It should be merged. I think the ディ was wrong.

Is ディ really an acceptable spelling?
By the way, should the entry be merged with the entry for デー?

@@ -4,4 +4,0 @@
-<r_ele>
-<reb>ディ</reb>
-<re_pri>gai1</re_pri>
-</r_ele>

ngrams

Merging 2482620.

@@ -8,0 +8,3 @@
+<reb>デーパック</reb>
+</r_ele>
+<r_ele>
@@ -9,0 +12,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・パック</reb>

Added additional dotted reading(s) via batch update.Added additional dotted reading(s) via batch update.
-*- via bulkupd.py -*-

@@ -6,0 +6,3 @@
+</r_ele>
+<r_ele>
+<reb>デイ・パック</reb>

ngrams

@@ -8,0 +8,3 @@
+<reb>デーベッド</reb>
+</r_ele>
+<r_ele>
@@ -9,0 +12,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・ベッド</reb>

Added additional dotted reading(s) via batch update.Added additional dotted reading(s) via batch update.
-*- via bulkupd.py -*-

@@ -6,0 +6,3 @@
+</r_ele>
+<r_ele>
+<reb>デイ・ベッド</reb>

ALC:
業務の計画を立てる 
set up an operation
業務の非効率性 
operational inefficiency
& lots of other examples.

@@ -16,0 +16,1 @@
+<pos>&adj-no;</pos>
@@ -24,0 +25,2 @@
+<gloss>(business) operations</gloss>
+<gloss>operational</gloss>

ＫＯＤ追加語彙: "hard to eat; difficult to get down; unpalatable."

Worth keeping, I think.

@@ -16,0 +16,1 @@
+<gloss>unpalatable</gloss>

necessary entry?  clear A+B; no other dictionary bothers with it

@@ -4,0 +4,3 @@
+<k_ele>
+<keb>食べにくい</keb>
+</k_ele>

Not a noun.

@@ -11,1 +11,1 @@
-<pos>&n;</pos>
+<pos>&adj-i;</pos>

Genius WaEi et al

@@ -20,0 +20,3 @@
+<gloss>severe</gloss>
+<gloss>grave</gloss>
+<gloss>acute</gloss>

meikyo

@@ -16,0 +17 @@
+<pos>&vt;</pos>
@@ -24,0 +26 @@
+<pos>&vt;</pos>

Raising this again as a correspondent has pointed out that in many cases in Japanese it can only mean prepayment; not COD. That second sense, which looks rather contradictory, seems mainly/only to be applied in cases such as business-reply mail and collect telephone calls, where the payment is guaranteed. For regular COD you use 着払い.
I'm not sure how to reflect this - I have proposed some examples which may help.

@@ -17,0 +17,1 @@
+<xref type="see" seq="1603800">前払い</xref>
@@ -24,1 +25,2 @@
-<gloss>payment on delivery</gloss>
+<xref type="see" seq="1423230">着払い・ちゃくばらい</xref>
+<gloss>payment on delivery (e.g. business-reply mail, collect phone calls)</gloss>

GG5, 新和英中辞典

Splitting senses.
See rejected 2794040.

@@ -17,0 +17,1 @@
+<gloss>payment in advance</gloss>
@@ -18,0 +19,5 @@
+<gloss>prepayment</gloss>
+</sense>
+<sense>
+<pos>&n;</pos>
+<pos>&vs;</pos>
@@ -19,0 +25,3 @@
+</sense>
+<sense>
+<pos>&n;</pos>

mk

@@ -16,0 +17 @@
+<pos>&vi;</pos>

@@ -24 +23,0 @@
-<pos>&vs;</pos>

@@ -25,0 +25,1 @@
+<xref type="see" seq="1692020">駐在所・ちゅうざいしょ</xref>

ALC:
（人）を海外に駐在させる 
station someone abroad
海外駐在 
overseas ［foreign］ posting

@@ -19,0 +19,2 @@
+<gloss>(job) posting</gloss>
+<gloss>being stationed (overseas, etc.)</gloss>

GG5

@@ -16,0 +17 @@
+<gloss>indecipherable</gloss>

n-grams
判読し難い	14
判読しがたい	34

@@ -3,0 +4,3 @@
+<k_ele>
+<keb>判読しがたい</keb>
+</k_ele>
@@ -10,0 +14 @@
+<pos>&exp;</pos>

Not a noun.

@@ -11,1 +11,1 @@
-<pos>&n;</pos>
+<pos>&adj-i;</pos>

n-grams

@@ -6,0 +6,6 @@
+</k_ele>
+<k_ele>
+<keb>かすめ取る</keb>
+</k_ele>
+<k_ele>
+<keb>掠めとる</keb>

n-grams

@@ -9,0 +9,3 @@
+</k_ele>
+<k_ele>
+<keb>藁ぶき</keb>

More gloss: alt wording (〜葺き meaning both “thatching” and “thatched roof”)

@@ -15,0 +15,1 @@
+<gloss>straw-thatching</gloss>

GG5

Specifically *straw* (as per 藁)

@@ -15,1 +15,1 @@
-<gloss>thatched roof</gloss>
+<gloss>straw-thatched roof</gloss>

@@ -25,0 +26 @@
+<gloss>going on about one's love affairs</gloss>

gg5

@@ -24 +24,2 @@
-<gloss>speaking fondly of</gloss>
+<gloss>boastful talk about one's love life</gloss>
+<gloss>speaking fondly about one's sweetheart</gloss>

Google n-grams confirm this.

n-grams
のろけばなし	6
ノロケばなし	1
のろけ話	728
惚気話	139
ノロケ話	358

@@ -24 +23,0 @@
-<misc>&uk;</misc>

@@ -3,0 +4,3 @@
+<k_ele>
+<keb>動かしがたい</keb>
+</k_ele>

As with 複合動詞, many of these compounds don't get into Japanese refs because they are formed dynamically. I think they are useful for learners, and probably worth keeping.

necessary?

Google N-gram Corpus Counts
╭─ーーーーーーーーー─┬────────┬───────╮
│ 筆舌に尽くしがたい │ 25,775 │ 70.4% │
│ 筆舌に尽くし難い　 │  8,071 │ 22.0% │
│ 筆舌につくしがたい │  1,401 │  3.8% │ - sK
│ 筆舌に尽しがたい　 │    580 │  1.6% │ - add, sK
│ 筆舌に尽し難い　　 │    507 │  1.4% │ - add, sK
│ 筆舌につくし難い　 │    289 │  0.8% │ - sK
╰─ーーーーーーーーー─┴────────┴───────╯

@@ -11,0 +12,9 @@
+<ke_inf>&sK;</ke_inf>
+</k_ele>
+<k_ele>
+<keb>筆舌に尽しがたい</keb>
+<ke_inf>&sK;</ke_inf>
+</k_ele>
+<k_ele>
+<keb>筆舌に尽し難い</keb>
+<ke_inf>&sK;</ke_inf>
@@ -14,0 +24 @@
+<ke_inf>&sK;</ke_inf>

n-grams

@@ -5,0 +5,3 @@
+<keb>筆舌に尽くしがたい</keb>
+</k_ele>
+<k_ele>
@@ -6,0 +9,6 @@
+</k_ele>
+<k_ele>
+<keb>筆舌につくしがたい</keb>
+</k_ele>
+<k_ele>
+<keb>筆舌につくし難い</keb>

Noun?

@@ -11,1 +11,2 @@
-<pos>&n;</pos>
+<pos>&exp;</pos>
+<pos>&adj-i;</pos>

GG5

Not a noun.

@@ -11,1 +11,1 @@
-<pos>&n;</pos>
+<pos>&adj-i;</pos>
@@ -13,0 +13,1 @@
+<gloss>hard to refuse</gloss>

correcting a reading problem

@@ -10,0 +11,3 @@
+<reb>やくしにくい</reb>
+</r_ele>
+<r_ele>
@@ -11,0 +15 @@
+<re_restr>訳し難い</re_restr>

GG5, Eijiro

I think these compound adjectives are worth keeping for beginners. The ~にくい versions are usually more common; in this case about 15:1.

@@ -4,0 +4,3 @@
+<k_ele>
+<keb>訳しにくい</keb>
+</k_ele>
@@ -13,0 +16,1 @@
+<gloss>untranslatable</gloss>

from the danger range of expressions; A+B and of very limited applicability; no other dictionary bothers with it

ngrams

@@ -8,0 +8,3 @@
+<reb>デーサービス</reb>
+</r_ele>
+<r_ele>
@@ -9,0 +12,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・サービス</reb>

Added additional dotted reading(s) via batch update.Added additional dotted reading(s) via batch update.
-*- via bulkupd.py -*-

@@ -6,0 +6,3 @@
+</r_ele>
+<r_ele>
+<reb>デイ・サービス</reb>

ngrams

@@ -8,0 +8,3 @@
+<reb>デートレーダー</reb>
+</r_ele>
+<r_ele>
@@ -9,0 +12,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・トレーダー</reb>

Added additional dotted reading(s) via batch update.Added additional dotted reading(s) via batch update.
-*- via bulkupd.py -*-

@@ -6,0 +6,3 @@
+</r_ele>
+<r_ele>
+<reb>デイ・トレーダー</reb>

That field isn't the place for those sorts of examples.

@@ -19 +18,0 @@
-<s_inf>1のゾロ目 a roll of all ones</s_inf>

In case anyone actually wanted to use ゾロ目, this is how it works...

See my treatise on ピンゾロ entry.  The most common (and most readily comprehensible) way to say "a roll of two ones" in Japanese is not ピンゾロ, but 1のゾロ目.  

This is also the safest translation of "snake eyes" in most cases(if you want to translate "bad luck" you'll need to add something extra...).

Also, it appears that this has a broader application than dice.  In numerology, for example, 5のゾロ目 is any number consisting of only 5s (also called "Angel Numbers" apparently).  55
555
5555 etc.

May also be used to talk about phone numbers with repeating digits in certain sections.

I don't know anything about numerology, but if you google "5のゾロ目" you'll get quite a large number of results that go well beyond dice.

Just throwing that out there. Not sure if it's worth a gloss: "a number consisting of a single repeating digit". (And maybe "Angel Number(numerology) if we're feeling adventurous, although apparently 56565656 is also an angel number if a sequence repeats, so maybe not).

@@ -18,0 +19 @@
+<s_inf>1のゾロ目 a roll of all ones</s_inf>

ゾロ目	93243	54.8%
ぞろ目	71007	41.8%
揃目	184	0.1%
ゾロめ	368	0.2%
ぞろめ	5257	3.1%

Simplifying.

@@ -12 +12 @@
-<ke_inf>&iK;</ke_inf>
+<ke_inf>&sK;</ke_inf>
@@ -15,4 +14,0 @@
-<reb>ゾロめ</reb>
-<re_restr>ゾロ目</re_restr>
-</r_ele>
-<r_ele>
@@ -20,2 +15,0 @@
-<re_restr>ぞろ目</re_restr>
-<re_restr>揃目</re_restr>

@@ -33,0 +34 @@
+<field>&math;</field>

I don't think that one really works very well.

@@ -24 +23,0 @@
-<gloss>almost</gloss>

@@ -23,0 +24 @@
+<gloss>almost</gloss>

なんちゃって	1084983 
なんつって	131618
* https://detail.chiebukuro.yahoo.co.jp/qa/question_detail/q1118991806
* https://hinative.com/ja/questions/8203635

Merging   2846211 submission.

@@ -6,0 +7,3 @@
+<r_ele>
+<reb>なんつって</reb>
+</r_ele>
@@ -11 +14 @@
-<gloss>Not!</gloss>
+<gloss>not!</gloss>

「なんてね」に合わせて

@@ -8 +8 @@
-<pos>&int;</pos>
+<pos>&exp;</pos>
@@ -10,2 +10,2 @@
-<gloss>just kidding!</gloss>
-<gloss>how corny!</gloss>
+<gloss>just kidding</gloss>
+<gloss>Not!</gloss>

@@ -14,4 +14,3 @@
-<gloss>It's hard to tell a poor thinker from a sleeping one</gloss>
-<gloss>They to whom only bad ideas come might as well be asleep</gloss>
-<gloss>Poor thinking is futile</gloss>
-<gloss>Inadequate ideas are worse than none at all</gloss>
+<gloss>it's hard to tell a poor thinker from a sleeping one</gloss>
+<gloss>they to whom only bad ideas come might as well be asleep</gloss>
+<gloss>inadequate ideas are worse than none at all</gloss>

@@ -12,2 +12,2 @@
-<xref type="see" seq="2416750">下手の考え休むに似たり・へたのかんがえやすむににたり</xref>
-<misc>&id;</misc>
+<xref type="see" seq="2416750">下手の考え休むに似たり</xref>
+<misc>&proverb;</misc>

新和英中辞典, ルミナス

Aligning glosses.

@@ -12,2 +12,7 @@
-<xref type="see" seq="1281000">考え・1</xref>
-<gloss>take a rest rather than useless thinking</gloss>
+<xref type="see" seq="2416750">下手の考え休むに似たり・へたのかんがえやすむににたり</xref>
+<xref type="see" seq="2416750">下手の考え休むに似たり・へたのかんがえやすむににたり</xref>
+<misc>&id;</misc>
+<gloss>It's hard to tell a poor thinker from a sleeping one</gloss>
+<gloss>They to whom only bad ideas come might as well be asleep</gloss>
+<gloss>Poor thinking is futile</gloss>
+<gloss>Inadequate ideas are worse than none at all</gloss>

prog: "It is a waste of time to think a matter over when one 
has no good ideas."
新和英中辞典　”Inadequate ideas are [Inadequate thinking is] 
worse than none at all."

this and twin entry 下手の考え休むに似たり (which is also in 
daijs) should have the same definition.
the current defs for 下手の考え休むに似たり are:
"It's hard to tell a poor thinker from a sleeping one" &
Mickle fails that fools think"
the "Mickle..." seems to be a jmdictism? very few hits 
outside the Japanese web.

@@ -14,4 +14,3 @@
-<gloss>It's hard to tell a poor thinker from a sleeping one</gloss>
-<gloss>They to whom only bad ideas come might as well be asleep</gloss>
-<gloss>Poor thinking is futile</gloss>
-<gloss>Inadequate ideas are worse than none at all</gloss>
+<gloss>it's hard to tell a poor thinker from a sleeping one</gloss>
+<gloss>they to whom only bad ideas come might as well be asleep</gloss>
+<gloss>inadequate ideas are worse than none at all</gloss>

@@ -12,2 +12,2 @@
-<xref type="see" seq="2407990">下手な考え休むに似たり・へたなかんがえやすむににたり</xref>
-<misc>&id;</misc>
+<xref type="see" seq="2407990">下手な考え休むに似たり</xref>
+<misc>&proverb;</misc>

新和英中辞典, ルミナス

Aligning glosses.

@@ -12,0 +12,2 @@
+<xref type="see" seq="2407990">下手な考え休むに似たり・へたなかんがえやすむににたり</xref>
+<xref type="see" seq="2407990">下手な考え休むに似たり・へたなかんがえやすむににたり</xref>
@@ -14,1 +16,3 @@
-<gloss>Mickle fails that fools think</gloss>
+<gloss>They to whom only bad ideas come might as well be asleep</gloss>
+<gloss>Poor thinking is futile</gloss>
+<gloss>Inadequate ideas are worse than none at all</gloss>

Entry created

As Richard pointed out in a comment on a deleted fork, ディ is turning up. We have it in the "バースデー;バースディ;バースデイ" entry. Richard wrote:

On the other hand, in support of ディ, I get results such as this:
Google n-gram counts
ホワイトデー: 663888
ホワイトデイ: 17366
ホワイトディ: 15101
.
But it seems strange that ディ could be used, 
since it seems to indicate pronunciation "dee" rather than "day".
Could it be just a common misspelling, perhaps?
Kojien, for example, has デイ and デー but not ディ.

I suggest adding ディ, tagged as "ik".

@@ -11,0 +11,4 @@
+</r_ele>
+<r_ele>
+<reb>ディ</reb>
+<re_inf>&ik;</re_inf>

デイ was [gai1] on the other entry

@@ -10,0 +10,1 @@
+<re_pri>gai1</re_pri>

temp approve because other entry is deleted

Keep the priority.

@@ -6,0 +6,1 @@
+<re_pri>spec1</re_pri>

デイトレード	606,445	99.8%	
デートレード	1,312	0.2%

The デートレード entries in daijr/s redirect to デイトレード. I think we can hide it.

@@ -8,3 +7,0 @@
-<reb>デートレード</reb>
-</r_ele>
-<r_ele>
@@ -14 +11,2 @@
-<reb>デー・トレード</reb>
+<reb>デートレード</reb>
+<re_inf>&sk;</re_inf>

@@ -17,0 +18 @@
+<field>&stockm;</field>

@@ -18 +18 @@
-<gloss>day trade</gloss>
+<gloss>day trading</gloss>

ngrams

@@ -8,0 +8,3 @@
+<reb>デートレード</reb>
+</r_ele>
+<r_ele>
@@ -9,0 +12,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・トレード</reb>

Merged with 1082540.

Added additional dotted reading(s) via batch update.Added additional dotted reading(s) via batch update.
-*- via bulkupd.py -*-

@@ -6,0 +6,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・パック</reb>

Entry created

ngrams

@@ -8,0 +8,3 @@
+<reb>デーホーム</reb>
+</r_ele>
+<r_ele>
@@ -9,0 +12,3 @@
+</r_ele>
+<r_ele>
+<reb>デー・ホーム</reb>

Daijr, GG5, ルミナス

I think this is "just" a compound adjective.

@@ -24,1 +24,0 @@
-<pos>&exp;</pos>

GG5, リーダーズ+プラス, Eijiro

Useful expression. Could be split into がたい/にくい versions although they share a (not very common) kanji form.

I think they can be regarded as "複合形容詞", which Eijiro says is a "compound adjective". Not in any refs, and only 217 ngram hits. I'll drop the "exp".

@@ -17,1 +17,0 @@
-<pos>&exp;</pos>

n-grams

Only one or two other ○○難い entries have [exp] right now, while most don't.

@@ -6,0 +6,3 @@
+</k_ele>
+<k_ele>
+<keb>考え難い</keb>

GG5 examples, Eijiro.

I think this idiomatic enough and common enough (250k in the ngrams) to include. 2 examples in the Tanakas.

daijs
G n-grams:
前彼	18018
前カレ	2893

@@ -6,0 +7,3 @@
+<k_ele>
+<keb>前カレ</keb>
+</k_ele>
@@ -8,0 +12,5 @@
+<re_restr>前彼</re_restr>
+</r_ele>
+<r_ele>
+<reb>まえカレ</reb>
+<re_restr>前カレ</re_restr>
@@ -12,0 +21 @@
+<misc>&col;</misc>

Daijr

I should have noticed that.

Oops, sorry. I intended to delete the 社員 at the end of the kanji.

@@ -5,1 +5,1 @@
-<keb>国外駐在社員</keb>
+<keb>国外駐在</keb>

That should do  it.

@@ -14,1 +14,1 @@
-<gloss>expatriate</gloss>
+<gloss>expatriate (employee)</gloss>

ALC:
国外駐在社員 
expat employee
海外駐在 
overseas ［foreign］ posting
海外駐在の家族 
expat ［transplanted］ family
海外駐在員の家族 
expatriate families
海外駐在費 
expatriate expense

Is there any way of indicating that "expatriate" here is intended as the adjectival sense only (not noun)?

Eijiro, http://cosmos.nobody.jp/grammar/adjectives_compound.html

Not common, but fills a gap.