[om] Re: Bugs in OpenMath standard draft 3

David Carlisle davidc at nag.co.uk
Thu Nov 20 13:51:03 CET 2003


Thanks for your comments.
Except as noted below I've fixed these in our CVS source, and so they
should appear in the next public draft (hopefully in a day or so).


   1) In the contents list the links F.1 - F.3 appear to be broken. in fact
   looking at the xhtml source, the links have href="#". I'm not sure what
   that means as far as the DocBook source is concerned, but surely something
   equaly trivial!


   2) Chapter 3 paragraph 1 states:

   "We first do so at an informal level (Section 3.2) and next by means of an
   abstract grammar description (Section 3.1)"

   It strikes me that 3.2 comes after 3.1 not before it, I guess that either
   these sections should have their order shifted or maybe this paragraph
   should be changed.

Section intro changed.

   3) there is a spurious ',' in section 3.1.3 iii)


   4) in section 3.2 the paragraph on Symbols is missing a space during

Seems to have been fixed already in our CVS sources.

   5) In section 4.1.1 there is a typo in paragraph 1 during
   "...it is generated from the using the compact syntax given below."

This paragraph has been re-worded.

   6) I am not a Relax NG schema guru, but from my reading I don't see that
   the Relax NG schema for OpenMath 2 allows for indexed 'OMV's as described
   elsewhere in the text.

Fixed. OMV now allows a 0-or-1 element children (Although personally i
think we may need to look at this again, I think we may want to restrict
this to (say) literal integers or back it out, as allowing arbitrary
content here raises some difficult issues about determining identity
of two openmath fragments, which previously has never been required.

   7) In section 4.1.2 the subsection on Symbols frequently refers to the CD
   'transc' "which is part of the collection maintained by the OpenMath
   Society" It has long been 'transc1'.


   8) also in section 4.1.2 the subsection on Floating-point numbers gives a
   regular expression for dec attributes, maybe for hex attributes it should


No, the hex attribute encodes the bytes of the IEEE double so there is
no - or e, if we gave a regexp (perhaps we should) it would be [0-9A-F]+

   9) In section on Acyclicity constraints, there is a sentence
   terminating with:

   "which is not admitted by the structure of OpenMath objects described
   in section 3."

   It would be helpfull if section 3 was linked to this (please).


   10) In section 4.2.2 during the sentence:
   "are encoded in two ways depending on whether The string is encoded in utf-16 ..."


   "...of Unicode are used. as a byte (if ..."

I changed "." to "," and left "as" as "as"

   11) in section 4.2.3 4'th paragraph the encoding for external refereces is
   incorrect; it says:
   "[30] for (short) external references, [30+128] for long external

   I believe this should be:

   "[31] for (short) external references, [31+128] for long external

Already fixed in CVS (Michael I assume)

   12) I think that the example in Figure 4.3 has errors:
   surely the tag in byte 21 should have argument (the next byte) of 01 with
   description "to the second shared object" whilst the tag in byte 24 should
   have argument with description "to the first shared object"

Will check

   13) There is a typo in section 4.2.4 paragraph 3:
   "analogously for or [30+128] {_}."

Fixed ("or" deleted)

   14) In section 4.2.5 I think the example is incorrect. It doesn't include
   the cdbase attributes. I think that the referenes are incorrectly encoded.
   In fact I don't see them in the XML. surely this should correspond?

Clearly something wrong, will check.

   15) In section 5.3.2 the element CDCDBASE is refered to

fixed (CDBASE)

   16) In section the text in the title of figure 5.1 needs a space


   17) In section surely the first paragraph should be changed to
   refer to the Relax NG specification in Figure 5.1


   18) In section the description of CDVersion is missing a space

   "                           ... version of the Content Dictionary isto
   be taken as member of the CDGroup. ..."


   19) In Chapter 7, second sentence:
   "The things      are addressed"

This para has bee rewritten in the cvs sources.

   20) throughout Appendix A. The macro &OM; has not been expanded

I don't see this, perhaps it's already fixed?

   21) The Meta CD in Appendix A is totally out of date (it even has support
   for a 'Presentation' element) which I never new existed!) and should
   surely be replaced with the most resent?

I'll modify the build process to pull in the "current" meta cd (although
we don't currently have a meta cd fully describing OM2, but this should
be fixed before the draft is finalised. Thanks for flagging this. 

   22) In Appendix B the same comment as point 6) applies here (please ignore
   if this is my miss-understanding!)

See comment above.

   all the best,




This e-mail has been scanned for all viruses by Star Internet. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
om at openmath.org  -  general discussion on OpenMath
Post public announcements to om-announce at openmath.org
Automatic list maintenance software at majordomo at openmath.org
Mail om-owner at openmath.org for assistance with any problems

More information about the Om mailing list