Community
Participate
Working Groups
This is follow up to bug 432741. Its been found that a "list of three" png images "works well" for most known cases, but this deviates from what the Platform has traditionally specified, and it is unclear if and when additional image sizes and image types are needed or used. This deserves more investigation. On the one hand, we don't want to lose functionality in some situation, but on the other, we don't want to do things "like we always have" just for the sake of doing things like like we always have, when using a smaller list of PNG images solves known problems. In particular, see http://www.visualpharm.com/articles/icon_sizes.html and other references listed in https://wiki.eclipse.org/Platform-releng/Updating_Branding In addition, need to understand exact effects/requirements of Tycho builds, and to a lessor extent, PDE builds. (They do, apparently, in some circumstances "replace" the ico image in windows executable, for example, with the icons in "windowImages", presumably.
A few things I've learned, is that the reference for http://www.visualpharm.com/articles/icon_sizes.html is taking about what needs to be in the ICO file, for Windows. Eclipse (SWT) apparently does not use what's in the ICO file, but uses what's in the "windowImages" list. It'd be nice to have a similar reference specifically for Eclipse ... what, when and where does Eclipse/SWT, use the various sizes. (And if order in the windowsList mattters? Also, they don't show a picture of where the 256 pixel image is used but I know from experience Windows 7 uses it in the "status" bar, of Windows explorer. See the attachment in bug 432791 comment 10, where for a while, apparently, we were replacing the smaller images with "Luna" images, but the 256 pixel image was still the one from Kepler. = = = = = Also, I think I'm learning that the "reduced palette" images are only needed for Windows XP. The "visualpharm" reference doesn't say anything about that, but Windows XP docs do, and I'm beginning to wonder if we still need that? (I would say "no", since Windows XP is so old and now "out of service", but is grey area since from news reports, sounds like a lot of people still use it? = = = = = I also don't believe we pay proper attention to "high contrast" settings that I think are used by some visually impaired people. Its true that it will be "automatically applied" to what we have ... but, if you do that automatically, what we have looks extremely bad -- useless, basically (at least, for "logo" windows image -- whereas if "done right" we'd hand craft such images. = = = = = = And, I should add ... we've not done anything for the "Windows 8" image sizes. This might be "ok" for Luna, since we are barely calling it a supported platform, if we do at all ... but, if/when people start to make "RCP apps" for Windows 8, they will want to include those sizes, and the ICO in our executable will be the "wrong shape" to accommodate them. (no biggie, for now, just "dumping notes for the future").
Arun, can you shed any light on "how when and where" SWT uses the different size images in "windowImages" list? (since it apparently does not use the ones from the ICO file embedded in executable). Or, was this something that "changed" in e4?
(In reply to David Williams from comment #2) > Arun, can you shed any light on "how when and where" SWT uses the different > size images in "windowImages" list? (since it apparently does not use the > ones from the ICO file embedded in executable). Or, was this something that > "changed" in e4? I don't know the exact answers to these questions for sure but from whatever I know, Shell#setImages does not choose the images by itself from any particular locations, it relies on the callers of the API to provide the set of images and then tries to choose the best image among them for display based on the platform and the usage. So, we should probably be looking at the platform code that creates and supplies these images to SWT.... I'm adding Silenio to the bug as he will probably be able to throw more light on this and correct me if I'm wrong.
This bug hasn't had any activity in quite some time. Maybe the problem got resolved, was a duplicate of something else, or became less pressing for some reason - or maybe it's still relevant but just hasn't been looked at yet. If you have further information on the current state of the bug, please add it. The information can be, for example, that the problem still occurs, that you still want the feature, that more information is needed, or that the bug is (for whatever reason) no longer relevant. -- The automated Eclipse Genie.