bugFree UCS Outline Fonts - Bugs: bug #49487, Regressions in revision 3286

 
 

bug #49487: Regressions in revision 3286

Submitted by:  David Corbett <dscorbett>
Submitted on:  Mon 31 Oct 2016 01:44:30 PM UTC  
 
Category: overall font problemSeverity: 4 - Important
Item Group: glyph(s) missingStatus: Fix posted
Privacy: PublicAssigned to: Steve White <Stevan_White>
Open/Closed: OpenRelease: development

Add a New Comment(Rich Markup)
   

You are not logged in

Please log in, so followups can be emailed to you.

 

(Jump to the original submission Jump to the original submission)

Sun 09 Jul 2017 11:36:03 PM UTC, comment #7:

I think that was the last of the regressions of 3286.

David Corbett <dscorbett>
Sun 09 Jul 2017 10:39:15 PM UTC, comment #6:

That last patch was pretty messed up -- hard to make much sense of it.
I did see that it touched on the two issues you mentioned.

I had to do a lot of manual fixups before, I just missed this last patch (or got fatigued).
Anyway: Please check out the formation of IJ and ij ligature in small caps,
and note that the OS/2 character ranges table is no longer automatic.
(By the way, the reason for that is that different FontForge versions populate this table differently, and I am forced to switch between different versions in avoidance of their peculiar bugs.)

The last and scariest trick to get the IJ thing going was to re-order the Latin lookups to be like they are in Sans Bold.
I hope this doesn't make anything in this face any worse than they are in that face.

Let me know if you see any further problems.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Sun 09 Jul 2017 06:57:54 AM UTC, comment #5:

Hi David,

Thanks, I had somehow missed your heads-up of May 19.

Fortunately, in the file I posted, I made pretty extensive notes of what I though I was doing with each patches. Indeed no patch after the 3242-3245 was applied, the reason being that something in that patch fouls up subsequent patches.

In the "procedure" notes, I indicated that I didn't thing 3287 was important, and that 3286 was the unsuccessful attempt to perform a regression. That explains what you're seeing.

But now I look into the patch file for that last commit, and besides the ij changes, I see other little tweeks. I should review that. As to why I thought it was unimportant, it ends with a lot of validations -- which are unimportant. Maybe I only saw those.

It isn't a huge patch but it contains several meaningful changes. Besides that I tweaked a width, adjusted the anchors on macron, and made several alterations to lookups.

This will take some time to sort out.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Sun 09 Jul 2017 01:53:57 AM UTC, comment #4:

This is just a reminder that there are still regressions from revision 3420. The longer the wait before fixing them, the harder they will be to fix.

David Corbett <dscorbett>
Thu 18 May 2017 11:02:09 PM UTC, comment #3:

I’ll look at this in more detail later, but here’s what I’ve noticed. The most recent revision to FSaBO (3420) reverts all the changes since revision 3286; luckily there are only two: 3287 and 3340. It removes the OS/2 code page and Unicode ranges, which most of the other fonts include. It removes the substitutions from IJ and ij to sc.ij.

David Corbett <dscorbett>
Thu 18 May 2017 08:46:14 PM UTC, comment #2:

OK this is like the fourth time I've been through this.
The results may not be perfect, but they're better than before, I think.

It looks as if the original problem way back at rev. 3105 looks like an inadvertent re-encoding to "Custom". That in itself is bad, if only that it knocks this face out of synch with the other faces. Anyway, other bad things happened in between. Then I tried to fix it, and made matters worse.

This version seems to have the letters you mentioned, and auxiliary letters where they were expected to be.

I balled up the files I used to build this (attached), with notes indicating what I did by hand, etc. in "procedures".

Thanks for pointing this out, David.
I should make a script to check for such nonsense -- but... how do I know what horseplay I'll get into next?

(file #40735)

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 31 Oct 2016 02:40:51 PM UTC, comment #1:

OK then I screwed it up.
The intent was to skip a single commit where something had gone wrong, but this sounds like several commits.

Thanks David.

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Mon 31 Oct 2016 01:44:30 PM UTC, original submission:

Revision 3286 introduced some regressions to FreeSans Bold Oblique.

U+A7FA LATIN LETTER SMALL CAPITAL TURNED M should look like a Cyrillic small sha, but it now looks like a turned small capital M.

The following characters were removed.

  • U+05BA HEBREW POINT HOLAM HASER FOR VAV
  • U+05C5 HEBREW MARK LOWER DOT
  • U+05C6 HEBREW PUNCTUATION NUN HAFUKHA
  • U+05C7 HEBREW POINT QAMATS QATAN
  • U+1D6B LATIN SMALL LETTER UE
  • U+218A TURNED DIGIT TWO
  • U+218B TURNED DIGIT THREE
  • U+25CC DOTTED CIRCLE
  • U+A7B0 LATIN CAPITAL LETTER TURNED K
  • U+A7B1 LATIN CAPITAL LETTER TURNED T
  • U+A7B6 LATIN CAPITAL LETTER OMEGA
  • U+A7B7 LATIN SMALL LETTER OMEGA
  • U+FE00 VARIATION SELECTOR-1
  • U+FE01 VARIATION SELECTOR-2
  • U+FE02 VARIATION SELECTOR-3
  • U+FE03 VARIATION SELECTOR-4
  • U+FE04 VARIATION SELECTOR-5
  • U+FE05 VARIATION SELECTOR-6
  • U+FE06 VARIATION SELECTOR-7
  • U+FE07 VARIATION SELECTOR-8
  • U+FE08 VARIATION SELECTOR-9
  • U+FE09 VARIATION SELECTOR-10
  • U+FE0A VARIATION SELECTOR-11
  • U+FE0B VARIATION SELECTOR-12
  • U+FE0C VARIATION SELECTOR-13
  • U+FE0D VARIATION SELECTOR-14
  • U+FE0E VARIATION SELECTOR-15
  • U+FE0F VARIATION SELECTOR-16
David Corbett <dscorbett>

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach File(s):
   
   
Comment:
   

Attached Files
file #40735:  patching.tgz added by Stevan_White (828KiB - application/x-compressed-tar)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by Stevan_White (Posted a comment)
  • -unavailable- added by dscorbett (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only logged-in users can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 7 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Thu 18 May 2017 08:46:14 PM UTCStevan_WhiteAttached File-=>Added patching.tgz, #40735
      CategoryNone=>overall font problem
      Severity3 - Normal=>4 - Important
      Item GroupNone=>glyph(s) missing
      StatusProceeding=>Fix posted
    Mon 31 Oct 2016 02:40:51 PM UTCStevan_WhiteStatusNone=>Proceeding
      Assigned toNone=>Stevan_White

    Back to the top


    Powered by Savane 3.1-cleanup1