Difference between incomplete and incorrect
I'm writing a book in DocBook. It is the first time I have used oXygen to write in a schema that I did not create myself, and I am therefore relying a lot more than I normally do on the cues that the editor gives me. One of those cues are the validation errors. The problem is, they are all signaled with RED. That makes sense if you have entered something illegal, but most of the time they appear because you have begun to create a complex structure and that structure is currently incomplete. From an author's point of view, that's not an error, I just haven't finished yet. Where oXygen can generate the whole structure, it does, so no error is shown, but there are many cases where it can't because what is required is one of one or more options. In these cases, the structured does not become legal until the author adds additional elements. The thing is, RED means stop, but the author, in this case should continue. RED means you did something wrong, but in these cases, the author simply hasn't finished yet. I find the editor is constantly slowing me down by throwing up these errors that I have to analyse each time to determine if I need to go on or if I need to back up. When I design a schema, I try to avoid creating these kinds of structures, for just this reason, but sometimes they cannot be avoided. My question is, is there a way, and do folks think it is desirable, to present a different cue for incomplete structures than for erroneous ones. At minimum, use a different color, such as yellow, for incomplete. Ideally, find some way to indicate that the author is on the right track and needs to continue. Thoughts? Mark
Hi Mark, Thank you for your feedback! Jing has an option for Relax NG validation that we expose in Options->Preferences -- XML / XML Parser / RELAX NG [ ] Check feasibly valid that does not mark anymore as validation errors the missing elements or attributes - in this mode a document is considered feasibly valid if all it takes to make it full valid is to add content, elements or attributes to it. Such a validation mode is useful indeed during authoring. AFAIK there is no similar support for XML Schema or DTDs. However, we do intercept all errors reported by Xerces to update the location information - oXygen changes the location information for many errors from the end of a start tag to the actual location of an attribute or of an attribute value the error actually refers to - so it may be possible to identify errors that refer to missing required content and change them to warnings, thus getting yellow markers instead of red markers, or remove them completely and getting a similar behavior with the check feasibly valid option from the Relax NG validation. This probably needs to be controlled by an option, because a validate action that returns only warnings is considered a successful validation, so the user should be aware that he decided to ignore some type of errors. Best Regards, George -- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com On 4/26/13 10:15 PM, Mark Baker wrote:
I’m writing a book in DocBook. It is the first time I have used oXygen to write in a schema that I did not create myself, and I am therefore relying a lot more than I normally do on the cues that the editor gives me.
One of those cues are the validation errors. The problem is, they are all signaled with RED. That makes sense if you have entered something illegal, but most of the time they appear because you have begun to create a complex structure and that structure is currently incomplete. From an author’s point of view, that’s not an error, I just haven’t finished yet.
Where oXygen can generate the whole structure, it does, so no error is shown, but there are many cases where it can’t because what is required is one of one or more options. In these cases, the structured does not become legal until the author adds additional elements. The thing is, RED means stop, but the author, in this case should continue. RED means you did something wrong, but in these cases, the author simply hasn’t finished yet.
I find the editor is constantly slowing me down by throwing up these errors that I have to analyse each time to determine if I need to go on or if I need to back up. When I design a schema, I try to avoid creating these kinds of structures, for just this reason, but sometimes they cannot be avoided.
My question is, is there a way, and do folks think it is desirable, to present a different cue for incomplete structures than for erroneous ones. At minimum, use a different color, such as yellow, for incomplete. Ideally, find some way to indicate that the author is on the right track and needs to continue.
Thoughts?
Mark
_______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
On 04/26/2013 02:15 PM, Mark Baker wrote: ...
My question is, is there a way, and do folks think it is desirable, to present a different cue for incomplete structures than for erroneous ones. At minimum, use a different color, such as yellow, for incomplete. Ideally, find some way to indicate that the author is on the right track and needs to continue.
Thoughts?
Hi Mark, That's an interesting idea and what follows doesn't address it at all. Instead, this is my approach to the problem: In cases where there's more than one possible next element often one of the elements is the most commonly used. For example, if you enter an itemizedlist, it could take a title or info as a child, followed by a listitem, but usually you just want a listitem. The listitem in turn could take many block elements as a child, but 99% of the time you want a para. If you're writing in author view, then you can add Actions that enter a typical collection of markup instead of just the element and then REMOVE the element from the content completion list. For example, you could have an itemizedlist action that inserts <itemizedlist xmlns="http://docbook.org/ns/docbook"><listitem><para/></listitem></itemizedlist> at the caret position and then bind that to the Content Completion menu as well as any contextual menus or toolbars. If you remove the itemizedlist element from the content completion stuff, then when you press enter and type itemizedlist, you'll always get your action and not the bare element. Now it's up to the user to know that she can modify that template and add a title to the itemizedlist or use something other than a para inside the listitem, but for usability, I think this approach wins. It would be really cool if that stuff were configured out of the box for supported schemas at least for the most important elements. It's not the kind of thing that each user in your org should configure for themselves. Instead all the config should be done and distributed to them in a framework. Regards, David
Hi David, In this context we have recorded an improvement to provide access to what elements are available at the current position so that you can disable the action in contexts when that element cannot be inserted. Best Regards, George -- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com On 4/28/13 7:34 AM, David Cramer wrote:
On 04/26/2013 02:15 PM, Mark Baker wrote: ...
My question is, is there a way, and do folks think it is desirable, to present a different cue for incomplete structures than for erroneous ones. At minimum, use a different color, such as yellow, for incomplete. Ideally, find some way to indicate that the author is on the right track and needs to continue.
Thoughts?
Hi Mark,
That's an interesting idea and what follows doesn't address it at all. Instead, this is my approach to the problem:
In cases where there's more than one possible next element often one of the elements is the most commonly used. For example, if you enter an itemizedlist, it could take a title or info as a child, followed by a listitem, but usually you just want a listitem. The listitem in turn could take many block elements as a child, but 99% of the time you want a para.
If you're writing in author view, then you can add Actions that enter a typical collection of markup instead of just the element and then REMOVE the element from the content completion list.
For example, you could have an itemizedlist action that inserts <itemizedlist xmlns="http://docbook.org/ns/docbook"><listitem><para/></listitem></itemizedlist> at the caret position and then bind that to the Content Completion menu as well as any contextual menus or toolbars. If you remove the itemizedlist element from the content completion stuff, then when you press enter and type itemizedlist, you'll always get your action and not the bare element.
Now it's up to the user to know that she can modify that template and add a title to the itemizedlist or use something other than a para inside the listitem, but for usability, I think this approach wins.
It would be really cool if that stuff were configured out of the box for supported schemas at least for the most important elements. It's not the kind of thing that each user in your org should configure for themselves. Instead all the config should be done and distributed to them in a framework.
Regards, David
_______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
On 04/29/2013 07:11 AM, Oxygen XML Editor Support wrote:
In this context we have recorded an improvement to provide access to what elements are available at the current position so that you can disable the action in contexts when that element cannot be inserted.
Thanks, that would help a lot! David
I am interested in converting Relax NG to XSD 1.1. Any progress/prospects for this? I might be interested in assisting with the Trang development for this. Regards, Tara
Hi Tara, I have not seen any work in that direction. As you probably know Trang is available on Google code at http://code.google.com/p/jing-trang/ James Clark is the project owner, I am a committer on the project as I made the NVDL implementation, initially as a Jing fork called oNVDL and then I merged that into Jing. I think that everyone will welcome a conversion to XML Schema 1.1 if you can contribute that to the project. Best Regards, George -- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com On 4/30/13 8:28 PM, Tara Athan wrote:
I am interested in converting Relax NG to XSD 1.1. Any progress/prospects for this?
I might be interested in assisting with the Trang development for this.
Regards, Tara _______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
It would greatly help the XML community, if the oXygen and other companies' NVDL, Schematron and other schemas, which were not written in XSD be converted to XSD1.1. XHTML5 cannot be properly implemented until it has an XSD1.1 schema. I use XSD1.1 because it has many advantages over XSD1.0, such as: 1) restriction of complexTypes (templates or generics) with the parent in one schema and the child in a second schema. 2) The ability with an openContent element to interleave the elements of two schemas with the elements from each schema being invisible to the other. This means that with the advent of an XSD1.1 schema that describes xhtml5, we can use standard elements like <p>, <h1>, <h2> etch that can be formatted with standard CSS style sheets and included in EPUBs. My goal is to have both the forms and reports that contain elements from my XML schemas (www.cytometryml.org) be xhtml5 pages. 3) The availability of assertions, permit the creation of constraints between datatypes and replaces the need for Schematron. Unfortunately, the one XSD schema for XHTML5 from www.xmlmind.com presently does not validate. Prior to the development of a complete translation program, the creation of a partial translator that works where there is a one to one relationship between XSD1.1 and the other schema language could still be used to expedite the process. Bob Leif -----Original Message----- From: oxygen-user-bounces@oxygenxml.com [mailto:oxygen-user-bounces@oxygenxml.com] On Behalf Of George Cristian Bina Sent: Wednesday, May 01, 2013 8:32 AM To: Tara Athan Cc: oxygen-user@oxygenxml.com Subject: Re: [oXygen-user] relax ng conversion to xsd 1.1 Hi Tara, I have not seen any work in that direction. As you probably know Trang is available on Google code at http://code.google.com/p/jing-trang/ James Clark is the project owner, I am a committer on the project as I made the NVDL implementation, initially as a Jing fork called oNVDL and then I merged that into Jing. I think that everyone will welcome a conversion to XML Schema 1.1 if you can contribute that to the project. Best Regards, George -- George Cristian Bina <oXygen/> XML Editor, Schema Editor and XSLT Editor/Debugger http://www.oxygenxml.com On 4/30/13 8:28 PM, Tara Athan wrote:
I am interested in converting Relax NG to XSD 1.1. Any progress/prospects for this?
I might be interested in assisting with the Trang development for this.
Regards, Tara _______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
_______________________________________________ oXygen-user mailing list oXygen-user@oxygenxml.com http://www.oxygenxml.com/mailman/listinfo/oxygen-user
participants (6)
-
David Cramer -
George Cristian Bina -
Mark Baker -
Oxygen XML Editor Support -
Robert Leif -
Tara Athan