Douglas W Philips wrote: > > > I'm writing, instead of Chris, because as we attempted our own > work-around for the Microsoft limit we ran into what we think might be > an XML Mind bug. > > Since part of our 40-character problem came from using very descriptive > id attributes on our concept elements, we tried to follow this advice: > http://tech.groups.yahoo.com/group/dita-users/message/14257 > and give each of our concepts a short, fixed, id attribute. >
Eliot Kimber wrote: --- In the case where every topic is the root of its own document, the topic ID can be the same *for every topic*. To make this point I have standard practice of using the value "topicid" for the IDs of all my root topics. There is absolutely no need to generate unique topic IDs for document-root topics as a matter of standard practice. --- We understand this and we agree that the DITA standard allows to do this (though we'll *never* do this for our own documents, because this really looks and feels like a dirty trick). However there is a rationale behind XMLmind DITA Converter requiring all topic IDs to be distinct. See below. > Unfortunately that did not work, ditac was unable to find most of our > cross references. > > I have created a sample set of files that illustrates this issue. > We're using version 1.1 (since 1.2 just came out) installed in the > standard location. > > ditac.bbt is the batch file used to invoke ditac and provoke what we > think the bug is. > (I had to change the extension in order to be allowed to send it as an > attachment to this email, it should be a .bat file) > > In this sample example, we do get a warning which we should not get. > The warning can be made to go away by changing the concept id attribute > in one of the files (and the corresponding xref in the other file. > > Please let me know if you need any additional detail and/or if you have > any trouble with the attachments. > XMLmind DITA Converter works internally with IDs, and not with URLs. That's why it requires all topic IDs to be distinct. This was always the case but, as of version 1.2, it now reports a *fatal* *error* when this is not the case. We don't think it's a bug. For us, having distinct topic IDs is a way to create easy-to-read, predictable, cross-references in the generated document. Example: your DITA document contains: * <xref href="C2.xml#turnip2/greens"/> * <xref href="C1.xml#turnip/root"/> You now convert your DITA document to PDF. Having unique topic IDS allows us to convert the links to: * <fo:basic-link internal-destination="turnip2__greens"/> * <fo:basic-link internal-destination="turnip__root"/> Without this requirement, XMLmind DITA Converter needs to create synthetic, unique, link IDs and targets in the PDF. Generating synthetic, unique, link IDs and targets may be OK for PDF, RTF, .docx, etc, but is not so nice in XHTML, HTML Help, Eclipse Help, etc. In the case of XHTML, HTML Help, Eclipse Help, etc, that is, for online help systems, link targets must be easy-to-read and predictable. Even if this was a bug, it cannot be fixed without a major rewrite of XMLmind DITA Converter. We currently do not plan to rewrite XMLmind DITA Converter in the near future. We'll of course document this idiosyncrasy. We apologize for the inconvenience. We would understand very well if you decided to stop using our product.
