pre-apply-curry:
186.2u 2.2s 3:10 99% 0+0k 0+1io 0pf+0w

after:
79.9u 1.5s 1:22 98% 0+0k 0+9io 0pf+0w

after vector concatenate-as:
73.9u 1.6s 1:16 98% 0+0k 0+12io 0pf+0w

after vector member?:
68.6u 1.5s 1:11 98% 0+0k 0+1io 0pf+0w

after vector empty?:
68.8u 1.8s 1:12 97% 0+0k 0+13io 0pf+0w

after vector every?:
71.2u 1.8s 1:15 97% 0+0k 0+6io 0pf+0w
68.5u 1.6s 1:14 93% 0+0k 0+8io 0pf+0w
68.2u 1.6s 1:10 98% 0+0k 0+13io 0pf+0w

after eliminating member?("</n"):
59.6u 1.6s 1:02 98% 0+0k 0+10io 0pf+0w
60.1u 1.3s 1:02 98% 0+0k 0+12io 0pf+0w

after vectorizing subsequence-position:
50.4u 1.7s 1:13 70% 0+0k 0+2io 0pf+0w
50.6u 1.7s 1:13 71% 0+0k 0+6io 0pf+0w

after special casing subsequence-position:
48.2u 2.0s 1:13 68% 0+0k 0+0io 0pf+0w
48.0u 1.8s 1:10 70% 0+0k 0+13io 0pf+0w

After fixing as(<string>) stuff:
48.1u 2.0s 1:13 68% 0+0k 0+6io 0pf+0w

> After eliminating as(<string>) stuff:
> 52.2u 1.8s 1:15 71% 0+0k 0+3io 0pf+0w
> 51.9u 1.8s 1:14 72% 0+0k 0+12io 0pf+0w

48.9u 1.8s 1:11 70% 0+0k 0+2io 0pf+0w

> After using replace-elements!:
> 51.4u 1.7s 1:14 71% 0+0k 0+3io 0pf+0w
> 51.5u 1.6s 1:13 71% 0+0k 0+14io 0pf+0w

> After vectorising replace-elements!:
> 49.5u 1.9s 1:12 70% 0+0k 0+1io 0pf+0w
> 49.5u 1.5s 1:11 71% 0+0k 0+14io 0pf+0w

49.0u 1.7s 1:11 71% 0+0k 0+1io 0pf+0w
49.3u 1.8s 1:15 67% 0+0k 0+5io 0pf+0w
48.4u 1.8s 1:11 70% 0+0k 0+14io 0pf+0w

after optimizing subseq element:
46.8u 2.0s 1:09 70% 0+0k 0+10io 0pf+0w
46.9u 2.0s 1:12 67% 0+0k 0+5io 0pf+0w

after vectorizing do:
47.8u 1.8s 1:10 70% 0+0k 0+14io 0pf+0w
47.8u 1.8s 1:10 70% 0+0k 0+14io 0pf+0w

after removing do_eol member? calls:
47.0u 1.8s 1:09 70% 0+0k 0+8io 0pf+0w
47.3u 1.7s 1:09 70% 0+0k 0+14io 0pf+0w

> after "optimizing" sfind:
> 46.8u 1.8s 1:09 70% 0+0k 0+1io 0pf+0w
> 46.8u 1.9s 1:09 70% 0+0k 0+14io 0pf+0w

> eliminating test fun from sfind
> 45.3u 1.7s 1:07 69% 0+0k 0+2io 0pf+0w

> after vectorizing fill!:
> 46.3u 1.8s 1:08 70% 0+0k 0+8io 0pf+0w

47.4u 2.3s 1:13 67% 0+0k 0+14io 0pf+0w
47.3u 1.7s 1:09 70% 0+0k 0+12io 0pf+0w
47.0u 1.8s 1:09 70% 0+0k 0+14io 0pf+0w

new byte ops:
44.3u 1.8s 1:03 72% 0+0k 0+7io 0pf+0w
44.2u 1.8s 1:01 74% 0+0k 0+11io 0pf+0w

new version 4/19/94:
43.2u 2.8s 1:25 54% 0+0k 44+7io 0pf+0w
42.9u 2.6s 1:11 63% 0+0k 2+12io 0pf+0w
43.1u 2.5s 1:14 61% 0+0k 0+11io 0pf+0w
42.8u 1.7s 1:04 69% 0+0k 0+7io 0pf+0w
42.8u 1.8s 1:04 68% 0+0k 0+12io 0pf+0w

with full SAM cache:
40.8u 2.0s 1:04 66% 0+0k 44+1io 0pf+0w
41.0u 2.0s 1:08 63% 0+0k 0+2io 0pf+0w
40.9u 1.7s 1:04 65% 0+0k 0+11io 0pf+0w
40.7u 2.1s 1:09 62% 0+0k 0+10io 0pf+0w
40.8u 1.6s 1:05 65% 0+0k 0+12io 0pf+0w

after eliminate *max-short*:
39.3u 1.7s 1:02 65% 0+0k 0+2io 0pf+0w
39.1u 1.8s 1:01 66% 0+0k 0+9io 0pf+0w

after "optimizing" min and max:
39.2u 2.1s 1:05 63% 0+0k 15+12io 0pf+0w
39.1u 1.5s 1:01 65% 0+0k 39+6io 0pf+0w
39.0u 1.5s 1:01 65% 0+0k 0+12io 0pf+0w

after "optimizing" subsequence postion for byte-strings:
> 39.6u 1.9s 1:02 67% 0+0k 12+12io 0pf+0w
> 40.1u 1.7s 1:02 67% 0+0k 0+10io 0pf+0w
39.3u 1.6s 1:00 67% 0+0k 0+8io 0pf+0w
39.1u 1.7s 1:00 67% 0+0k 0+12io 0pf+0w
39.0u 1.8s 1:01 66% 0+0k 0+1io 0pf+0w
39.2u 1.9s 1:03 65% 0+0k 0+12io 0pf+0w

(replace on every line -- without full SAM cache:
50.0u 2.0s 1:16 68% 0+0k 42+5io 0pf+0w
replace on every line -- with full SAM cache:
41.8u 2.1s 1:05 67% 0+0k 41+2io 0pf+0w)


After writing a proper replace-string/substring-position:
39.0u 2.4s 1:14 55% 0+0k 0+4io 0pf+0w
38.3u 1.5s 1:00 66% 0+0k 0+11io 0pf+0w
38.4u 1.5s 0:59 67% 0+0k 0+14io 0pf+0w

After *fixing* SAM cache:
36.6u 1.6s 0:57 66% 0+0k 0+14io 0pf+0w
36.7u 1.7s 0:59 64% 0+0k 0+14io 0pf+0w
36.5u 1.6s 0:57 66% 0+0k 0+6io 0pf+0w

> After adding subclass specializers to vec.c:
> 37.6u 2.1s 1:02 63% 0+0k 14+3io 0pf+0w
> 37.6u 1.4s 0:58 66% 0+0k 0+11io 0pf+0w
> 37.6u 1.5s 1:00 64% 0+0k 0+15io 0pf+0w

36.6u 1.9s 0:58 66% 0+0k 0+3io 0pf+0w

After adding limited int support to SAM cache:
36.3u 1.7s 0:59 64% 0+0k 12+12io 0pf+0w
36.2u 1.7s 0:57 65% 0+0k 0+3io 0pf+0w
36.4u 1.6s 0:57 65% 0+0k 0+11io 0pf+0w

35.7u 2.0s 0:59 62% 0+0k 1+17io 0pf+0w
35.7u 1.8s 0:58 64% 0+0k 0+12io 0pf+0w
35.6u 1.5s 0:56 65% 0+0k 0+13io 0pf+0w

After correctiong "simple" calculations for SAM cache:
35.3u 2.1s 0:58 64% 0+0k 41+10io 0pf+0w
35.5u 1.7s 0:57 64% 0+0k 0+9io 0pf+0w
35.5u 1.8s 0:56 65% 0+0k 0+9io 0pf+0w
35.5u 1.5s 0:56 65% 0+0k 0+9io 0pf+0w

After fixing bracketing problem with SAM cache:
36.0u 2.0s 1:00 63% 0+0k 31+5io 0pf+0w
36.0u 1.8s 0:58 65% 0+0k 0+2io 0pf+0w
36.0u 1.5s 0:57 65% 0+0k 0+10io 0pf+0w

Unknown changes to 5/3/94:
40.0u 1.9s 1:03 65% 0+0k 60+11io 0pf+0w
40.0u 2.3s 1:06 63% 0+0k 0+4io 0pf+0w
40.0u 2.8s 1:08 62% 0+0k 0+10io 0pf+0w

Swap in hash tables for solist:
43.7u 2.4s 1:07 68% 0+0k 0+5io 0pf+0w
43.5u 2.4s 1:07 68% 0+0k 0+3io 0pf+0w
43.5u 2.3s 1:07 68% 0+0k 0+12io 0pf+0w

Add new key-hash function:
43.8u 2.1s 1:07 68% 0+0k 0+5io 0pf+0w
43.9u 2.1s 1:07 68% 0+0k 0+12io 0pf+0w

Correctly optimize new key-hash:
42.8u 2.4s 1:07 66% 0+0k 0+1io 0pf+0w

Add optimized key-hash for strings:
42.1u 2.4s 1:05 67% 0+0k 0+2io 0pf+0w
42.3u 1.9s 1:06 67% 0+0k 0+12io 0pf+0w
42.2u 2.3s 1:06 67% 0+0k 0+12io 0pf+0w

Optimized byte-... copy-sequence (no optimized key-hash):
41.9u 2.5s 1:06 67% 0+0k 0+0io 0pf+0w

Added better method for collection "as":
40.2u 1.7s 1:01 67% 0+0k 0+0io 0pf+0w

Added a sequence "as":
39.4u 1.9s 1:01 67% 0+0k 0+1io 0pf+0w
"Optimized" (should improve lists, while pessimizing vectors):
41.5u 2.1s 1:04 67% 0+0k 0+0io 0pf+0w

Added vector "as":
41.3u 1.8s 1:03 68% 0+0k 0+0io 0pf+0w
40.7u 2.1s 1:03 67% 0+0k 0+0io 0pf+0w