Adding more reference lines - attn.: Xavier [solved]

Bonjour Xavier,
I have started to cross reference Hungarian coins and experienced that I am running out of lines. Currrently it is maximum 5 references.
If you look at this, you'll find the references full, and 2 more references are mentioned in the comments, and remain unsearchable:
https://en.numista.com/catalogue/pieces33905.html

Would it be a difficult technically to add more cells?
Just like with the ADD 5 more years button on the bottom of years?

Thanks for your consideration,
Imre

I also face this issue sometimes, increasing to at least 10 would be good.

This is a great suggestion! Especially for ancient coins this would be very useful!
10-12 reference lines would be nice

technically speaking, I am not sure whether there should be a limit.
I guess it can be handled just like the year lines on a coin page

Five reference books looks enough for near all the countries. Do you know any example that needs 6 books?

The example you provide only needs four: EH, H, Dav and KM.
https://en.numista.com/catalogue/pieces33905.html

The "user's guide" of the multi-reference feature says to write one reference number per book in References, and all the numbers in Comments in year lines.

I have a few where Im at the limit with more to add such as here for example https://en.numista.com/catalogue/pieces185330.html we will add more references so such cases are only increasing.

Yes, there are many!

https://en.numista.com/catalogue/pieces188440.html

This is one I happened to add yesterday, it would need 9.

Most Celtic coins will have 6-12. Laufenburg coins can have more than 9, Neuchatel gold coins and bracteates, St Gallen thalers and gold coins and Lausanne also can have more than Some Basel gold coins too I think. Just to name a few that I've encountered.


Also more than one ref. number for a catalog would be good (like for No date lines). And if there are non-consecutive numbers, these could be added on a separate line.

best
strato

On top of that, would be nice to have each reference in a new line, because currently it looks a bit messy:



and last but not least, referees should be able to choose default sorting by reference for their country.
Let's say, there at least 3 different references used for Roman coinage. For consistency, as a referee I'd love to set how to sort coins by references, for example: 1. sort by RIC catalogues; 2. RCV; 3. Cohen and so on. Then the references would be sorted and shown in a uniform way.

Absolutely agree what you guys have just suggested.
Sorting of references is indeed quite a mess, personally I always put KM# first as that is what most people use but what i put after that is often not sorted in a uniform way.

Quote: "stratocaster"​Yes, there are many!
​
​https://en.numista.com/catalogue/pieces188440.html
​
​This is one I happened to add yesterday, it would need 9.
​
​Most Celtic coins will have 6-12. Laufenburg coins can have more than 9, Neuchatel gold coins and bracteates, St Gallen thalers and gold coins and Lausanne also can have more than Some Basel gold coins too I think. Just to name a few that I've encountered.

​I didn't be aware of this high number of reference books used in German or Swiss states. I agree more than five fields are needed.

When this feature was added, we had a few dozens of reference books, now there are 792. By the way 85 never used, and 59 with only one coin.

But is there really a "need" for 12 different "references".
In the ref-list I see many collections which are at most index lists not really catalogs and if the coins were referenced only in a particular list there would be no need for so many slots.
If we had a similar system for sourcing the information on our coin pages like a wiki I would understand your concerns more or am I missing something important.

If you are talking about the case were every year line has its own number but from the same catalog (most often the case for specified catalogs where they would run out of things to do if they wouldn't record every little fart) I would get that. Here an additional field for reference per line would be helpful (right now it is most often in the non-searchable comment line).

Hi Idolenz, zegeri,

For example for Appenzell AR/IR, a Swiss canton:
1. KM# - for general collectors
+ Y# (?) I am not familiar with the Yeoman catalog, but maybe some people still use it
2. Schön# - for Europeans and German speakers

3a. Fr# - for gold coin collectors
3b. Dav# - for collectors of Thalers
3c. Horat# - for collectors of small coins

4. HMZ# - for Swiss coins
5. Divo# - for Swiss coins, a bit older but more complete
6. Hofer# - this is even older and I don't know much about it, but I still see it

Then catalogs for Appenzell. 7 and 8 mostly, 9 is a bit older.
7. Tobler, 1969#
8. Jaeger, Lavanchy 1963#
9. Tanner, 1950#
+10. Reinhart 2019 - for coins of Bodensee, which also includes Appenzell

If the coin also happens to be a Pattern:
11. Richter PR#
+ Other references for patterns I am less familiar with (Divo, Lavanchy etc)
So easily more than 5.

Re. ancient coins. Generic catalogs (RIC, RCV, Cohen) are not complete, so more specific catalogs are essential. These are for a particular hoard or collection (private/ museum) and can have conflicting attributions, so more refs can create a more complete description of the coin.

best
strato

Quote: "zegeri"​Five reference books looks enough for near all the countries. Do you know any example that needs 6 books?
​
​The example you provide only needs four: EH, H, Dav and KM.
​https://en.numista.com/catalogue/pieces33905.html
​
​The "user's guide" of the multi-reference feature says to write one reference number per book in References, and all the numbers in Comments in year lines.
​

​At Uruguay Marcos Silvera Antunez (SA reference) uses 1 number for each year of the coin, and adds sub levels for variants and alignment data, I can use the same reference for variations/alignments, but not for the years.

His structure is ##.#.#, where first is coin identification, starting with 0 (as said 1 per year), middle is variant, starting with 1, and last is alignment: 1 for coin alignment, 2 for medal, 3 for 90º/270º, and 4 for all the rest in between.

Here is a sample of one of the coins (there were no variants nor alignment variations on them):



It makes 8 different references for SA book, and also are specified KM# 20 and Schôn# 2 (Yeoman Y# 16a still was not added).

So no, I don't need 6 books, I need 8 references for 1 book and 3 more books.

Hi
Based on only five available catalogue references, which can be added to any coin/token, it might be a useful approach to select top five references for each state, separately for coins and tokens (to please majority of users, other references can be added in "comment field". In future top five may also change...
lp

Thanks to everyone for joining this idea!
... and specially for providing such powerful "case studies"" to support the initiative.
Thank you, stratocaster!

I also love the idea referees being able to set a "default reference" and would like improve on that: what if when sorting by reference, NUMISTA would kindly ask: "which one, please?" - then both referees and users can sort the coins according to their prefence.
Thank you, druzhynets!

@zegeri:
The original dilemma of mine is stemming from the fact that reference books (i.e. catalogs) have a different aspect or perspective to organize the coins. Let me give 3 examples, and - kind of prove - that even with 3 reference books if I do a thorough job I cannot fit into 5 lines.
In the Hungarian example, we have 3 major (ignoring specialist catalogs) sources:
1) H# by Lajos Huszár, who was a scientific numismatist, historian and his works are very precisely documenting different versions based on evidences of other sciences (heraldry, history, symbolism, documents and stamps of rulers etcc.)
2) ÉH# by Emil Unger, who was a collector himself and used a thorough, but pragmatic approach
3) KM# by Krause, we all know, vague and generous, listing sometimes different rulers under 1 number, then annoyingly saying: "varieties exist" - these varieties are: different portraits, different coats of arms, adding & missing symbols - really like all - I mean all - Queen Elizabeth II 1 pound coins would be under one number and ignore the shields and symbols of the 4 kingdoms at the reverse totally plus the different portraits.

Now, the dilemma is:
1) If I create pages per H#, then I shall add sometimes 2-3 ÉH# to be searchable, and the KM number (5 lines done) and here is the key problem: - practically (using my original coin up in the thread as an example) that would result in 9 results for a search of KM#241 - pretty confusing for the average user
2) If I create pages per ÉH#, then I shall add 3-4 H# numbers to be searchable, plus the KM#241 (5 lines done) to all of this four... this will create 4 KM#241 pages, which is still more user-friendly.
3) If I create pages per KM#, - as it is now - I shall add 4 ÉH# and 9 H# numbers: that is 15 reference numbers!

I will go for the middle option: 1 ÉH#; 2-4 H#; 1 KM# number, then the 5 lines are done, and then comes all the others, Schön, other Hungarian catalogs, Davenport, etc... as listed by starocaster...

@adanieluy
You wrote in the other thread:

Just a crazy thought: could be possible to modify to add more than 1 number on each book, separated by marks (commas, slashes, etc) to make them searchable, something like words added for search the tags on "search for a coin" start page? (Book: SA, #s: 20,21,24,28, that would mean to search on SA value 20 or 21 or 24 or 28)

... and @zegeri
I was not aware of that peculiarity of Spanish catalogs, that number all coins differently by year - till yesterday, when I bought my first Spanish catalog in Sevilla.
Your "crazy thought" might be a good solution for some (but not all) the issues mentioned above. That is probably a technical issue at the discretion of Xavier; the key thought is to make all reference numbers relevant: searchable through advanced search.
... I am still convinced that 5 lines will not be enough to do a good job with the user experience in focus.

Thanks for your thoughts,
Imre

Bonsoir Xavier,

Let me then summarize the three findings from this thread and the other one, which has evolved thanks to the contributors:

1) Allow more than 5 - potentially unlimited - reference sources, surely needed;
2) With the same reference item, allow sequential numbers to be searchable, i.e. H# 1212, 1213, 1214 - in the same row - as suggested by adanieluy ...
3) Make referees - or even better - everyone; to set default references or their preferences in their search for reference sequence, i.e. in advance search: sort by reference: allow to choose by which reference. ... by druzhynets

Looking forward to your kind reaction, as always, mate,
Imre

I recieved a message from someone about being able to search multiple reference numbers on pages that group reference numbers. I told them that, if a suggestion already exists, I would bump the topic. The idea is very similar to imreh's Point 2, so... I am giving this thread a little bump.

Quote: "imreh"​2) With the same reference item, allow sequential numbers to be searchable, i.e. H# 1212, 1213, 1214 - in the same row - as suggested by adanieluy ...

While I think it would be a great idea (it would definitely help within China), I imagine that, if each indiviudal number appeared within the reference section, the reference section would look very crowded.

What might look nice is that only the very first number appears within the reference section. Then, when hovering above that number, all other numbers appear in some sort of drop down (which would basically be like hiding those additional numbers to avoid a cluttered reference section).

druzhynetsOn top of that, would be nice to have each reference in a new line, because currently it looks a bit messy:



and last but not least, referees should be able to choose default sorting by reference for their country. 
Let's say, there at least 3 different references used for Roman coinage. For consistency, as a referee I'd love to set how to sort coins by references, for example: 1. sort by RIC catalogues; 2. RCV; 3. Cohen and so on. Then the references would be sorted and shown in a uniform way.

Agree!

N#5310

 

the km220 (not a KM#) has 5 sub-numbers: km220.1 ; kmm220.2, km220.3 km220.4 and km220.5. Even if I change the km220 I will only have room for 4, since the Schoen# is also taken.

 

Please make room for at least 10 references for every coin.

 

I think, I've asked for that once before, but I couldn't find the reference.

 

Thanks in advance

Sjoelund

I think, I've asked for that once before, but I couldn't find the reference.

 

https://en.numista.com/forum/topic122388.html 

BTW, if we have the 5 km# (km220.1; km220.2; km220.3; km220.4 and km220 .5) referenced for one coin entry, knowing that the km220 doesn't exist and has never existed, I believe that it would still be possible to search for all the km# starting with 220 and get them properly if in one or several n#!

 

It's possible that the search times will be prolonged, but with the speed of the processors doubling every now and then, it's a false problem nowadays. 

Sjoelund

BTW, if we have the 5 km# (km220.1; km220.2; km220.3; km220.4 and km220 .5) referenced for one coin entry, knowing that the km220 doesn't exist and has never existed, I believe that it would still be possible to search for all the km# starting with 220 and get them properly if in one or several n#!

 

It's possible that the search times will be prolonged, but with the speed of the processors doubling every now and then, it's a false problem nowadays. 

Naive question: if KM 220 does not exist at all, as you stated quite repeatedly, why did they sorted coins as 220.1, 220.2, 200.3 etc (sub-numbers of 220) instead of calling them 220, 221, 222, 223 etc?

It is a false debate in my opinion. KM #220 obviously exist. It is what we call a concept, meaning the sum of each of its extensions (the sub numbers).

It is a sophism to say it does not exist, exactly same as saying that notion of car does not exist because there are so many different cars.

Numista registers entries, which are categories of collectible items. These categories may or may not have codifications in certain references books, it changes absolutely nothing regarding their ontology.

 

I precise I totally support the fact to add more references in these entries anyway 😀

I know that it is a somewhat absurd debate, since in the end we almost all agree that more lines should be added for references... but the KM#220 does not exist. 

 

Numista has its own numbering, but when we put the references of other catalogs we must adjust to their exact coding. Krause uses xxx.x for what they consider variants, so KM#220 doesn't exist, there is KM#220.1, 220.2... If Krause's system had been 220, 221, 222, would we have put in Numista all variants under the KM#220 or would we have indicated 220, 221, 222...?

oynbcn

I know that it is a somewhat absurd debate, since in the end we almost all agree that more lines should be added for references... but the KM#220 does not exist. 

 

Numista has its own numbering, but when we put the references of other catalogs we must adjust to their exact coding. Krause uses xxx.x for what they consider variants, so KM#220 doesn't exist, there is KM#220.1, 220.2... If Krause's system had been 220, 221, 222, would we have put in Numista all variants under the KM#220 or would we have indicated 220, 221, 222...?

We would have followed our guidelines, meaning create entries only when differences are significant between versions of the same type.

References do not all use same granularity, I dont consider we should follow necessarily one system over another.

But Krause chose to put all those under KM 220 which makes it easy anyway, for this case :-)

We should all agree to disagree, we all know that km#220 factually does not exist in SCWC, but the conceptual umbrella number of 220.1, 220.2 etc is 220. In our heads it exists.

 

My last words to that, since it seems we all agree on more reference numbers in numista!

Any timeline about having more than 5 references? Its cumbersome to maintain extra references in comments

Sjoelund

BTW, if we have the 5 km# (km220.1; km220.2; km220.3; km220.4 and km220 .5) referenced for one coin entry, knowing that the km220 doesn't exist and has never existed, I believe that it would still be possible to search for all the km# starting with 220 and get them properly if in one or several n#!

 

It's possible that the search times will be prolonged, but with the speed of the processors doubling every now and then, it's a false problem nowadays. 

In specific cases, where there are sub-types to the main reference, maybe a move of referencing system to year lines would be simpler solution and obtaining the main reference in between the main references. Along the way also additional pictures of sub-types could be finally added (Chines cast coins for example). Sub-type references and pictures all in one place.

LP

Not possible to use in a search…

And also I'm not talking about sub references (which are another interesting topic, though) but about having more than 5 main references (for carolingian coins it happens a lot with the classical Prou, Gariel, Morrison + potential Dep, MEC, Nouchy, CNI, Krum, PA, Chwartz, etc etc)

Hello,

It's now possible to enter up to 10 references.
The next much needed step will be to have searchable references per year lines.

Xavier

Hello,

It's now possible to enter up to 10 references.
The next much needed step will be to have searchable references per year lines.

Thanks for the Info and having done it!

Xavier

Hello,

It's now possible to enter up to 10 references.
The next much needed step will be to have searchable references per year lines.

Good !

Would be also good to can move up and done the reference line as we will have a lot more now.