Thunderforge/scummvm-trac-migration-poc

SCI: LB1: Notebook rendering "INCOMPLETE" wrong

Closed this issue · 8 comments

The word "INCOMPLETE" is rendered incorrectly in the notebook. ScummVM is placing the last "E" on a new line.

Screenshot attached. Play-through with correct text at 18:35: https://www.youtube.com/watch?v=1XUptEPs1RA

Saved game attached at THE END screen right before notebook, made with dos version and nightly build.


Ticket migrated from Trac

  • Last modified: 2019-07-09 01:52:15
  • Owner: sluicebox
  • Reporter: sluicebox
  • Status: closed
  • Resolution: fixed
  • Keywords: has-pull-request

This text is drawn by a kDisplay call, with width set to 50. (Hardcoded in script 786.)

kDisplay: 786, 117, 100, 23, 43, 106, 50, 102, 4, 103, 65535, 105, 4

The text "INCOMPLETE" seems to be 52 pixels wide, even in this youtube play-through. Haven't investigated disasm yet to see what happens in SSCI.


Comment 1 migrated from Trac

  • Author: wjp

wjp, I know that we are currently not pixel perfect in some situations.
I noticed that while watching a Larry 5 Let's Play using ScummVM and the user look at some game credits windows and those windows weren't placed at the same pos as in SSCI.

But when trying to fix it, I encountered a few more issues. Like in some cases we directly use a Sierra-rect, but we shouldn't because Sierra-rects are inclusive and ours are exclusive.

You can also see this in QfG1 character creation when you compare the output of ScummVM and SSCI. They do not match perfectly.

So maybe this is also caused by that code, idk.


Comment 2 migrated from Trac

  • Author: m-kiewitz

sorry for typos, too late and need some sleep


Comment 3 migrated from Trac

  • Author: m-kiewitz

scummvm/scummvm#1737

  • Keywords: set to has-pull-request
  • Owner: set to sluicebox

Comment 4 migrated from Trac

  • Author: sluicebox

In [changeset:"ca66a776b0444a4af5cde844a924b8f25baffb77" ca66a776]:

#!CommitTicketReference repository="" revision="ca66a776b0444a4af5cde844a924b8f25baffb77"
SCI: Implement early GetLongest() behavior, bug #10000

Comment 5 migrated from Trac

  • Resolution: set to fixed
  • Status: newclosed

Comment 6 migrated from Trac

  • Author: bluegr

Attachment: incomplete_bug.png (4.66 KiB)

[[[Attachment]]]


Author: sluicebox

Attachment: laurabow.030 (23.73 KiB)

[[[Attachment]]]


Author: sluicebox