Die-section for Dédalo

Markus Möller

Hi all,

At the ECFN in London, @Alex, @David, @rca, Karsten and I had already briefly discussed the need to be able to document die combinations in Dédalo. I mean, Rahel in particular had some more concrete ideas about how something like that could be implemented. Currently, in our project on Celtic coins, we now have a whole series of die-data that we’d like to document.
My dissertation is also nearing completion, and by now I have so much die-data that I need a way to document it as well. Since the data is supposed to go into Dédalo anyway, it would be double the work if I had to set up an MS Access database or something similar at the same time.

I would be happy if we could discuss what an implementation might look like that would benefit as many numismatics users as possible.

I think it’s a “no-brainer,” as David would say (;, to be able to link each coin to an obverse and a reverse die. The question is where to store the obverse and reverse dies. Theoretically, you could also create them under “types,” but Karsten, in particular, has already told me that we’d then have “die-types” and that the term just causes confusion, especially during automated processing.

What do we need:

A definition of a class for dies.
A connection from each coin-side to a possible die. … this implies that we should be able on the coin level to indicate the obverse and reverse side. How to handle situations where we can not indicate this?
The possibility to add some uncertainty to a connection of 2.
(optional for now) There can be a recut or die crack. Could be “variations” of a die?

There are now also several databases designed specifically for stamps. Avoiding parallel development or design is another aspect we need to consider. https://numismatics.org/rrdp for the Roman Republican Die Project

rca

Markus Möller
Dear Markus, dear all @Alex , @David , @Manuel Gozalbes
for me, the simplest way still is to integrate the dies in the type section. I have separated obverse and reverse types below the "generic coin type" in any case, and then I can attribute these dies to a specific type: Die - Obverse/Reverse type - coin type (plus in my Haldenstein/our med/mod case "Sorte" = denomination/year und "Art" = denomination/time frame)
And in my coin I can link to a specific obverse or reverse type.
What is still missing are the corresponding labels in the type section - correct?
I gladly discuss this topic in a zoom call. And or - as we all will be in Bari - we put it on the agenda for our WG DEDALO meeting!
Best wishes,
Rahel

Alex

Markus Möller I would be happy if we could discuss what an implementation might look like that would benefit as many numismatics users as possible.

rca I gladly discuss this topic in a zoom call. And or - as we all will be in Bari - we put it on the agenda for our WG DEDALO meeting!

Yes! an important topic that we need to put in common...

Best

rca

we talked about it in Frankfurt already and did all agree - as far as I can remember. So final discussion/decission in Bari and definitely implement it?

Markus Möller

Thank you very much for the quick responses.

I would definitely like it if we could discuss this in Bari. I’ll be there on Tuesday for the working groups as well. I assume there’s a Dédalo session planned as part of Nomisma? I haven’t seen which groups are scheduled—I only have the general preliminary program for Bari.

My main goal in bringing this topic up again was so that we could give it some thought beforehand and then make a final decision in Bari about what we really need.

I think, from a conceptual standpoint, Karsten’s perspective is particularly important. If we need to have another discussion via Zoom beforehand, he should be there to share his input.

For me, regardless of the exact implementation, the most important thing is that dies can be specifically assigned to individual coins, since I usually have multiple dies per type. If we have dies in Types, then theoretically I should also be able to assign multiple dies to a single type, which in turn can be assigned to multiple coins, right?

@rca If I understand this correctly, you're treating the dies as a very detailed type—in other words, a sub-subtype? So, at the very bottom of the hierarchy? And then you'd just need to be able to label them as "die"? In that case, that would be the point where Karsten sees a problem—one that might have already been solved by labeling. He'll have to say so then.

best wishes

Markus

Manuel Gozalbes

Hello,
This is a very important issue, and we have gone back and forth on how to solve it without ever reaching a clear conclusion. I am sure the best approach would be to hold a meeting and make some decisions on the matter.
What happened or rather, what is currently happening is that we already have a Type section which could, in principle, be used in some way as an indication or reference in relation to dies. However, the Type section should really represent or function as a combination of dies, rather than as a die record itself. A die is, after all, an individual object and not the combination of dies. Because of this, it may be necessary to develop a new Dies section where the entire framework of relationships can be built, including links between dies and their connections to coins and types.
What I am certain about is that this is an important topic, and I think Rahel comments also makes sense: creating a typology that would allow the Type section to be used in whatever way is required. Then, if necessary, a separate section could be used to work directly with dies.
That said, I also remember that Karsten always point out to an interesting aspect of the problem: what exactly is a type? A type can simply be understood as a particular combination of dies—in other words, a pair of dies can itself define a type. This suggests that there is an intermediate level of definition that needs to be carefully considered.

Markus Möller

Manuel Gozalbes

I’m also not entirely sure whether it might not be better to have a separate “die” section—if only for analysis purposes. But it also depends a bit on what the majority of users actually need. For example, since we also have die chains that we can track, we could use the option to specify an order. That way, when assigning coins, we can indicate which ones appear earlier or later in the chain. But I can also imagine that this might be a little too specific.

rca

... yes, I forgot Karsten Tolle's input that a die is an object, wether preserved or not. So the dies are TCH - Tangible Cultural Heritage Objects (section tch1) that can then be linked to the typology section (numisdata3), an obverse or a reverse type,
@Manuel Gozalbes - in many medieval/modern and also celtic topics, we split the obverse & reverse types, and one specific coin can be described as a specific combination of an obverse and a reverse type. The same is true for Roman types - bust types on the obverse, "picture types" on the reverse. And in a die study I have to be able to link every coin to a specific obverse & reverse die. It may always be the same "type" in your definition.

Die = tch1, with description
Type = numisdata3 with description (obverse or reverse) of the die as a type (or - in the hierarchical order - a sub-sub-sub-sub-type of the obverse or the reverse; linking the tch1 object to the type
Type = numismata3 on a "higher hierarchical level" - either first obverse/reverse splitted, or directly combining both (as @Manuel Gozalbes will do).
Linking from the numismatic objects (numisdata4) to either a type for "a whole coin/combination of obverse & reverse" or to a obverse type and / or a reverse type separately, beeing able to specify/label what I chose.

rca

@Markus Möller - DIE CHAIN: not too specific at all, one of the very basic use cases! fresh die - broken a little - more broken - "fixed" - broken again ....