Developer Meetings/20071013

From K-3D

Jump to: navigation, search

To Attend

  • Meeting Time: 20071013 1700 UTC
  • Timezone Information: www.timeanddate.com
  • Meeting Protocol: XMPP (Jabber) Conference
  • Meeting Location: k-3d@conference.jabber.org

Agenda

  1. Meta-discussion: should we switch to audio/video for the meeting? The intent would be to improve the bandwidth / address RSI issues.
  2. Need to setup an umbrella organization to handle incoming revenues.
  3. What do we need to do before releasing 0.7?
  4. Technical Issue: standard modifier behavior
    • What to do with input data of the wrong class (i.e. do you discard patches if you're a polyhedron-modifier?)
    • Output selections - should all newly-created geometry be selected? All components?
  5. Our recruiting and retention are suboptimal ;) - and our mailing-list subscriptions have been flat for years. What do we do to shake things up?
    • Create contributor pages with pics, biographical info, etc.
    • Live demos using VNC / Skype / foo
    • Setup a marketing "team"
      • Develop slide presentations
      • Post releases to a concrete list of venues
      • Improve website
      • Ensure everyone has t-shirts / caps / foo
    • Attend open-source conferences
    • Attend SIGGRAPH
    • Developer sprints
    • Bug-days
    • Community bios with photos
  6. Animation: Where to go from the current state? Curve editor or other priorities?
  7. GSoC 2008 ideas

Action Items

  • [Shead] - Contact SPI and SFC.
  • [Duo] - Contact SPI projects re: SPI.
  • [Janssens] - Contact Inkscape project re: SFC.
  • [Duo] - Start work on a K-3D 0.6 script for freezing documents.
  • [Shead] - Work with Joe on Plugin Behavior Guidelines wiki page.
  • [Shead] - Start work on immediate 0.7 release.
  • [Janssens] - Work with Joe on Display Behavior Guidelines wiki page.

Discussion

(11:00:39) Tim Shead: Shall we get started?
(11:00:44) joe: sure
(11:01:00) Tim Shead: So welcome, everybody!
(11:01:59) Bart: Hi all!
(11:02:07) Tim Shead: I wanted to kick-off our first meeting by discussing if IM is the best way to work.
(11:02:27) Tim Shead: I know that there are potential RSI issues for some of us ...
(11:02:40) Tim Shead: ... and one of my major complaints with IM is the very low bandwidth.
(11:02:45) joaduo: HI
(11:03:20) Tim Shead: Ideally, I think we should be using audio for these meetings.
(11:03:30) Bart: Advantages are that you get a complete log, and it's very clear
(11:03:42) joe: I'm mostly ok with it, but it would be nice if there was some easy graphical meeting software to go with it, like an easy "whiteboard" type thing or an easy place to drop pictures
(11:03:55) Bart: Internationally, spoken English may be more difficult
(11:04:01) joe: i definitely think having jabber as a base is a great start
(11:04:10) joaduo: i agree
(11:04:10) joe: yes, Bart, very true
(11:04:39) joaduo: audio is great but sometimes there a lot of delay
(11:04:39) Tim Shead: So, for whiteboarding, there's a link posted in today's meeting page, has anyone tried it?
(11:04:48) joaduo: i tried it
(11:04:58) joaduo: doesnt work on the first try
(11:04:58) joe: no, not yet
(11:05:13) joaduo: i was searching for a better soft
(11:05:20) joaduo: but couldnt find
(11:05:24) Bart: just looked at the screenshots :)
(11:05:45) joe: that link looks really cool
(11:05:52) joaduo: coccinella on ubuntu feisty fawn. It freezes starting
(11:06:17) joaduo: i have runned it before, at that time it has some bugs...
(11:06:25) joaduo: so i wanted to try it now
(11:06:29) joaduo: but couldnt
(11:07:02) joe: you'd think that an easy to post to / update webpage would serve fine too, just have a window beside jaber, then no special software would be required
(11:07:31) joaduo: the wiki?
(11:07:46) centipede901 [centipede901@gmail.com/HomeF82A2FBD] entered the room.
(11:07:48) Tim Shead: That may be the best solution in the short-term.
(11:08:05) joe: as long as it was easy to upload graphics to that page etc. Something that showed time stamps for each post would be great to, because it would sync with the logs for the IM, and it would provide its own record
(11:08:38) joaduo: when you sign it includes the time stamp
(11:08:53) Tim Shead: In the wiki, just do "[[User:Tshead|Tshead]] 11:45, 13 Oct 2007 (MDT)" to get a name and timestamp.
(11:09:00) Bart: Maybe VNC+voice chat would be nice for some more involved demos
(11:09:09) joaduo: cool
(11:09:34) joe: does the wiki handle a bunch of people making posts simultaneously?
(11:09:39) joaduo: we use the article for the meeting?
(11:09:55) Tim Shead: Yes, let's use the meeting page.
(11:10:01) Tim Shead: Not sure about locking.
(11:10:01) joaduo: http://www.k-3d.org/wiki/Developer_Meetings/20071013/images for example
(11:10:22) joaduo: o directly Developer_Meetings/20071013/
(11:10:48) Tim Shead: My intent is to post the logs to the meeting page itself, I don't see any reason to separate images.
(11:11:08) Tim Shead: We'll know better what works when you guys start posting ;)
(11:11:18) joe: sweet
(11:11:30) Tim Shead: Bottom-line, let's go with IM+wiki for the short-term.
(11:11:38) Bart: ok
(11:11:51) Tim Shead: Second item on the agenda: setting-up a non-profit.
(11:12:38) Tim Shead: The K-3D project will be receiving either $500 or $1000 from Google for our participation in SoC 2007
(11:13:01) Tim Shead: I'm not sure of the amount because I forget whether they are paying us for failed projects or not?
(11:13:02) Tim Shead: Bart?
(11:13:32) Bart: Yes, they are, I think
(11:13:58) Bart: So it should be $1000 :)
(11:14:03) Tim Shead: Either way, we need to handle the money.  If it goes to an individual such as yours truly, we have to pay personal income taxes on it.
(11:14:29) Tim Shead: Fortunately, I made a temporary arrangement with Google to withold our payment for an indeterminate amount of time.
(11:14:45) Tim Shead: So we have some breathing-room to setup a non-profit.
(11:15:03) Tim Shead: There are a couple of links in the meeting page.
(11:15:45) Tim Shead: Both SPI and SFC are "umbrella" organizations that are incorporated in the USA ...
(11:16:33) Tim Shead: The idea is that FOSS projects can join either of these organizations.
(11:17:02) Tim Shead: The umbrella organization has nonprofit status, so they can receive money on behalf of their member projects.
(11:17:30) Tim Shead: For both organizations, donations are tax-exempt, which is nice.
(11:17:41) Tim Shead: And they handle all the paperwork / fees / etc.
(11:17:53) Tim Shead: Both provide some level of legal representation to member projects.
(11:18:20) Tim Shead: In general, I don't see us setting-up our own corporation, because it's the kind of work I hate :)
(11:18:37) Tim Shead: So I think we should look closely at both organizations, and see if we want to join either.
(11:18:45) Tim Shead: Initial thoughts?
(11:18:56) Bart: SPI had a 5% fee, any idea about SFC?
(11:19:17) joe: wow, 5% is a lot!
(11:19:23) Tim Shead: From their page, I believe there are no fees.
(11:19:53) Tim Shead: Of course, 5% beats 28%, which is what we're looking at if I accept the money personally ;)
(11:20:14) joaduo: we can consult other projects involved on that organization
(11:20:16) joe: ya for sure, I'm just thinking you can probably beat that...
(11:20:20) joaduo: to compare opinions
(11:20:28) dahnielson: I was planning to only lurk, but: Rolling your own organization is usually more work than it's worth, unless you plan to handle a lot of donations. Lots of paperwork and I believe you must get a certain amount of donations to keep the non-profit status.
(11:20:56) dahnielson: Of course, IANAL and IAMNAM (I Am Not An American) ;-)
(11:21:24) joe: yes, a custom organization is probably way too much work
(11:21:37) Tim Shead: I think rolling-our-own is off the table at this point.
(11:21:43) Bart: yeah
(11:21:53) joaduo: i meant, compare sfc and spi
(11:21:56) joaduo: .)
(11:22:03) Tim Shead: My intent was to contact both sfc and spi officially.
(11:22:20) joaduo: can we be in both?
(11:22:23) Tim Shead: What would be useful is if someone would volunteer to try and get feedback from member projects.
(11:22:42) Tim Shead: There's no value in doing both that I can see - they provide similar services.
(11:23:14) joaduo: ok
(11:23:46) Tim Shead: The V-word made it get real quiet.
(11:24:24) joaduo: V-word??
(11:24:29) Tim Shead: "volunteer"
(11:24:30) dahnielson: he eh
(11:24:31) Bart: I'm leaning more towards SFC... they have smaller projects, in SPI you have some huge projects, meaning K-3D would not be of uch importance to them...
(11:24:33) joaduo: aaah
(11:25:05) Bart: I could mail the Inkscape guy who was at the SoC summit
(11:25:10) Tim Shead: Cool!
(11:25:34) Tim Shead: So, I'll contact both, see what response we get, and we'll have more to go on next week.  Sound good?
(11:25:36) joaduo: ok, i try with spi
(11:25:44) joaduo: when is the deadline?
(11:25:50) joaduo: for the "report"?
(11:25:50) Bart: ok
(11:26:12) joaduo: ok next wekk
(11:26:15) joaduo: week
(11:26:20) joaduo: didnt see it
(11:26:41) Tim Shead: On a general note, let's say that action-items are normally for the next week -
(11:26:53) Tim Shead: anything more open-ended tends not to get done in my experience.
(11:27:20) Tim Shead: So on to item #3
(11:27:45) Tim Shead: Thoughts on what (if anything) we need to get done before a release of K-3D 0.7?
(11:28:00) Tim Shead: The issues involved -
(11:28:25) Tim Shead: On the one-hand, we haven't had a release of any kind for a long time.
(11:28:36) Tim Shead: That hurts us because people think nothing's going on.
(11:29:06) Tim Shead: On the other-hand, 0.7 is a long way from being 0.8, so there will be compatibility issues, regressions, partially-complete features, etc.
(11:29:30) joaduo: well on .7 its supposed that mess arrays are finished?
(11:29:34) joaduo: mesh
(11:29:43) Bart: Well, I think the main issues are animation and texturing... Not having those makes people won't see it as complete...
(11:29:55) joaduo: I desagree...
(11:29:58) joe: why not just release 0.7 and call it a development release?
(11:29:58) Bart: But we could market 0.7 as a development release
(11:29:58) joaduo: :)
(11:30:05) Tim Shead: As major features, it won't be 0.8 until we have animation and texturing.
(11:30:08) Bart: yes :)
(11:30:26) Bart: having releases will encourage a lot more testing
(11:30:29) joe: yes
(11:30:35) Tim Shead: I'm just talking about development releases to whet people's appetites and let them know good things are happening.
(11:30:50) Tim Shead: Has everyone here run 0.7?
(11:30:55) joe: regular releases are good even of development versions
(11:30:58) joe: yes, I have
(11:31:01) Bart: then nothing stands in our way!
(11:31:19) Bart: I should try to get the booleans working on win32, though
(11:31:19) joe: but only after quite a lot of difficulty getting it to compile
(11:31:28) Tim Shead: Anybody think we shouldn't release what we have right now, today?
(11:31:38) joe: I say go for it
(11:31:59) joaduo: i agree ...
(11:32:20) joe: also, since I can't seem to build on win32 without a lot of trouble, a new release means I can demo newer things when I show people the software
(11:32:40) Tim Shead: What's the problem with the build?
(11:33:27) Tim Shead: Actually we should talk about that off-line and not get distracted ... OK, we will do a 0.7 release ASAP
(11:33:48) joe: I think its the fact that all the dependencies aren't prepakaged an I had to compile boost etc and somewhere along the way I kept getting stuck. I haven't built on win32 since the old code::blocks system
(11:33:55) joe: sounds good!
(11:34:04) Tim Shead: Any last thoughts before moving on?
(11:34:30) Tim Shead: OK, item 4 ...
(11:34:35) joe: just that more releases, even point releases without much extra information are very good
(11:34:55) joe: to keep people aware of new development
(11:35:00) joe: and excited
(11:35:09) Tim Shead: Agreed, carpe diem!
(11:35:28) Tim Shead: So item 4 touches on a larger, very critical issue with backwards-compatibility.
(11:36:12) Tim Shead: As I am writing/rewriting plugins using the new mesh API, I have to deal with making them compatible with the originals.
(11:36:22) Tim Shead: e.g: rewriting TriangulateFaces.
(11:37:12) Tim Shead: One item that has come up is that I often find myself in the position of not wanting to duplicate behavior precisely, because it is suboptimal.
(11:37:40) Tim Shead: In many cases, the current behavior is what it is simply by default - I never thought about it at all!
(11:38:26) Tim Shead: This stuff is critical because any change, no matter how small, to the output of a plugin can wreck a saved document.
(11:38:41) Tim Shead: For this reason, I want to do the following:
(11:38:59) Tim Shead: * Create new plugins rather than rewriting old plugins.
(11:39:09) joaduo: cool
(11:39:12) Tim Shead: * Deprecate the old plugins
(11:39:27) Tim Shead: * Remove the old plugins at some point in the future (preferably prior to 1.0)
(11:39:45) joe: if we have freeze history and freeze transform, then we can also lock the end result down and not worry about compatibility of the entire pipeline
(11:40:11) Tim Shead: That's true, but we would need to either do it for the user, or get them to do it, before the deprecated plugins go away.
(11:40:30) joe: yes, but that's ok
(11:40:51) joe: just warn people that the new release doesn't support those old things
(11:41:00) Tim Shead: The other side to this, is that we need to have some better specs on how plugins should operate, to avoid this problem in the future.
(11:41:09) joe: at this point I don't think its a big deal because no one has any mission critical files
(11:41:11) Bart: Also, the deprecated plugins should be removed from the menu, and only used when loading an old document
(11:41:22) joe: the software as a whole isn't in wide use
(11:41:24) Tim Shead: Good point, Bart.
(11:41:37) joe: yes, bart, good point
(11:41:45) Tim Shead: Joe, you're the real artist here ... no documents you'll miss?
(11:41:45) joaduo: uhm... i think we can have the old k-3d "engine" working at the same time as the new one
(11:41:55) joaduo: well what bart said
(11:42:28) joaduo: Tim it would be great that plugins would have a long last live
(11:42:28) Tim Shead: joaduo: yes, we can leave deprecated plugins around indefinitely, but there is a long-term cost paid in terms of support and maintenance.
(11:42:36) joe: Tim: any document I wanted to keep, I would just want to delete the history and freeze transforms, then I'd have it in the new version anyway
(11:42:38) joaduo: yes...
(11:42:49) joaduo: i suppose we could set a treshold
(11:42:59) joaduo: but that would encourage developers
(11:43:14) joaduo: they would know that plugins will be used for a long time
(11:43:21) joaduo: i suggest :)
(11:44:03) Tim Shead: Anybody want to volunteer to write a script for K-3D 0.6 that freezes the pipeline as Joe suggests?
(11:44:29) joaduo: i don't know how to do it...
(11:44:38) joaduo: i suppose with python :D
(11:44:41) Tim Shead: That sounds like a volunteer!
(11:44:44) joaduo: :P
(11:45:27) joaduo: ok, i will try
(11:45:34) joaduo: that's for next week?
(11:45:44) joe: i do think that "freeze" is on our shortlist of most necessary features to add. It seems ridiculous to have harmonic quad remeshing, but not a freeze button
(11:46:19) Tim Shead: joaduo: you should try to have something, even if it's hello world ... the point is just to get started and make some progress.
(11:46:39) Tim Shead: Joe: fair enough, but it will be in 0.7.
(11:46:45) joaduo: but the freeze thing is to apply the freeze mesh plugin to the end of an object pipeline?
(11:47:15) Tim Shead: Yep, you would iterate over all the mesh instances and insert freeze on their inputs.
(11:47:19) joe: not just that, but also to delete any nodes that were used in the pipeline before the freeze, and not used by anything else
(11:47:40) joaduo: oh...
(11:47:40) joe: its also for scene optimization
(11:47:52) joaduo: and for desser?
(11:47:55) joaduo: dessert?
(11:47:56) joe: models in k3d currently get crazy heavy and complex even when simple
(11:48:13) joe: because there is no way to delete the huge huge modifier stack that gets created over time
(11:48:32) joe: literally thousands and thousands of nodes even for relatively simple models
(11:48:33) Tim Shead: This is very experimental - you should look at it as research.  You're not signing-on to deliver something in a week :0
(11:48:54) joaduo: yes i know :D
(11:49:24) Tim Shead: OK, so that's the backwards-compatibility half of this issue.
(11:50:05) Tim Shead: The other half is establishing some guidelines on source/modifier behavior, so we minimize these types of breakage in the future.
(11:50:12) Bart: I guess for 0.6 we should aim for a "backwards compatibility" script, and for 0.7 a button or context menu item that freezes the mesh when the artist is happy with it and wants to lighten the pipeline?
(11:50:20) Tim Shead: Bingo!
(11:50:26) joe: for sure
(11:51:00) joe: I should take a look at my currect build and suggest a place for it in the UI
(11:51:09) joe: I think a menu item is the best start
(11:51:18) joe: perhaps on the toolbar too
(11:51:25) joaduo: uhm...
(11:51:26) joe: its a very common button for artists to push
(11:51:35) Bart: Could the 0.7 freeze be made a little more dynamic, maybe disconnecting the prior nodes, storing them, and giving the artist the ability to resurrect them?
(11:51:43) joaduo: i guess there should be a better pipeline stack GUI
(11:51:55) Bart: i.e. also add an "unfreeze" button?
(11:52:16) joe: I think the unfreeze button doesn't make sense
(11:52:26) joe: unless it gets stuff from disk
(11:52:41) joe: because the whole point of freeze is to get the info out of RAM 
(11:52:51) Bart: ko
(11:52:52) Bart: ok
(11:52:54) Bart: :)
(11:52:55) joe: and have the CPU no longer calculating those modifiers
(11:53:27) Tim Shead: So, designing freeze as a feature is a good item for a wiki page.
(11:53:34) joaduo: yes...
(11:54:04) Tim Shead: I do want to touch on the kinds of questions to be answered for source/modifier behavior (also destinied for a wiki page) ...
(11:54:05) joaduo: but i mean, for example, there is no easy way to move modifiers in the pipeline stack
(11:54:32) joe: no, or to toggle them on and off, and I have a feeling that an on/off toggle would be easy to make
(11:54:40) Tim Shead: It's already in 0.7
(11:54:47) joe: really?
(11:54:48) joaduo: yes
(11:54:52) joaduo: tim made it
(11:54:53) joe: awesome!
(11:54:55) Tim Shead: I would not lie.
(11:54:58) joaduo: :)
(11:55:07) joe: you rule Tim, I will try this as soon as I get to my office
(11:55:24) joaduo: something alse
(11:55:26) joaduo: else
(11:55:44) joaduo: its not easy to insert a  modifier on the pipeline
(11:55:51) joe: that's right
(11:55:53) joaduo: if it is not on the top
(11:56:14) joaduo: i guess we should design that GUI
(11:56:18) joe: something quite important is to go to a particular point in the pipeline, and work there
(11:56:32) joe: the history panel should accomadate that eventually
(11:56:48) joaduo: although that could be for a next release
(11:56:49) joaduo: tim?
(11:57:01) Tim Shead: Pipeline editing is a huge topic, there's lots of work there, it's really too much for one meeting.
(11:57:02) joe: yes, we should let nothing delay the current release
(11:57:55) joe: also, I think before getting to pipeline editing, there's a lot of little things that need fixing from user test feedback. More on that later...
(11:57:58) Tim Shead: We are coming-up on one hour for the meeting.  Does everyone want to continue, or call it a day?
(11:58:19) joe: I wanted to talk about marketing a little, and about my experiences demoing software
(11:58:20) Bart: continue
(11:58:25) joe: and also about the connections I have
(11:58:37) joaduo: i guess the modifier thing should be talked on the mailing list
(11:58:41) joe: but after the other discussions are mostly good
(11:58:53) Tim Shead: OK, I'm going to suggest some topics to think about, and then move on.
(11:59:28) Tim Shead: Remember that meshes can contain any combination of geometry - polygons, patches, curves, etc.
(12:00:01) Tim Shead: What is the correct behavior for a modifier that only works with one type?  If there are other types in its input, does it pass them through unmodified, discard them, or what?
(12:00:13) joe: pass them through unmodified, for sure
(12:00:20) Tim Shead: Fair enough.
(12:00:22) joaduo: i agree
(12:00:47) Tim Shead: When it comes to selections, what should mesh modifiers produce as output?
(12:01:08) joe: this might be the case already, but in general, any modifier that only deforms things should automatically work on all types, because they are all based on points in space
(12:01:09) Tim Shead: The legacy TriangulateFaces selects some components (faces) but not others (edges), for example.
(12:01:22) Tim Shead: Joe: yes, that's already true.
(12:01:29) joe: sweet
(12:01:52) joe: I actually wrote up an answer to this earlier... hang on
(12:02:02) joe: for output selections
(12:02:50) joe:       This depends on workflow.    General rule of select new geometry makes sense, but often other    things are better. Extrude should only select end poly, not side    polys. Artists/users can say how they are likely to use a    modifier.  Also properties can exist that change what output    selection gets made.      
(12:03:02) joe: man, the formatting pastes in jabber badly
(12:03:04) joe: sorry
(12:03:21) Tim Shead: What about faces vs. edges vs. points?
(12:03:30) Tim Shead: Can we make any general statements?
(12:03:40) joe: again, that depends, and yes, we can make some general statements
(12:04:02) Tim Shead: OK, I'll start a wiki page on that, and we can hash it out as we go.
(12:04:06) joaduo: what if that depends on the plugin developer?
(12:04:09) joe: extrude face should select the face as output, extrude edge should select edges
(12:04:13) joaduo: i mean, his decision
(12:04:26) Tim Shead: These are guidelines.
(12:04:28) joe: however, connect verticies should output the edges, because they are what gets created
(12:04:47) Tim Shead: The point is that we never really thought it through in the past, so the current behavior is somewhat arbitrary.
(12:05:01) Tim Shead: For every guideline there will be exceptions.
(12:05:09) joaduo: ok
(12:05:11) joe: wings3d is so excellent because the developers have really thought about output selections
(12:05:25) joe: and the output selection is always ideal for the most likely next operation
(12:05:33) Tim Shead: That's what I mean.
(12:05:43) joaduo: joe, maybe we could create a table for every modifier
(12:05:47) joaduo: on the wiki
(12:05:48) joe: subdivide edge outputs vertices and then connect vertices outputs edges
(12:05:53) joe: I would be happy to do that!
(12:06:05) joe: which page should I edit?
(12:06:17) Tim Shead: I'll let you know once I create it ;)
(12:06:21) joe: sweet
(12:06:24) joe: please email me then
(12:06:54) Tim Shead: So I think that meetings that go beyond 90 minutes-or-so are counterproductive, so I want to move-on.
(12:07:00) joe: cool
(12:07:06) Tim Shead: The next item on the agenda is very open-ended
(12:07:19) joaduo: yes :D
(12:07:38) Tim Shead: It's about the need to really reinvigorate the community, and get things hopping in a way that we've never done before.
(12:07:58) Tim Shead: So, I'm going to hand it over to Joe, since he mentioned some related topics.
(12:08:05) joe: awesome!
(12:08:42) joe: Ok, so let me start by saying that my main use to this community is really in dealing with industry people and marketing etc
(12:09:05) joe: I live in a city that is a huge hotbed of game development and graphics work
(12:09:09) joe: (vancouver bc)
(12:09:28) joe: I also have contacts in pretty much every big studio, and a lot of them are even my clients
(12:09:42) joe: the point is, I have ample opportunity to get k3d in front of people
(12:09:50) joe: and I definitely like to do that
(12:10:37) joe: people are usually very impressed by many things, however, the problem I generally get is that people ask "when will it be ready" (for basic usage they mean)
(12:11:08) joe: I usually say "oh, a couple of years, but you can download it now and try it out"
(12:12:16) joe: but usually it comes down to the fact that certain core operations are expected, and when those certain basics don't work, people I show it to consider that its very experimental
(12:12:39) Bart: so we have to work harder :)
(12:12:43) joe: So one of the key things for more adoption and better marketability is a back to basics approach I think
(12:12:47) joe: its not working harder
(12:12:55) joe: its just doing to simple things first
(12:12:57) Tim Shead: I'm going to break in here ... yes, working harder is not the answer.
(12:13:15) Tim Shead: We're all volunteers, and I don't expect the level of effort to change.
(12:13:30) joe: a recurring theme is that k3d does super complex things, but very often misses the really simple things
(12:13:43) joe: for example, we have crazy "painters"
(12:13:52) joaduo: :)
(12:14:06) joe: a great archetechture really, however we don't have a simple way to toggle a viewport from wireframe to shaded mode!
(12:14:13) joaduo: i have a suggestion for making k3d more popular
(12:14:23) joe: That is a core function that a regular average Joe artists can't really live without
(12:14:35) joaduo: i agree
(12:14:46) joe: so when people ask me "when will it be ready," what the are really asking is, "when will the basics work"
(12:14:54) Bart: joe, what should that look like? I'm willing to do some UI work for things like that
(12:15:15) joe: also, as soon as we have the basics it will be easy for me to build a big community
(12:15:20) joaduo: can a i write my opinion
(12:15:22) joaduo: ?
(12:15:24) joe: sorry bart, hang on
(12:15:38) joaduo: joe i right
(12:15:50) joaduo: i already sent a mail about making first a modeler from k3d
(12:16:04) joaduo: i gues k3d is nearer from a modeler than an animation software
(12:16:09) joe: I have connections with the school board here in BC, and I can get a lot of schools using the software
(12:16:25) joe: joadua, I agree 100%
(12:16:40) joaduo: i we strengthen k3d on modeling we can get people on using k3d
(12:16:53) joaduo: so all we need is to complete the basics on modeling
(12:16:53) joe: I strongly think we should finish the poly modeling toolkit, make k3d obviously better than wings3d
(12:16:59) joe: we'll have a big influx of users right there
(12:17:15) joe: and with users comes brand recognition and awareness.
(12:17:42) joaduo: well... thats the idea
(12:17:53) joaduo: :)
(12:17:56) joe: also, we can't compete with Blender on features
(12:18:06) joe: we have to seperate ourselves for marketing
(12:18:20) joaduo: so, tim, bart? what do you think?
(12:18:22) joe: we need to say "look how clean and easy our software is"
(12:18:36) joe: "look how intuitive the basics are"
(12:18:52) Bart: I agree
(12:19:01) Tim Shead: I don't disagree with any particular point.  This kind of "refocus" is why I called the meeting.
(12:19:13) Bart: I'd like a prioritized list of what to do first, though :)
(12:19:17) joe: most every professional artist I know that has tried Blender, throws there hands up in the air and says "*$%& this! This interface is nothing like it should be"
(12:19:31) joaduo: haaa :)
(12:19:34) joe: but everytime I show people k3d, they are thoroughly impressed
(12:19:43) joe: that's a really accomplishment by you guys!
(12:20:04) joe: big time professional artists immediately take k3d way more seriously than Blender
(12:20:14) joe: because its obvious how much better thought out it is
(12:20:14) joaduo: i have another suggestion for the priority list
(12:20:28) joe: btw, I'm talking about Electronic Arts here as a prime example
(12:20:36) joaduo: i believe we could have two "roadmaps"
(12:21:17) joaduo: one looking on concrete short theme, and end user oriented (strengthing on modeling for example)
(12:21:23) joaduo: and another on the far future
(12:21:29) joe: good plan
(12:21:40) joaduo: so that developers dont loose motivation
(12:22:00) Tim Shead: Gonna break in again here ...
(12:22:33) joaduo: i read
(12:22:53) joaduo: "one looking on concrete short term
(12:22:54) joaduo: "
(12:23:21) Tim Shead: I would put more focus on short-term projects that developers will find interesting.
(12:23:36) Tim Shead: We know from experience that that's the stuff that gets done.
(12:23:47) Tim Shead: Shutting-up now.
(12:23:59) joe: one example was how awesome Romain did in hammering out those transforming tools
(12:24:10) joe: that was a real turning point in usability for the software
(12:24:22) joe: and it only took a relatively short time compared to bigger features
(12:24:46) Bart: ok, so we need guidance on what the artists want most at this point
(12:24:55) joe: I actually think one of our biggest priorities with modeling is just adding a few more things with those transform tools
(12:25:10) joe: I've gone out and done a lot of user feedback sessions
(12:25:18) joe: I've posted the results on the wiki a few times
(12:25:45) joe: I can do more, I can even do them with industry pros, and then complete Noobs
(12:25:47) joe: whatever
(12:26:02) joe: the results are often the same actually. I get very consistent feedback
(12:26:25) joe: "wow! I love the feel of this program! How do I move an objects pivot point?"
(12:26:48) joe: so I could put most common things up there
(12:27:01) joe: I could also bring in a few other people to discuss things
(12:27:27) joe: Also, for testing, I have many students who like to try things out, and they make great testers too.
(12:27:34) Tim Shead: We're coming-up on the end of the meeting, gang.
(12:27:40) joaduo: ok, the question is: can we focus on modeling tools rather than on animation till we have enough users? then hoping to have more manpower?
(12:27:55) joe: I would really hope so
(12:28:02) Tim Shead: Sounds fine to me.
(12:28:40) joaduo: bart?
(12:29:07) joe: Tim, if you send me wiki links, I would love to pop on there and edit some things/do whatever is useful for you guys.
(12:29:12) joe: Also, one last point
(12:29:26) joe: in terms of me actually helping to develop k3d
(12:29:35) Bart: ok with me
(12:29:54) joaduo: ok, we should start a topic on this on the mailing list
(12:30:12) Tim Shead: I'm going to claim the closing statement, just waiting on Joe.
(12:30:26) Bart: The animation stuff was about to get more complex and time-consuming. I wasn't sure if that was the most urgent thing to do right now, anyhow
(12:30:27) joaduo: ok, reading
(12:30:35) joe: would it be possibly to make it so that scripts can register menu entries? like having a startup scripts folder where the startup scripts make new entries appear in the menu that run certain script functions
(12:30:48) Tim Shead: Yes.
(12:30:49) joe: because then I could directly contribute to stuff in the UI
(12:30:52) Bart: joe, I have that working here now!
(12:31:00) joe: right now, I have a real hard time getting anything into the ui
(12:31:12) joe: a simple menu entry would do wonders for me!
(12:31:18) joe: I'm actually pretty decent in python
(12:31:24) joaduo: cool
(12:31:30) joe: its just my c++ that sucs
(12:31:33) joe: sucks
(12:31:47) Bart: Joe, Tim and I talked about writing an OGRE importer, and it was decided we would go with a script. As part of that, I now have code that reads scripts from a dir and adds them to a menu, just like the create node menu
(12:32:01) joe: awesome!
(12:32:12) joe: thanks bart!
(12:32:18) Bart: All I need to do is make an example script, and I can commit this
(12:32:23) joe: at some point I must get trained on how to use it!
(12:33:07) joe: also, its great that way because we can bind menus to hotkeys, so its easy to bind our scripts to hotkeys that way!
(12:33:09) joaduo: ok...
(12:33:16) joe: anyway, I'm done
(12:33:20) joe: thanks for hearing me out!
(12:33:22) joaduo: me too
(12:33:37) joaduo:  GSoC 2008 ideas?
(12:33:44) Tim Shead: OK, I'm going to wind things up.  If you refresh your copy of the meeting page, you will see tha
(12:33:47) joaduo: or thats for the next meeting?
(12:34:01) Tim Shead: Next meeting. 
(12:34:09) Tim Shead: There was never any chance of getting to all those items.
(12:34:19) joe: can you send an invite on the mailing list Tim?
(12:34:22) Bart: :)
(12:34:35) Tim Shead: You mean on an ongoing basis?
(12:34:54) Tim Shead: If you refresh your copy of the meeting page, you will see that I've added all of the action items.
(12:34:56) joe: yes, just whenever there's a new meeting scheduled, just email it out there so we know!
(12:35:17) joe: because I'll forget to check that page, but I will check my email!
(12:35:19) Tim Shead: Well, the next meeting is same time next week, right?
(12:35:27) joe: ok, great
(12:35:39) Tim Shead: There are also links to the two new wiki pages that we will be working on.
(12:35:48) joe: awesome
(12:35:57) joaduo: we could use rss from the wiki
(12:36:01) Tim Shead: So we have only begun to scratch the surface of item 5
(12:36:04) joe: paste a link to the page with the links?
(12:36:04) joaduo: http://www.k-3d.org/wiki/Developer_Meetings
(12:36:08) joe: sweet
(12:36:32) joaduo: i mean... modify that page?
(12:36:36) joaduo: o directly a mail
(12:36:38) joaduo: ?
(12:36:48) Tim Shead: I don't understand the question.
(12:37:30) Tim Shead: OK, I'm going to make a closing point.
(12:37:43) joaduo: ok
(12:37:47) joaduo: i post a mail
(12:37:48) joaduo: :D
(12:37:53) joaduo: sorry
(12:37:59) Tim Shead: We've discussed altering our priorities on what the developers should be working on.
(12:38:10) Tim Shead: That's cool, I think it will be a big help.
(12:38:14) joaduo: my nephew
(12:38:32) Tim Shead: However, I'm going to point out that talking about priorities is something that we've done many times.
(12:39:10) Tim Shead: So development prioritization, in-and-of-itself, is not going to get the kind of radical improvement in community that I'm looking for.
(12:39:34) Tim Shead: I really want us to start thinking outside the box on how to energize the community.
(12:39:58) Tim Shead: So next week, I will be pushing everyone for fresh ideas on how to get going.
(12:40:14) Tim Shead: We need to break out of the current rut.
(12:40:46) Tim Shead: If there's one thing I took away from the SoC summit, it's that we're doing a tiny fraction of the things we could be doing.
(12:41:11) Tim Shead: Cool?
(12:41:18) joaduo: ok!
(12:41:21) Bart: yep!
(12:42:08) Tim Shead: We shouldn't be putting the responsibility for making contacts all on Joe's shoulders ...
(12:42:19) joe left the room (Replaced by new connection).
(12:42:36) joe [joe@celestinestudios.com/Home] entered the room.
(12:43:17) Bart: agreed
(12:43:36) joe: also, I'm only in one region, since we are all over the place, we can cover a lot of ground 
(12:43:52) Tim Shead: I'm pretty darned uncomfortable going-out and making contacts, but I'm going to be pushing all of us to "press-the-flesh".
(12:44:24) Tim Shead: Great!
(12:44:31) Tim Shead: So let's call it a meeting.
(12:44:42) joe: cool. Thanks guys, that was fun!
(12:44:43) Tim Shead: Same time next week ...
(12:44:47) Tim Shead: Cheers,
Personal tools