<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">John: Yes. If you can design a system that is quickly expandable and all you have to do is pop in some properties, I think the majority of tweening is done on known prop names. <div><br></div><div>For special cases there could be a few generic classes that tween 'a number on an object', 'an object in an object' and 'an array in an object'. But I agree with you that if you have a solid set of typed ones those are the ones you will grab for first, and save the generic ones for special situations where you want to tween some random variable once and it doesn't require creating a reusable class. (Of course for those situations you can also just use a LinearGo with an update callback so in that way, the generic tweens are barely needed at all once you have a strong set of typed tweens.)</div><div><br></div><div>m<br><div><br></div><div><br><div><div>On May 7, 2008, at 8:07 AM, John Grden wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite">The thing about strict / strong typing is that you gain performance benefits, coding benefits, compiling security as well as control in a development environment where you have several devs working on the same project.<br> <br>Now, in the api I'm developing, I'm starting to see how easy it would be to abstract the properties class so that if you knew what properties you would ever want to manipulate, you could create your own properties class. For the most part, if we really think about it, you can create classes ( or maybe just one big one ) for DisplayObjects to start with. If you used Fuse like I did for sequencing events more than actual tweening, you can certainly create your own properties class for those as well. I mean, it makes too much sense especially in light of what you gain in terms of what I mentioned above.<br> <br>It doesn't cut you off from doing anything at all, and you gain quite a bit. The cost is adding 1 properties class 1 time.<br><br>Thoughts?<br><br><div class="gmail_quote">On Wed, May 7, 2008 at 12:44 AM, Sebastian Weyrauch <<a href="mailto:go@tweego.org">go@tweego.org</a>> wrote:<br> <blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> <div link="blue" vlink="purple" style="" lang="DE"> <div><p><span style="font-size: 11pt; color: rgb(31, 73, 125);">Hi,</span></p><p><span style="font-size: 11pt; color: rgb(31, 73, 125);">think this method of John is really cool too. I hate string properties too. But as you mentioned: For universal handling this is no real possibility :( If the advantage is much bigger than the disadvantage, you can give up strict typing, at least in my eyes.</span></p><div><span style="font-size: 11pt; color: rgb(31, 73, 125);"> </span><br class="webkit-block-placeholder"></div><p><span style="font-size: 11pt; color: rgb(31, 73, 125);">Maybe you're right. A universal tween class can not be that good as specialized classes. But I think it can be really hard to know and use several tween classes. And all popular tween classes are universal tween classes. But perhaps this will change in future. We will see.</span></p><div><span style="font-size: 11pt; color: rgb(31, 73, 125);"> </span><br class="webkit-block-placeholder"></div><p><span style="font-size: 11pt; color: rgb(31, 73, 125);">best regards</span></p><p><span style="font-size: 11pt; color: rgb(31, 73, 125);">sebastian</span></p><div><span style="font-size: 11pt; color: rgb(31, 73, 125);"> </span><br class="webkit-block-placeholder"></div> <div> <div style="border-style: solid none none; border-color: rgb(181, 196, 223) -moz-use-text-color -moz-use-text-color; border-width: 1pt medium medium; padding: 3pt 0cm 0cm;"><p><b><span style="font-size: 10pt;">Von:</span></b><span style="font-size: 10pt;"> <a href="mailto:golist-bounces@goasap.org" target="_blank">golist-bounces@goasap.org</a> [mailto:<a href="mailto:golist-bounces@goasap.org" target="_blank">golist-bounces@goasap.org</a>] <b>Im Auftrag von </b>Moses Gunesch<br> <b>Gesendet:</b> Mittwoch, 7. Mai 2008 00:06<br> <b>An:</b> Mailing list for the Go ActionScript Animation Platform<br> <b>Betreff:</b> Re: [Golist] This looks a bit better as far as syntax</span></p> </div> </div><div><div></div><div class="Wj3C7c"><div> <br class="webkit-block-placeholder"></div><p>Hehe, sorry. And they let me write for a book! ;)</p> <div><p>Let me try again....</p> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p><span><u>Turn ons:</u></span> What you're doing is better for strict typing because the properties are real properties – that is cool. I like how you have x() etc., that's really cool. In mine you had to do properties as strings like "x".</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p><span><u>Turn offs:</u></span><span> I'm starting to lean against multi-property tween classes in general. Because, it's easy to forget how complex just ONE tween can be. Like, think of a color tween or a filter tween – multiple properties in arrays, nested in subproperties of display objects... gets hairy. Start values and relative values are another can of worms.</span></p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>Lately I've been thinking that<i> </i>a tween instance is more like an atom than a molecule. That's why I'm now gravitating back toward individual-property tween classes. That, and OverlapMonitor would be able to handle every property individually. </p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>Have not built this out though.... </p> </div> <div><p>But I'm wondering if there would some way to do that cool property syntax thing you're doing, but at the PlayableGroup level?</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>- m</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><div> <br class="webkit-block-placeholder"></div> <div> <div><p>On May 6, 2008, at 4:38 PM, John Grden wrote:</p> </div><p><br> <br> </p><p style="margin-bottom: 12pt;">So I'm confused a bit -seems like you said you were leaning towards 1 property per tween that exists, but you like what I'm doing with that last code sample<br> <br> LOL I know I missed something<br> <br> I think that I started to get what you meant though - having those apart of a PlayableGroup etc is essentially what we end up with, is that about right?</p> <div><p>On Tue, May 6, 2008 at 3:17 PM, Moses Gunesch <<a href="mailto:moses@goasap.org" target="_blank">moses@goasap.org</a>> wrote:</p> <div><p>Yeah that's good. I created a system called OpenTween – don't think I ever posted it though because I wasn't sure I liked the abstraction layer of putting a ton of information and functionality into property classes which is what happened when I tried this.</p> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>My property inputs were</p> </div> <div> <div><p> propName: <span style="color: rgb(55, 55, 55);">String</span>, </p> </div> <div><p> endVal: *=<span style="color: rgb(127, 0, 85);">null</span>, </p> </div> <div><p> endValRelative: <span style="color: rgb(55, 55, 55);">Boolean</span>=<span style="color: rgb(127, 0, 85);">false</span>, </p> </div> <div><p> startVal: *=<span style="color: rgb(127, 0, 85);">null</span>, </p> </div> <div><p> startValRelative: <span style="color: rgb(55, 55, 55);">Boolean</span>=<span style="color: rgb(127, 0, 85);">false</span></p> </div> </div> <div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>Then I had this functionality in there from Fuse, that lets you omit either start or end and it will figure it out. That's probably why it got complicated. (I will share OpenTween if you want.)</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>Anyway yeah I think there is something to a multi-property approach for sure, but I've been leaning back toward just extending LinearGo in the simplest way possible – one property per tween class. </p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p><b>Here is why:</b> if every single property were in a separate tween class, OverlapMonitor works the best in that it can allow just one property to be paused, stopped and so on. That's how ZigoEngine worked, it atomized every tween internally. I've been thinking PlayableGroup could be extended to handle multi-prop tweens by creating separate tween instances, then it could let you drill into them to grab children by property etc. </p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>One could argue that it's slower to not block all similar tweens, but it also provides more control and is less abstract, and I've found that in real practice there are usually only a very small number of "blocked" tweens at once... pretty soon you want to change delay/duration/easing on one property or another and you end up with a group anyway.</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>That said, I'd love to see your concept taken all the way John...!</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> <div><p>:-)</p> </div> <div><div> <br class="webkit-block-placeholder"></div> </div> </div> <div><div> <br class="webkit-block-placeholder"></div> <div> <div> <div> <div><p>On May 6, 2008, at 3:54 PM, John Grden wrote:</p> </div><div> <br class="webkit-block-placeholder"></div> </div> </div> <blockquote style="margin-top: 5pt; margin-bottom: 5pt;"> <div> <div><p><b>Was this:</b><br> tween_0 = new Tween3D(target, 1, Equations.easeOutCubic);<br> tween_0.x = 0;<br> tween_0.y = 50;<br> tween_0.rotationZ = 0;<br> sequence.addStep(tween_0);<br> <br> sequence.lastStep.advance = new OnDurationComplete(.2); // advance early/overlap<br> tween_0b = new Tween3D(target, 1, Equations.easeOutCubic);<br> tween_0b.z = 200;<br> sequence.addStep(tween_0b, true); // 2nd param groups it with previous step. param is "addToLastStep"<br> <br> tween_1 = new Tween3D(target, 1, Equations.easeOutCubic);<br> tween_1.x = -10;<br> tween_1.y = 85;<br> tween_1.rotationZ = 15;<br> sequence.addStep(tween_1);<br> sequence.lastStep.advance = new OnDurationComplete(.25); // advance early/overlap<br> <br> tween_2 = new Tween3D(target, 1, Equations.easeOutBounce);<br> tween_2.rotationX = 0;<br> tween_2.rotationY = 0;<br> sequence.addStep(tween_2);<br> <br> <b>Is now this:</b><br> <br> tween_0 = new Tween3D(target, [Go3D.x(0), Go3D.y(50), Go3D.rotationZ(0)], 1, Equations.easeOutCubic);<br> sequence.addStep(tween_0);<br> sequence.lastStep.advance = new OnDurationComplete(.2); // advance early/overlap<br> <br> tween_0b = new Tween3D(target, [Go3D.z(200)], 1, Equations.easeOutCubic);<br> sequence.addStep(tween_0b, true); // 2nd param groups it with previous step. param is "addToLastStep"<br> <br> tween_1 = new Tween3D(target, [Go3D.x(-10), Go3D.y(85), Go3D.rotationZ(15)], 1, Equations.easeOutCubic);<br> sequence.addStep(tween_1);<br> sequence.lastStep.advance = new OnDurationComplete(.25); // advance early/overlap<br> <br> tween_2 = new Tween3D(target, [Go3D.rotationX(0), Go3D.rotationY(0)], 1, Equations.easeOutBounce);<br> sequence.addStep(tween_2);<br clear="all"> <br> <br> I'm still thinking about this approach, but thought I would throw it out to you guys to see what you thought. Right now, there's static methods in Go3D that return a Go3Dproperty. Tween3D has an array called propertyChanges and if there is an array in the propertyChanges argument, I just set it straight away - no parsing required. It's all ready to go and is filled with Go3DProperty objects.<br> <br> Thoughts?<br> -- </p> </div> </div><p>[ JPG ] _______________________________________________<br> GoList mailing list<br> <a href="mailto:GoList@goasap.org" target="_blank">GoList@goasap.org</a><br> <a href="http://goasap.org/mailman/listinfo/golist_goasap.org" target="_blank">http://goasap.org/mailman/listinfo/golist_goasap.org</a></p> </blockquote> </div><div> <br class="webkit-block-placeholder"></div> </div> </div><p style="margin-bottom: 12pt;"><br> _______________________________________________<br> GoList mailing list<br> <a href="mailto:GoList@goasap.org" target="_blank">GoList@goasap.org</a><br> <a href="http://goasap.org/mailman/listinfo/golist_goasap.org" target="_blank">http://goasap.org/mailman/listinfo/golist_goasap.org</a></p> </div><p><br> <br clear="all"> <br> -- <br> [ JPG ] _______________________________________________<br> GoList mailing list<br> <a href="mailto:GoList@goasap.org" target="_blank">GoList@goasap.org</a><br> <a href="http://goasap.org/mailman/listinfo/golist_goasap.org" target="_blank">http://goasap.org/mailman/listinfo/golist_goasap.org</a></p> </div><div> <br class="webkit-block-placeholder"></div> </div> </div> </div></div></div> </div> <br>_______________________________________________<br> GoList mailing list<br> <a href="mailto:GoList@goasap.org">GoList@goasap.org</a><br> <a href="http://goasap.org/mailman/listinfo/golist_goasap.org" target="_blank">http://goasap.org/mailman/listinfo/golist_goasap.org</a><br> <br></blockquote></div><br><br clear="all"><br>-- <br>[ JPG ] _______________________________________________<br>GoList mailing list<br><a href="mailto:GoList@goasap.org">GoList@goasap.org</a><br><a href="http://goasap.org/mailman/listinfo/golist_goasap.org">http://goasap.org/mailman/listinfo/golist_goasap.org</a><br></blockquote></div><br></div></div></body></html>