You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm using lwip as part of a project at LinkedIn with @chriseppstein, which involves generating custom PNG images from other images. Since this is computationally expensive, we are currently saving certain information to check if our PNGs need to be regenerated, such as the modification times of the source images. We would ideally like to save this information in the PNG metadata, perhaps in a a tEXt data chunk, and were considering writing this feature for lwip. Is this a feature you would consider accepting, and if so do you have any opinions on how that should be implemented?
The text was updated successfully, but these errors were encountered:
I would start by learning how the PNG format defines storing text data in the file. Then this would have to be supported by the decoder and the encoder.
Additionally, the lwip Image type would have to support storing and accessing this data. I.e., passing initial data in the constructor, and providing a 'getMetadata' and 'setMetadata' methods on the prototype.
When an image is decoded from PNG, its metadata would be passed into the Image constructor. When an image is encoded into PNG, the metadata on the Image object would be passed to the encoder.
For all other formats (JPEG and GIF) it can ignored for now.
I'm using lwip as part of a project at LinkedIn with @chriseppstein, which involves generating custom PNG images from other images. Since this is computationally expensive, we are currently saving certain information to check if our PNGs need to be regenerated, such as the modification times of the source images. We would ideally like to save this information in the PNG metadata, perhaps in a a tEXt data chunk, and were considering writing this feature for lwip. Is this a feature you would consider accepting, and if so do you have any opinions on how that should be implemented?
The text was updated successfully, but these errors were encountered: