Closed patricksurry closed 7 months ago
I meant that the code assumes n < 256 even tho it's passed on the stack as a 16bit word. I can clarify
On Wed, Apr 17, 2024 at 9:27 AM SamCoVT @.***> wrote:
@.**** commented on this pull request.
In taliforth.asm https://github.com/SamCoVT/TaliForth2/pull/67#discussion_r1568845294:
@@ -268,6 +268,108 @@ _nibble_to_ascii:
rts
+ +find_header_name:
- ; """Given <266 char string ( addr n -- )
I'm assuming you wanted "256" here
— Reply to this email directly, view it on GitHub https://github.com/SamCoVT/TaliForth2/pull/67#pullrequestreview-2006048271, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABA5MKVOE5TSMP2LILYEJZ3Y5Z2DXAVCNFSM6AAAAABGLFCKOKVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDAMBWGA2DQMRXGE . You are receiving this because you authored the thread.Message ID: @.***>
This looks really good. I'm not surprised there was some low-hanging savings here, as wordlists are a recent addition to Tali and they were "tacked on" rather than fully integrated from the start. There may be more optimizations available here, but I think you've gotten most of it already in this PR.
I can probably review and merge this PR this coming weekend. You're welcome to push any other changes/tweaks you find in the meantime.
I thought your original comment was clear enough - it just needed "256" instead of "266". It looks good in your recent commit.
ha, my brain still didn't see the doubled six, i was focused on the < sign
On Wed, Apr 17, 2024 at 9:43 AM SamCoVT @.***> wrote:
I thought your original comment was clear enough - it just needed "256" instead of "266". It looks good in your recent commit.
— Reply to this email directly, view it on GitHub https://github.com/SamCoVT/TaliForth2/pull/67#issuecomment-2061294777, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABA5MKVDSOJBAZHH3SY47SLY5Z365AVCNFSM6AAAAABGLFCKOKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANRRGI4TINZXG4 . You are receiving this because you authored the thread.Message ID: @.***>
i'll stop messing with this branch and once we finalize I'll look at splitting up the native_words.asm
Saves 100 bytes or so and avoids duplication; equivalent performance