the script

Whereas English is written with
A a B b C c D d E e
F f G g H h I i J j
K k L l M m N n O o
P p Q q R r S s T t
U u V v W x , ; . :
! ? " - 1 2 3 4 5 6
7 8 9 0 (  ) [  ] /  \
% & # ' + * ~ ^ { }
(80 chars)
the first qurʾān manuscripts just have
  ا   ٮ‍   ح‍ د ر و  ه‍ ط ك‍ ل‍
م‍ س‍ ص‍ ع‍ ڡ‍ ڡ ٯ ع ص س
* م ل ك ‌ـه‌ ح ٮ ں ى لا
(30 chars)

Most people know that the earliest manuscripts
do not have diacritical dots, no ḥamza sign, no numerals,
no shadda, no hyphen, colon, just an end of aya sign,
but hardly anybody is aware of two facts:

There is no space between words.
{Th. Bauer is wrong ("Words are set apart by greater spaces" in Peter T. Daniel, ed. p. 559).}
There is no hyphenation: end of line is insignificant.

Start letters and End letters are distinct letters
(although standing for the same sound, they carry a different meaning), whereas Start and Middle forms, End and Iso forms are "just" a consequence of the preceding letter.
As "conservative, liberals, god" are different from "Conservative, Liberals, God"
= A and a are not the same letter
ح‍ and ح are not the same letter
Just as capital letters carry a meaning (person, majesty, name, start of a sentence--in German: noun),
End (resp. Iso) carries the meaning: "end of the word".
Therefore there was no "space between words" -- or was it the other way round?
And because there is no End-waw (and because two alifs can not stand WITHIN a word),
after waw at the end of a word an alif was added: the word border runs between the two alifs.
Lakhdar-Ghazal saw a core letter and end markers:
ح ع م
ٮ ل ى
س ص ں
That does not work for all letters and not for all calligraphic styles.
Unicode sees colon, space, Non-Joiner as triggering the end form of ONE letter.
That is clearly wrong for the early manuscripts.

Bauer's "each letter may occur in four different positions: initial, medial, final, and isolated" is a truism, but it shows, that he noticed that the common statement "each letter has four forms/graphic shapes" is untenable, both because many have only one form (in typewriter script), and many have more than twenty (in "high" naskhi). Not trivial is "the common designation of the Arabic script as "consonantal" is incorrect, since the long vowels are represented but consonant gemination is not." (Bauer in Daniel p.561) -- although not ALL long vowels are represented (as Bauer knows of course), and some short vowels are represented and diphthongs as well.