Allow upto 512 images on a WPS

[knight4led]

I believe using a : (colon) character as a seperator will eventually lead to the need to have %: in the rest of the WPS because for now it is used with the RTC on many targets without the leading %. I don't know if this will confuse the WPS parser, but it may confuse WPS writers.

[Lear]

It is only used as a separator when parsing a small number of tags (one so far), so the it wouldn't confuse the parser at least. Using ':' in combination with the RTC tags is not a problem.

Adding '%:' would only be needed if there was a need to enter ':' for other uses than a separator within a tag using ':' as a separator. In this case (as well as for the other image-related tags) it is used to separate numbers and filenames, so it isn't needed yet.

[fml2]

I'd put much brain energy in a solution that would work without a new special symbol. Special symbols should be kept to a minimum IMHO. Why can't we use something like %evA (as proposed by lowlight)? This command would also need the grid size specified somehow (can '|' be used for this?)

[Lear]

Why can't we use something like %evA (as proposed by lowlight)? This command would also need the grid size specified somehow (can '|' be used for this?)

%evA would not need a grid size, since that is implied from the BMP dimensions and number of entries, as specified when loading it. (But if a grid size - or more likely, image number, when used in non-enumeration cases - were to be specified, '|' could not be used.)

[kahuna_999]

the %evA solution proposed by lowlight only addresses one part of this patch.
I agree it would be a more elegant solution to enumerated tags such as volume and battery levels.

My patch allows you to use any part of a bitmap anywhere on the screen, so the need for a new seperator is still valid.

[fml2]

My patch allows you to use any part of a bitmap anywhere on the screen, so the need for a new seperator is still valid.

But can't this be done with regular image tags? I.e. the new feature is mostly interesting for enumerations, for other cases we already have a solution. Or does anybody need more than 52 regular images?

[Llorean]

The more image files you have in a WPS, the longer it takes to load at boot. A single image, where only the parts are used in multiple places, would probably result in a much faster loading WPS than one that contains 52 images.

[fml2]

The more image files you have in a WPS, the longer it takes to load at boot. A single image, where only the parts are used in multiple places, would probably result in a much faster loading WPS than one that contains 52 images.

True. But how many 'static' images are there? I'm not a WPS designer, but I think that there are many images because every single image in every enumeration has to be declared. If that's done with an %evA tag the image number will be significantly cut.

[Llorean]

Even then, if you reduce it to 22 images, that's still more than the possibility of a single image plus backdrop. I don't see why the possibility for a faster load time is a bad thing, even if it requires a new separator. Is it just fear of confusion that makes the separator bad?

[fml2]

Is it just fear of confusion that makes the separator bad?

It means more complicated syntax (to learn), more complicated parser code and, probably, more tags needed (%:). Faster WPS loading at no other cost is fine but there is the said penalty.

And IMHO images for a enumeration can be thought of as a package and hence get a dedicated tag. Whereas using a package for 'standalone' images, while improving loading speed, would make WPS less readable (for a designer).

Are there great speed related problems? My WPS is text only so I may have missed something.

[Llorean]

As was said above, %: should not be needed. If someone can't learn the additional syntax for the one tag, then they can avoid using that tag.

As for designer readability: In the end, should the focus be on "make the WPS easier for designers to make" or "reduce the cost of using pretty WPSes to the end user"?

A complex WPS with many images can add a measurable time to the boot. I don't know the maximum time it can add, but the difference can be noticed by a casual user.

[fml2]

As for designer readability: In the end, should the focus be on "make the WPS easier for designers to make" or "reduce the cost of using pretty WPSes to the end user"?

IMHO we should consider both. An extremely bad state on either side should be avoided.

But, as I already said, I don't use a WPS stuffed with images so my opinion is only based on purity and aesthetics.

[knight4led]

I was not opposed to using : as a seperator, nor the addition of a %: tag down the road should it be necessary. I was simply stating if the first option is implemented, the second may be either a necessary or a good idea.

Personally I think that if it is decided that this new seperator is good for this tag, maybe it should be used for all enumeration tags that are non-conditional.

Progress in the form of faster loading and more versatility to the WPS parser is, in my opinion, better than trying to force constants based on previous versions of Rockbox.

[Yotto]

For the record, I would welcome complexity in WPS code to be able to make nicer WPS's. I've thought of another cool use for this, a smooth circular volume "bar" around... anything.  A play button, album art, whatever.

[lowlight]

FYI...my patch for enumerated image sets is here: http://www.rockbox.org/tracker/task/7345
If you try it, please, post your comments there.