none
Where do the 20.1.10.60 ST_TextAutonumberScheme (Text Auto-number Schemes) lists come from? RRS feed

  • Question

  • Product/Technology: PowerPoint/PresentationML

    The 20.1.10.60 ST_TextAutonumberScheme (Text Auto-number Schemes) lists are not documented anywhere that I can find. I've found on http://www.w3.org/TR/css3-lists/ the lists here are probably pretty similar or even the same to what Open XML implements, but I just needed to confirm the numbering/ordering lists.

    BTW, the numbering (digits) seems to be set by local settings of Windows (e.g.. 1, 2, 3 for English and ١, ٢, ٣ for Arabic) - there doesn't seem to be any rules within the actual file that forces any choice in code set. What I mean is that even if the paragraph/run is set to lang="ar-LB" (for Arabic Libya), unless Windows is set to Arabic, it will still display in Latinized Arabic numbers (1, 2, 3 instead of ١, ٢, ٣). However, I noticed that if I set Windows to Persian ("fa-IR") and create a numbered list of 4 - 6, it will display ۴, ۵, ۶  but then when I change Windows to Arabic these are are not changed to ٤, ٥, ٦. But if I change to English, they become 4, 5, 6. What should we be looking for when dealing with this? Is there a combination of values that will force a numbering system to be displayed one way and not another?

    Friday, October 21, 2011 6:47 PM

Answers

  • Hi Okatu,

    Thank you again for bringing to our attention that the implementation details of the ST_TextAutonumberScheme should be documented in the Microsoft documentation. If you want to see some specific case to be documented, please let me know.  You can send your suggestion to “dochelp (at) microsoft (dot) com” too, indicating that it is for me.

    Thanks, Vilmos

    Thursday, December 29, 2011 10:54 PM

All replies

  • Hi, Okatu,

      Thanks for your question.  One of our team members will work on it and respond to you soon.


    Hongwei Sun -MSFT
    Friday, October 21, 2011 8:48 PM
  • Hi Okatu,

    I am the engineer who will be working with you on this issue. I am currently researching the problem and will provide you with an update soon.

    Thanks,
    Vilmos Foltenyi | Escalation Engineer | Open Specifications Support Team
    Email: vilmosf@microsoft.com
    Phone: +1 (425) 421-2642
    Time zone: (UTC-08:00) Pacific Time (US & Canada)

    Saturday, October 22, 2011 5:26 PM
  • Thanks Vilmos, looking forward to hearing from you.

    I have a number of questions regarding the enums. Do you mind if I continue to add individual questions to this as we go? For example, I've noticed that arabic1Minus only includes the Arabic alphabet - not Farsi or Arabic-based alphabets (e.g. when the run is set to "fa-IR" it still displays only Arabic letters and not Persian ones, like چ). I have changed both my system and Office to Farsi and it still doesn't display Persian letters. Is this expected behavior?

    Monday, October 24, 2011 9:30 PM
  • Hi Okatu,

    Let’s concentrate on one problem, e.g. the first one after BTW in your original posting. You are saying that you have a document in three versions: the original one, one where the paragraph language is set to “ar-LB” and one where the run language is set to “ar-LB”. In all cases the numbering digits  are displayed according to the Windows language setting, the language setting in the document is ignored.

    I would like to analyze these documents. To send them as attachments, send them to "dochelp (at) microsoft (dot) com", and ask to forward it to me. Please give additional information including which versions of Windows and Word are you using, how the documents were created and whatever else you think could be useful.

    Thanks, Vilmos

    Friday, October 28, 2011 8:43 PM
  • Vilmos,

    I am simply providing examples of where the standard does not define anything. I realize it is because of Windows settings for Arabic/English because I disovered that on my own. You'll notice this is *not* the case for Hebrew, Thai or Hindi. But a definition of this does not exist in the ECMA/ISO documents.

    You can create the files I proposed above very easily. Just install a Persian keyboard and change to it. Also, ensure the DefaultLanguage of PowerPoint is Persian. Create a numbered list. Then install an Arabic keyboard and change the DefaultLanguage of the PPTX to Arabic. You'll see that 4, 5, and 6 (in Persian) do not change to Arabic. Now change everything to English. You'll see that all numbers change to latinized Arabic.

    What I'm asking for is a full definition of the numbering standard. I can continue to provide examples, but that still does not detail out what the standard expects. So let's go back to the original question "Where do the 20.1.10.60 ST_TextAutonumberScheme (Text Auto-number Schemes) lists come from?"

    Once you can provide that definition I may have more individual quesitons.

    Thanks.

     

     

     

    Friday, October 28, 2011 9:59 PM
  • Hi Okatu,

    Thank you for your patience, I’m still researching the issue.

    Vilmos

    Wednesday, November 2, 2011 4:15 PM
  • Vilmos,

    Checking in on this as it's been more that 2 weeks since I raised the question "Where do the 20.1.10.60 ST_TextAutonumberScheme (Text Auto-number Schemes) lists come from?". Any updates?

    Thursday, November 10, 2011 2:37 AM
  • Hi Okatu,

    No update, yet; I'm still researching the issue.

    Vilmos

    Tuesday, November 15, 2011 1:22 AM
  • Hi Okatu,

    Thank you for your patience. Coverage of auto number digits depending on host operating system language setting was not covered under our current released documentation. Based on your feedback the product group is looking at adding content for a future release of the documentation covered in the ISO 29500 Implementer notes.

    Please could you describe in details what kind of information would you like to see in the documentation. For example, how the rendering should depend on the “Use Native Digits” setting (Control Panel -> Region and Language -> Formats -> Additional Settings -> Numbers).

    Thanks, Vilmos

    Tuesday, December 6, 2011 8:31 PM
  • Thanks. I'd like to see as much detail as possible. The first is around the apparent inconsistency with Latinized Arabic numbers and actual Arabic numbering (i.e. to quote you: how the rendering should depend on the “Use Native Digits” setting (Control Panel -> Region and Language -> Formats -> Additional Settings -> Numbers))

    Secondarily, most numbering doesn't have anything to do with the OS.

    For example ea1ChtPeriod is always going to be traditional Chinese. But what isn't defined is how that is implemented. The description simply states "EA: Traditional Chinese w/ single-byte period". This does not tell us which numbers are Type A or Type C or, for that matter, what Type A or Type C actually are. Who defined Type A and C? Are they subject to change in the PRC or ROC? I just happen to know Chinese, so I figured it out on my own through trial and error in PowerPoint, but unless you know Chinese, you'll never be able to figure it out as an implementer.

    arabic1Minus is inconsistent when changing between RTL languages, such as Arabic/Dari to Farsi/Urdu. As mentioned above, 4, 5, 6 do not change when switching from to Arabic<->Perso-Arabic scripts. Please also see the example of Persian چ  when it comes to alphabets.

    In Roman numerals, such as romanLcParenR, what are the heuristics expected? God forbid someone has 1999 bullet points on a slide, but if they do, is it written MDCCCCLXXXXVIIII, MCMXCIX, or MIM?

    Is there a fall-back mechanism if a system doesn't have the Wingdings font for circleNumWdBlackPlain? Also it says in the description "11-arabic numbers" - are those 11 and up numbers dependent on the OS or...?

    circleNumDbPlain appears to be based off of the minor font based on which script is being used (+mn-lt, +mn-ea or +mn-cs), but I'm not 100% sure of that.

    arabic2Minus is Abjad. The definition of Abjad needs to be made clear as there are a few different ways of looking at Abjad. Numerology in Arabic-only Arabic letters?

    In Latin scripts, are values like alphaLcParenR expected to be 26 letters only in the order they appear as used in English or does this depend on which Latinized language is being used?

    Finally, is it expected that implementation results be the same across host operating systems? In other words - if I create a complex numbering example on the PC, is it expected to look exactly the same on a Mac?

    There are probably many more questions (I know nothing about Hindi or Thai and so I haven't tackled those values yet).

    Thursday, December 8, 2011 8:00 PM
  • Hi Okatu,

     

    Thank you again for bringing to our attention that the PowerPoint implementation details of the ST_TextAutonumberScheme from the Office Open XML File Formats, ECMA-376, is missing from the MS documentation. In the next version of the MS documentation this might be included.

     

    In the Office Open XML File Formats – Part 1, ECMA-376, 3<sup>rd</sup> Edition / June 2011, the ST_TextAutonumberScheme is defined in 20.1.10.60. The definition is general, doesn’t contain enough information which is necessary for a PowerPoint presentation developer and, as mentioned above, the accompanying MS documentation doesn’t have any implementation detail.
    I summarize the topics in this item, this will be filed as request for addition to the PowerPoint documentation, this is a non-exhaustive list of cases which should be covered.

     

    Does the OS, the paragraph / run language setting influence the autonumber display, especially when the language orientations are different.

     

    In the Chinese autonumberings, ae1Cht…, Type A and Type C are mentioned. These are not well known classifications, in the PowerPoint documentation they should be explained.

     

    There are no definitive rules for displaying Roman numerals, the rules used in PowerPoint should be documented (the current rules can be deducted by setting the start number, e.g. 1999 – MCMXCIX).

     

    How the circleNumWd… types behave if the Wingdings font is not available, is there any fallback mechanism? The circleNumWdWhitePlain type definition says 0 – 10 circle, is any way to display 0 circle?

     

    The alpha… types based on the 26 letter English alphabet. Is any way to use another Latin letter based alphabet or, e.g. the Greek letters?

     

    Please feel free to add any case, situation what you would like to see in the documentation.

     

    Thanks, Vilmos

    Friday, December 23, 2011 9:57 PM
  • This is great. There are more use-cases in my sample above that can be used as well. Thanks for following up on this.
    Monday, December 26, 2011 11:54 PM
  • Hi Okatu,

    Thank you again for bringing to our attention that the implementation details of the ST_TextAutonumberScheme should be documented in the Microsoft documentation. If you want to see some specific case to be documented, please let me know.  You can send your suggestion to “dochelp (at) microsoft (dot) com” too, indicating that it is for me.

    Thanks, Vilmos

    Thursday, December 29, 2011 10:54 PM