fbungarz wrote:There is probably good reason, why Hert decided to finally abolish non-standardized custom XMP.
Reading through old posts recently, Hert believed that very few people used custom XMP. And by eliminating those features, he can reduce the amount of testing that goes into every new release.
When Adobe introduced XMP as so-called eXtensible Metadata Platform, IDI took the eXtensible quite literally; for years IDI was the only software that allowed users to deign their own custom XMP without major programming skills. I was one of the early adopters, excited that I now could write specimen collection data directly into my photos using my own custom XMP. Perfect ...
I can't be 100% certain, but I believe that iMatch supports/supported some form of custom XMP. I know a group of people who among themselves devised a set of extensions that they used to share photos.
Well, almost. The problem: virtually no other tools will ever read "my" custom XMP. Thus the decision to abandon my own custom XMP in favor of using IDI to "translate" the DarwinCore metadata standard for biological diversity metadata into DarwinCore XMP. And I am very grateful to Hert that with version 3.3. DarwinCore XMP is at last natively supported by PSu. Granted, DarwinCore XMP is still not widely adopted, but at least it is based on a metadata standard now widely employed by Natural History Collections and thus at least compatible with that standard!
So true about lack of third-party support for XMP. I attribute that to poor marketing and "industry evangelism" on the part of Adobe. But it's not only XMP. To a lesser extent, Adobe's DNG isn't supported either. Yes, some guys had an "openraw" idea about 8 years ago, but that idea went basically nowhere, despite the claims on the website.
https://www.openraw.org. Adobe's DNG is the closest thing to an open raw format, but I claim that it has had only limited success since Nikon and Canon have both continued to use their proprietary formats. Again, an evangelism failure.
PSu's own "custom" ICS schema also has at least the one advantage that it is PSu's own quasi-standard. Thus, there is at least a chance that some day other tools will be able to read (and write?) it - unlike IDI's custom XMP it is not entirely arbitrary.
Agree with you. However, that support is limited to only text fields right now. Since I just finished experimenting with custom fields, I have some questions, which I will post in a separate thread.
The once liberating promise of XMP to be indeed eXtensible has unfortunately not delivered.
I would argue that Adobe focused only on the "engineering" aspects of XMP, without considering "deployment and usage" across a range of user types and scenarios. BTW, you see tech companies make this sort of mistake only all the time.
Managing metadata in entirely obscure fields that no tool will ever be able to access defies the purpose of adding metadata to these images. So, anybody is probably well advised to migrate their own IDI custom XMP into PSu's ICS XMP.
I agree again, and that is my plan going forward for my own custom metadata for railroad photography subjects.
All that said, I regret that custom XMP is not generally more widely supported. It is quite ironic that even Lightroom, Adobe's own brainchild, only supports standard XMP...
Yes, because again, Adobe didn't think through the issues.
I don't want to hijack my own thread, but I will say this. Custom XMP isn't just one "problem." It's a range of problems. It's special interests such as yours or museums or art collectors. it's also hobbyists, and it's also people who just want custom fields for their own use and have no interest in sharing that metadata with anyone else. But I don't think that this forum is even the right place to discuss these issues.
Phil
but better late than never. 
I lived through the Penn-Central debacle and then ConRail. I still can't get used to the idea that NS owns most of the old PRR. 