The new emerging NLE for GNU/Linux
State Draft
Date Mi 08 Dez 2010 11:32:32 CET
Proposed by Ichthyostega <>

The Lumiera website is assumed to accumulate a lot of content. Thus we need to care about making that content accessible, to help finding the relevant topics and to keep the overall structure intact. This RfC is to collect, discuss and agree upon the guidelines and requirements.


Issues to care


The page hierarchy becomes at least 5 levels deep, likely even deeper. When reading a page, the current subtree leading down to this page should be right at hand; especially access to the siblings and the parent’s siblings is important. For re-accessing content, it is necessary to be able to drill down to an known location (“within the design docs, detailing the application, I need the configuration section”)
→ we need an auto generated navigation and an embedded menu tree widget in the web pages.


There should be an easy way to categorise individual pages by keyword(s) and an automatically generated indexing by tags, possibly with an per tag overview page.


The usual site search. It should include the contents of the issue tracker. Even today such a scoped search is valuable and even necessary for working with the informations collected within the Lumiera project


Each relevant page needs to be reachable. There are some additional pages and especially subdirectories which should not be linked into the website navigation. Moreover, all (internal) links on the pages should be valid.
→ this could be addressed by a sanity checker script

Usage situations

(a) working on content

Working on content should be readily accessible for everyone. One time contributions are especially encouraged. This leads to the following usage scenario:

A contributor has some informations to share or wants to do some additions or modifications. (S)he locates somehow the place where relevant informations are stored, adds some text, possibly adds a new page or splits another page in two.

Note: no awareness of the issues of navigation can be assumed. The occasional contributor won’t notice any concern which isn’t right at hand.

(b) maintaining a subsystem

Some person(s) will be responsible for a subsystem or some segment of the informations on the website. This responsibility is content centric. It might include frequent rearranging, regrouping and reordering of pages to accommodate the increasing scope of informations.

Note: while here some awareness of website organisational issues can be assumed, any requirement to care for external organisational issues is a burden and distracts from the actual work to be done — thus it is likely to be short circuited or postponed “for later”. Note especially, reorganising content in a subsection must not incur the burden of re-doing the same reorganisation steps mirrored in some central navigation configuration or table of contents. (this is a knock out criterion)

(c) maintaining the website

The website maintainer is responsible for the overall sanity of the website, without being familiar with all details of ongoing work in some part or section of the information. Another concern here is the outward perception of the website, which might incur changes on the top level navigation or some rewording of overview pages.

Note: this kind of work is rather unrewarding. There is the danger of collisions with the work of the subsystem maintainer

Conclusion: Requirements for any navigation solution

  • ability to to pick up a nested page structure

  • ability to cope with any additions and changes in the lower levels automatically, without help by the user

  • ability to override:

    • not including some subdirectories

    • including links-to-external at arbitrary positions

optional/additional features

The following features would be handy, but can be considered optional

  • ability to change the displayed title of a page in the navigation

  • ability to control the ordering of pages in the navigation

  • complete manual override of the visible content of a specific subdirectory


  • define markup for the various features (✔ done)

  • get a technical solution for the menu to work (✔ done)

  • write a script to traverse contents and generate the menu (✔ done)

  • test and integrate it into the website (✔ done)


Maintaining the navigation within a website beyond just some pages is a daunting task. When frequent rearrangements of pages are to be expected, the only viable solution is to automate this task. Moreover, there needs to be a secondary path to each page, asside of the direct links which might or might not be provided. A automatically generated navigation menu separate of the actual page content helps to address these issues.


State → Draft

A Menu generator script based on these principles is deployed and working since a looooong time now. We still need to build the tagging facility though. This is covered by another RfC.


So 07 Okt 2012 07:30:17 CEST <>