Talk:Death Star: Difference between revisions
Content added Content deleted
(→Perl 6) |
(→Perl 6: 'twas me) |
||
| Line 5: | Line 5: | ||
:::Interestingly, this other buglet I've been fixing distorts the ellipse on the other side, by making the outline of the entire figure always a circle, when it needs to show background for rays where both points of the positive sphere are eaten by the negative sphere. Most of the chosen spheres don't do this, but if you increase the size of the "small" sphere to be equal or larger than the "large" sphere, you'll see the intersection of the two spheres is far from an ellipse in programs with this bug. So while it's not important for this task, it would be better if the algorithms were more correct in case someone wants to use them elsewhere. (It would probably also be better for that reason to name them "negative" and "positive" spheres in the solutions, since in general the negative sphere doesn't have to be smaller than the positive sphere.) --[[User:TimToady|TimToady]] 02:31, 21 August 2011 (UTC) |
:::Interestingly, this other buglet I've been fixing distorts the ellipse on the other side, by making the outline of the entire figure always a circle, when it needs to show background for rays where both points of the positive sphere are eaten by the negative sphere. Most of the chosen spheres don't do this, but if you increase the size of the "small" sphere to be equal or larger than the "large" sphere, you'll see the intersection of the two spheres is far from an ellipse in programs with this bug. So while it's not important for this task, it would be better if the algorithms were more correct in case someone wants to use them elsewhere. (It would probably also be better for that reason to name them "negative" and "positive" spheres in the solutions, since in general the negative sphere doesn't have to be smaller than the positive sphere.) --[[User:TimToady|TimToady]] 02:31, 21 August 2011 (UTC) |
||
:::: <s>Originally I wrote the C code as a quick hack, mostly for laughs. I noticed it may have this bug, and later fixed it in the perl code. I probably should check the C example again.</s> Apparently I already fixed it at some point, huh. --[[User:Ledrug|Ledrug]] 03:06, 21 August 2011 (UTC) |
:::: <s>Originally I wrote the C code as a quick hack, mostly for laughs. I noticed it may have this bug, and later fixed it in the perl code. I probably should check the C example again.</s> Apparently I already fixed it at some point, huh. --[[User:Ledrug|Ledrug]] 03:06, 21 August 2011 (UTC) |
||
:::::Er, actually, you're not going insane; I just fixed it several hours at [http://rosettacode.org/mw/index.php?title=Death_Star&curid=9404&diff=118150&oldid=118135]. |
|||
Revision as of 06:14, 21 August 2011
Perl 6
The perl6 output doesn't look quite right. --Ledrug 03:25, 20 August 2011 (UTC)
- Agreed. I will look at it. --TimToady 14:31, 20 August 2011 (UTC)
- Fixed, thanks. --TimToady 16:41, 20 August 2011 (UTC)
- Interestingly, this other buglet I've been fixing distorts the ellipse on the other side, by making the outline of the entire figure always a circle, when it needs to show background for rays where both points of the positive sphere are eaten by the negative sphere. Most of the chosen spheres don't do this, but if you increase the size of the "small" sphere to be equal or larger than the "large" sphere, you'll see the intersection of the two spheres is far from an ellipse in programs with this bug. So while it's not important for this task, it would be better if the algorithms were more correct in case someone wants to use them elsewhere. (It would probably also be better for that reason to name them "negative" and "positive" spheres in the solutions, since in general the negative sphere doesn't have to be smaller than the positive sphere.) --TimToady 02:31, 21 August 2011 (UTC)
Originally I wrote the C code as a quick hack, mostly for laughs. I noticed it may have this bug, and later fixed it in the perl code. I probably should check the C example again.Apparently I already fixed it at some point, huh. --Ledrug 03:06, 21 August 2011 (UTC)- Er, actually, you're not going insane; I just fixed it several hours at [1].
- Interestingly, this other buglet I've been fixing distorts the ellipse on the other side, by making the outline of the entire figure always a circle, when it needs to show background for rays where both points of the positive sphere are eaten by the negative sphere. Most of the chosen spheres don't do this, but if you increase the size of the "small" sphere to be equal or larger than the "large" sphere, you'll see the intersection of the two spheres is far from an ellipse in programs with this bug. So while it's not important for this task, it would be better if the algorithms were more correct in case someone wants to use them elsewhere. (It would probably also be better for that reason to name them "negative" and "positive" spheres in the solutions, since in general the negative sphere doesn't have to be smaller than the positive sphere.) --TimToady 02:31, 21 August 2011 (UTC)
- Fixed, thanks. --TimToady 16:41, 20 August 2011 (UTC)