon after they are finished, one by one, instead of releasing them all in a big lump as we're forced to do with 1.9.7.
  • When you say "just as fast as before", are you referring to the slow progress of the last several years or referring to the speed of the last couple months? I recall in a recent update Dani saying that from this point forward, you will be working on all 4 cylinders. So I'm hoping and guessing you mean that you will continue working at the heightened pace of the last couple months.
  • gary wrote:
    When you say "just as fast as before", are you referring to the slow progress of the last several years or referring to the speed of the last couple months?
    I'm referring to the average speed we've had since March 2011.
  • Well the way Dani was talking, it makes me think that progress will be even faster and sustained than in 2011. The term "all four cylinder's" was something he said recently.

    Edit: Wait, you said it:

    "The ending of the ringzero branch will also leave us in the fortunate situation of having only one primary development branch. Both I and DaniJ will be able to contribute more directly to the main codebase, and all the changes will end up in the unstable releases much faster. Development-wise, we will be firing on all cylinders after this."
  • That is correct. Remember that for the entirety of 2011, the development efforts have been split between two branches (DaniJ's and mine) and thus less than half of what we've been doing was actually visible in the releases. The "all cylidenders" refers to the dual impact of 1) being able to concentrate all efforts into the master branch, as we've begun doing since the beginning of the year, and 2) after we finish the work started in the 1.9 beta versions, we can move on to more interesting features waiting on the roadmap.

    However, Doomsday is still a hobby project for me and DaniJ and as such more important matters may take precendence from time to time; that's why I said "average" speed.
Sign In or Register to comment.