[codenameone-discussions] Re: an IOS platform issue with images

2016-06-21 Thread Shai Almog
Pull requests are reasonably easy to do. The main motivation of using that over an issue or comment is: a. They get done. b. You get the credit. I would not use the word unrelated since the values are clearly related. -- You received this message because you are subscribed to the Google Groups

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-21 Thread Dave Dyer
It's unreasonable for me to go to all the effort to resync my sources and construct a pull request, only to be rejected because I can't guess the specific wording you will find acceptable. Something to the effect of Note that RGB values produced by getRGB may not be identical to those you exp

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-20 Thread Shai Almog
If you have something specific in mind you can submit a pull request with the javadoc changes you think are appropriate for this. -- You received this message because you are subscribed to the Google Groups "CodenameOne Discussions" group. To unsubscribe from this group and stop receiving email

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-19 Thread Dave Dyer
well, you could retain the original rgb array that was used to create the image, but that Screams waste of memory. My test indicates this is also a problem for images created by the embedded image readers. Basically this means getRGB is useless for processing images with a transparency componen

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-19 Thread Shai Almog
That's possible. Especially with image manipulation which is pretty tricky to get right on iOS as it's graphics pipeline is totally different. I'm not sure if it's fixable though, if I recall that code we added the color to workaround a different problem in that logic where transparency didn't

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-18 Thread Dave Dyer
The value I put in is 0x, the value I get back out Is 0x00ff but the effect ix not limited to fully transparent pixels. Near!y transparent come back as nearly white. The "white" pollution continues up through the transparency scale, so my smooth gray shadows become tinged with white

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-18 Thread Shai Almog
Where did you say you put in black and got back white in the question? I just re-read it and can't see that neither in the sample nor the wording. If the alpha component is transparent the color component is ignored. -- You received this message because you are subscribed to the Google Groups "

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-18 Thread Dave Dyer
I would say that putting out black and getting back white, as is the case here, is a rather spectacular example of "not guaranteed" On Friday, June 17, 2016 at 11:31:16 PM UTC-7, Shai Almog wrote: > > Assuming out is an ARGB array which you didn't state. This still isn't > guaranteed to work as

[codenameone-discussions] Re: an IOS platform issue with images

2016-06-17 Thread Shai Almog
Assuming out is an ARGB array which you didn't state. This still isn't guaranteed to work as the colors are applied to a native surface which doesn't have to support all the colors. -- You received this message because you are subscribed to the Google Groups "CodenameOne Discussions" group. To