more on word support

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

more on word support

Bruce D'Arcus
So based on back-and-forth with the product manager responsible for
the new bib support in Word 2007*:

1) they won't support footnote/endnote citations in v1
2) seems (?) they don't support first/subsequent distinctions in author-year
3) they think it perfectly fine to have styles implemented in raw XSLT
(they don't appear interested in using CSL or a CSL-like abstraction)

So this tells us where we can differentiate OOo. There's going to be a
lot of frustration with their default support, particularly among the
historians.

Also, on 3, it should possible to swap in a citeproc-like solution,
and so get support for CSL in Word through the back door. Hmm ...
wonder if I should try to productize citeproc for the Word market?
;-)

Bruce

* see comments at
<http://blogs.msdn.com/joe_friend/archive/2006/07/13/664960.aspx>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: more on word support

David Wilson-7
I read through the exchange and I think you put your position very well  and
the response was rather defensively or evasive.

David

On Saturday 15 July 2006 10:52 am, Bruce D'Arcus wrote:

> So based on back-and-forth with the product manager responsible for
> the new bib support in Word 2007*:
>
> 1) they won't support footnote/endnote citations in v1
> 2) seems (?) they don't support first/subsequent distinctions in
> author-year 3) they think it perfectly fine to have styles implemented in
> raw XSLT (they don't appear interested in using CSL or a CSL-like
> abstraction)
>
> So this tells us where we can differentiate OOo. There's going to be a
> lot of frustration with their default support, particularly among the
> historians.
>
> Also, on 3, it should possible to swap in a citeproc-like solution,
> and so get support for CSL in Word through the back door. Hmm ...
> wonder if I should try to productize citeproc for the Word market?
> ;-)
>
> Bruce
>
> * see comments at
> <http://blogs.msdn.com/joe_friend/archive/2006/07/13/664960.aspx>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]

--
-------------------
David N. Wilson
Co-Project Lead for the Bibliographic
OpenOffice Project
http://bibliographic.openoffice.org

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: more on word support

Bruce D'Arcus

On Jul 16, 2006, at 5:38 AM, David Wilson wrote:

> I read through the exchange and I think you put your position very
> well  and
> the response was rather defensively or evasive.

I think it's worth paying close attention to what they've done not just
for interoperability's sake, but also because they seem to have very
similar design goals, as well as constraints. To quote Jennifer:

> Please keep in mind that this is a v1 feature, and our first goal was
> to make the tools extremely useful for high school and college
> students. We designed this feature to be a platform so that anyone –
> us, in later versions; any 3rd party like EndNote; or you(!) could
> build tools on top of ours. That’s why everything in this feature is
> XML-based. Certainly there are more things we could have delivered
> with more time and person-power, but we tried first to make sure our
> platform was solid.

And I think with some caveats, they will have met these goals. I am
particularly intrigued by their no-local-database approach, where the
editing forms are only editing XML data embedded in the file package.

This is something we need to seriously consider for OOo (though we can
do a better job).

Bruce
---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: more on word support

David Wilson-7
On Sunday 16 July 2006 9:47 pm, Bruce D'Arcus wrote:

>
> And I think with some caveats, they will have met these goals. I am
> particularly intrigued by their no-local-database approach, where the
> editing forms are only editing XML data embedded in the file package.
>
> This is something we need to seriously consider for OOo (though we can
> do a better job).

I have been think about this question, what do we need a local SQL database
for? And how does it need to be integrated with the Bibliographic
application?

SQL databases are good for very large quantities of data, and would  essential
if the the volume of data was greater that could could be held in list in
memory.(which is now a very big list)  This would not be the case for the
citations to a single document. SQL databases are also good for sharing data
and update facilities amongst many users.

An SQL database is not necessary to store the local citations in documents
even temporarily as they are worked on. However Bib users may want to store
collections of citations, in some manner, and SQL databases are probably a
mechanism we may want to support. If only to provide a browse and 'insert
into document' function, as would do for internet / remote database search
and insert.

What the "no-local-database approach, where the editing forms are only editing
XML data embedded in the file package" implies though is OOoBib would NOT be
providing a mechanism or maintaining your collection of citations. That is,
looking through the collection and spotting a error and fixing it.
Perhaps suggest we advise people to use one of the many third party tools for
that purpose. We can certainly do this in the early stages as we develop the
application.

One of the difficulties with a building a close connection between the xml
local storage and a SQL database, is that the xml data will support formatted
text, included embedded document objects (mathematical formula etc).
SQL database are ascii based. Of course a one way to deal with this is to have
a ascii version of each potentially formatted field and the formatted
version. (you need the plain ascii field for searching)

If we want to store and maintain bibliographic collections it would be easier
if the collections were stored in a xml database such as eXist, and the field
conversion problems disappear.

I am thinking that the xml database for local storage of citation collections
would be the a good choice as the some the local editing tools would work in
the same way on the in-document citations and the xml database citations.
This is least work option for storage of collections of citations. With a SQL
database we would need to build two sets of editing tools - xml and SQL.

I do not have any firm ideas about this, these are just my musings.

David

--
-------------------
David N. Wilson
Co-Project Lead for the Bibliographic
OpenOffice Project
http://bibliographic.openoffice.org

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: more on word support

Bruce D'Arcus
On 7/16/06, David Wilson <[hidden email]> wrote:

> I have been think about this question, what do we need a local SQL database
> for? And how does it need to be integrated with the Bibliographic
> application?

And come to think of it, do we need to have source editing within OOo
at all? One could imagine, for example, that editing a source would
take one back to one's outboard database.

> An SQL database is not necessary to store the local citations in documents
> even temporarily as they are worked on. However Bib users may want to store
> collections of citations, in some manner, and SQL databases are probably a
> mechanism we may want to support. If only to provide a browse and 'insert
> into document' function, as would do for internet / remote database search
> and insert.

That's the thing, in the way we've approached this, there really is no
distinction between "internet/remote" and "my references"; they're
accessed in the same way. And given projects like SmartFox, Connotea,
CiteUlike, Refworks, etc., is tthere really any doubt that the future
of reference management is the web?

> One of the difficulties with a building a close connection between the xml
> local storage and a SQL database, is that the xml data will support formatted
> text, included embedded document objects (mathematical formula etc).
> SQL database are ascii based. Of course a one way to deal with this is to have
> a ascii version of each potentially formatted field and the formatted
> version. (you need the plain ascii field for searching)

SQL database can store pretty much anything, so there'd be no problem,
for example, storing unicode-encoded math titles, or even adding some
minimal markup? Maybe there are some tricks for searching it?

> If we want to store and maintain bibliographic collections it would be easier
> if the collections were stored in a xml database such as eXist, and the field
> conversion problems disappear.

I'm not really sure an xml db is needed. I doubt know about anybody
else, but my data in RDF/XML is maybe 1MB.

> I am thinking that the xml database for local storage of citation collections
> would be the a good choice as the some the local editing tools would work in
> the same way on the in-document citations and the xml database citations.
> This is least work option for storage of collections of citations. With a SQL
> database we would need to build two sets of editing tools - xml and SQL.

Well, I think the questions we're left with are:

1) must OOoBib include reference management facilitties (actual
database support), or can we leave it to third-parties?

2)  if we do, SQL, flat XML, or XML DB?

3)  if we don't, how do we allow users to edit the source for a
citation? Edit the local embedded XML though an OOo form (probably
xforms) and sync it to the outboard database, or just link their and
let that application handle it?

My current thinking is:

1)  no; it's be nice to have, but not essential

2) SQL or flat XML

3)  link; no OOo form support for this

> I do not have any firm ideas about this, these are just my musings.

Right, me neither.

Bruce

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: more on word support

ptsefton
In reply to this post by David Wilson-7
David is proposing that the scope of the OOo bibliogrpahy tools be kept to
an absolute minimum. I would support this.

I would like to see:

1. An external reference database / research tool that could be used with
both Word and Writer. It should be able to store not jsut metadata but
full-text articles,  notes, webpage snapshots & other supporting data.
I am still not convinced that storing reference details with a document is a
good idea; the biggest reason is that even if interop can be sorted out this
will still only work with Word 2007 and a future version of Writer.

2. An external CSL editor that can export compiled XSLT for Word - and maybe
Python code  for Writer. (I don't agree with Bruce that this needs to be
built in to Word or that the size of XSLT files is a problem since they
would be auto-generated from a CSL file).

3.  Interoperable citation markers that will allow cross-word-processor
teams to work together.


Microsoft are clearly not going to be swayed by lobbying, and it's not clear
to me how much Sun will do to push this stuff through. Why not concentrate
on building stand-alone tools that work with the current installed base of
word processing software? That is, build a better EndNote.

(That's what I hope my team will be doing over the coming year - in
alignment with the work going on here)

On 7/17/06, David Wilson <[hidden email]> wrote:

>
> On Sunday 16 July 2006 9:47 pm, Bruce D'Arcus wrote:
>
> >
> > And I think with some caveats, they will have met these goals. I am
> > particularly intrigued by their no-local-database approach, where the
> > editing forms are only editing XML data embedded in the file package.
> >
> > This is something we need to seriously consider for OOo (though we can
> > do a better job).
>
> I have been think about this question, what do we need a local SQL
> database
> for? And how does it need to be integrated with the Bibliographic
> application?
>
> SQL databases are good for very large quantities of data, and
> would  essential
> if the the volume of data was greater that could could be held in list in
> memory.(which is now a very big list)  This would not be the case for the
> citations to a single document. SQL databases are also good for sharing
> data
> and update facilities amongst many users.
>
> An SQL database is not necessary to store the local citations in documents
>
> even temporarily as they are worked on. However Bib users may want to
> store
> collections of citations, in some manner, and SQL databases are probably a
> mechanism we may want to support. If only to provide a browse and 'insert
> into document' function, as would do for internet / remote database search
> and insert.
>
> What the "no-local-database approach, where the editing forms are only
> editing
> XML data embedded in the file package" implies though is OOoBib would NOT
> be
> providing a mechanism or maintaining your collection of citations. That
> is,
> looking through the collection and spotting a error and fixing it.
> Perhaps suggest we advise people to use one of the many third party tools
> for
> that purpose. We can certainly do this in the early stages as we develop
> the
> application.
>
> One of the difficulties with a building a close connection between the xml
> local storage and a SQL database, is that the xml data will support
> formatted
> text, included embedded document objects (mathematical formula etc).
> SQL database are ascii based. Of course a one way to deal with this is to
> have
> a ascii version of each potentially formatted field and the formatted
> version. (you need the plain ascii field for searching)
>
> If we want to store and maintain bibliographic collections it would be
> easier
> if the collections were stored in a xml database such as eXist, and the
> field
> conversion problems disappear.
>
> I am thinking that the xml database for local storage of citation
> collections
> would be the a good choice as the some the local editing tools would work
> in
> the same way on the in-document citations and the xml database citations.
> This is least work option for storage of collections of citations. With a
> SQL
> database we would need to build two sets of editing tools - xml and SQL.
>
> I do not have any firm ideas about this, these are just my musings.
>
> David
>
> --
> -------------------
> David N. Wilson
> Co-Project Lead for the Bibliographic
> OpenOffice Project
> http://bibliographic.openoffice.org
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>


--
Peter (pt) Sefton
Toowoomba 4350
Queensland, Australia
Phone: +61 4 1032 6955
Web: http://ptsefton.com
Email: [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: more on word support

Bruce D'Arcus

On Jul 16, 2006, at 10:23 PM, pt wrote:

> 1. An external reference database / research tool that could be used
> with
> both Word and Writer. It should be able to store not jsut metadata but
> full-text articles,  notes, webpage snapshots & other supporting data.

Right. I think this is slowly being taken care of by others. The
Firefox Scholar plug-in (aka SmartFox), for example, is likely to be
really good, and will be free (GPL). They also have interest in serving
as data sources for word-processors.

> I am still not convinced that storing reference details with a
> document is a
> good idea; the biggest reason is that even if interop can be sorted
> out this
> will still only work with Word 2007 and a future version of Writer.

Why? I'm pretty sure there must be a way to store that data in earlier
versions of the apps. Apps like Endnote, for example, have been
embedding data in Word docs for years. It's just never been
standardized.

In our case, in any case, we're not dependent on embedded data; it's a
convenience. I think Microsoft, by contrast, is dependent on it.

The difference is that I'll be pushing to ensure that the identifiers
for citations are uris, where we recommend best practices to make it
easy to reconstitute data as needed. MS, by contrast, is using dumb
local natural language ids for linking. E.g. they always assume the
data is embedded.

> 2. An external CSL editor that can export compiled XSLT for Word - and
> maybe
> Python code  for Writer. (I don't agree with Bruce that this needs to
> be
> built in to Word or that the size of XSLT files is a problem since they
> would be auto-generated from a CSL file).

The code to generate static XSLTs of this sort is not easy to write,
and it makes everything more, not less, complex, doesn't it?  For
example, how do styles get stored, and how would a user add a style?
What happens if you need to update a style?

CSL is simple enough that it's really not hard to write a parser for
it. The SmartFox guys will be doing just that using Javascript (E4X to
be specific, which has XML extensions).

> 3.  Interoperable citation markers that will allow cross-word-processor
> teams to work together.
>
> Microsoft are clearly not going to be swayed by lobbying, and it's not
> clear
> to me how much Sun will do to push this stuff through.

MS won't be swayed to do the right thing aside from what's in their own
interest. I am trying to point out to them where those overlap (for
example, they don't support footnotes within the citation fields; am
making sure that's not a file format restriction, which would be bad
for them too), and am optimistic interop will be fairly good.

FWIW, MS is using standard field support to implement the coding for
citations.

As for OOo and Sun, I think we need at minimum to get the new citation
field implemented and exposed so that projects like your's and
Matthias' can easily interest with it.

> Why not concentrate on building stand-alone tools that work with the
> current installed base of
> word processing software? That is, build a better EndNote.

I agree.

> (That's what I hope my team will be doing over the coming year - in
> alignment with the work going on here)

Cool. You should hook up with the SmartFox guys. They'll be releasing a
beta sometime in the next few months, complete with support for CSL and
the biblio schema I've been working on ;-).

Bruce

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]