Post XjR26vcvAX3

Paul Strack Jan 15, 2018 (20:31)

I am in the middle of a major overhaul of the Eldamo data model to reflect its new “dual mission” of both (1) describing Elvish as Tolkien wrote it and (2) suggesting possible uses for Neo-Elvish writing.

These changes are big enough to be considered “breaking changes” in the data model, so I wanted to give those who are using Eldamo data for their own online dictionaries a preview of these changes before I release it. In particular, a large number of “Sindarin” and “Quenya” entries are being removed from Eldamo. These are all empty entries that existed solely as means of indicating with Noldorin and (Middle) Quenya might be used in Neo-Sindarin and Neo-Quenya writing. Further details can be found here:

Basically, I am becoming more strict in which words the model says are Sindarin and Quenya words, limiting them (mostly) to words that Tolkien himself included as part of those languages in the 1950s and later.

I will have more to say about how I am organizing Neo-Quenya and Neo-Sindarin, but I need to get further in the reorganization before I am ready to do so. Basically, I think Neo-Quenya and Neo-Sindarin need to be “emergent” properties of the model: derived from the base data rather than incorporated directly into the data itself. That is because there are many ways to define Neo-Quenya and Neo-Sindarin, and I hope the core data model itself can remain sufficiently neutral that it can be still be used as the basis for a variety of approaches to Neo-Quenya and Neo-Sindarin writing.

For a preview of the data model with the empty “Sindarin” and “Quenya” entries removed, see here:

If you have questions, please feel free to ask, either here or on github.
Eldamo : Version History
[Home]. Version History. v0.6.0. Clean up old and out-dated phonetic artifacts. Most of the remaining data entry for PE19. First beta release! v0.5.9 (December 9, 2017). Data entry for PE18. v0.5.8 (November 11, 2017). Data entry for PE21. More Early Quenya phonetic analysis. v0.5.7 (August 20, 2017).

Ицхак Пензев Jan 15, 2018 (22:53)

That's a serious change. I need to think even, what I should ask about the way new system works. I'll keep a v.0.6.0 offline copy for a while for comparison.

Paul Strack Jan 15, 2018 (23:02)

+Ицхак Пензев There is a reason why I waited until the data entry was done before making these changes. I want people to be able to use the older releases if they don’t like the changes. However, from an end-user perspective, I hope to be able to preserve the existing functionality in (as yet unreleased) v0.6.1.

For example, I plan to have a search option for “Sindarin + Noldorin”, which would be more or less equivalent to the current “Sindarin” search.

Ицхак Пензев Jan 16, 2018 (07:02)

If, for example, one could search for Late Quenya plus Middle Quenya minus Early Qenya, that would be great.

Paul Strack Jan 16, 2018 (07:48)

+Ицхак Пензев Yes, that’s what I meant, except I gave Sindarin/Noldorin as an example. Basically, the ability to search for late and middle period words but to exclude the early period, as many people consider those to early to include.

I haven’t decided yet how exactly to implement it, though.

Tamas Ferencz Jan 16, 2018 (10:37)

+Paul Strack are you planning to extend/strengthen the semantic classification of words? I think many would find it useful when searching for a word of they could list semantically related entries.

Ицхак Пензев Jan 16, 2018 (14:00)

+Paul Strack you say the period (Early, Middle, Late etc.) will be generated from the reference. I hope the database will include precise dates for documents. As we see, some sources have various documents from various periods. Like, e.g. PE22 with EVS1 & QVS from 40s and LVS from late 60s.
(Besides, please accept my friend request in the Discord).

Paul Strack Jan 16, 2018 (16:03)

+Tamas Ferencz Yes, now that the data entry is done, semantic classification is a priority.

Paul Strack Jan 16, 2018 (16:09)

+Ицхак Пензев More precise dating is something I want to do eventually, but the model does not support it well at the moment. I am referring to the current, somewhat crude classification of Early (1910 - 1930), Middle (1930 - 1950) and Late (1950 - 1970). More precise dating is a challenge, since individual sources can reference items from multiple periods on the same page (for example, in footnotes).

Ицхак Пензев Jan 17, 2018 (12:17)

Well, at least it takes into account that PE22:92-140 is MQ. and PE22:141-168 is LQ. This is very good, at the moment.