Differences

This shows you the differences between two versions of the page.

--- design:producing-output [2013/08/28 21:40] – Paul added some changes from Alice paul
+++ design:producing-output [2013/10/30 16:34] (current) – rsewikiadmin
@@ Line 1: / Line 1: @@
 This page describes how RFC output is produced in the the current system and proposes a method for it to be produced in the future.
+====== Pre-production ======
+  - Internet drafts are generated with a variety of tools.
+  - Some of these tools generate XML2RFC input, which is then used to generate text.
+  - Some of the tools generate text directly.
+  - When ready to become an RFC, the text version is sent to the RFC Editor, along with either the corresponding XML2RFC input or the corresponding nroff input
 ====== Current Production ======
-**//Alice needs to check whether the statements in this section are true.//**
+  - If the input to the RFC Editor includes XML, the document is edited directly.
-  - If the input to the RFC Editor is XML, the document is edited directly.
   - When editing is complete, xml2rfc v2 is used to generate text (unless errors are such that xml2rfc v1 is used instead).  AUTH48 state is initiated using the text output. Once approvals are complete, xml2rfc v2/v1 is used to generate nroff.
   - The nroff is is edited to fix the placement of page breaks and any remaining quirks.
   - The nroff is used to produce paginated text.
   - The paginated text is used to produce PDF that looks just like the paginated text.
+  - The RFC Editor publishes the text and PDF files.
   - Separately, the Tools Team uses the paginated text to produce HTML that looks just like the paginated text but has additional links.
@@ Line 21: / Line 27: @@
 ====== Proposed Future Production ======
-  - If the input to the RFC Editor is XML, the document is edited directly.
+  - A tool (tentatively called "RFCToolv3") is used to validate I-Ds against the new DTD
-  - When editing is complete, RFCToolv3 is used to produce HTML that is round-trippable with the XML; this will most likely be done using XSLT processing.
+    - if the I-D is in text format, then an initial step of creating an XML file is necessary
-  - RFCToolv3 uses the HTML to produce unpaginated text.
+    - Tool will need to be able to combine multiple XML files (i.e., images, references)
-  - RFCToolv3 uses the unpaginated text to produce paginated text.
+  - If the input to the RFC Editor is XML, the document is edited directly. If the input is a different format, the RFC Editor first converts it into XML using tools and (probably) hand-editing.
-  - RFCToolv3 uses the HTML to produce PDF format 1 that looks much like the HTML (including having live links), and also has relevant headers and footers.
+  - When editing is complete,  AUTH48 state is initiated. The RFC Editor and the authors interact, and the interactions are reflected in the XML.
-  - RFCToolv3 uses the paginated text to produce PDF format 2, hopefully having live links. (Both PDF formats are produced for US Letter and A4 page sizes.)
+  - Once approvals are complete, the RFC Editor finalizes the XML and uses RFCToolv3 to create all the non-normative formats (see [[design:formats|Thoughts on Non-Canonical Formats]]).
-  - RFCToolv3 uses the HTML to produce EPUB.
+  - The RFC Editor publishes the XML and all of the other formats.
-If the input to the RFC Editor is nroff, the RFC Editor first turns it into XML, then the steps above
-are used.
-If the input to the RFC Editor is plain text, the RFC Editor first turns it into XML, then the steps above
-are used.
+Note: Authors may run this tool prior to submission in order to gain understanding on what their final RFC is expected to look like (modulo AUTH48 changes).  This includes
+  * creation of an XML file (if submitted file is TXT)
+  * addition of anchors and tags
+  * creation of all publication outputs (HTML, TXT, EPUB, PDF)
+    * make it optional which outputs are produced by authors?