Here is a quote from the Drupal.org Roadmap:
... Each year Drupal.org and its sub-sites serve millions of visitors, and thousands of active contributors. It is one of the largest continuously operating Drupal sites in the world. And because of that, it has 15 years of legacy content and features. At this scale, it's impossible to make real improvements to Drupal.org without a prioritized roadmap, focusing on a few, high impact features at a time. ...
To understand what the D8 "upgrade challenge" would be about, have a look at these drupal.org projects (quotes are from the links):
A project with issue tracker that you can use to report spam, broken links, or user account problems on Drupal.org website.
A project with issue tracker for Drupal.org non-documentation content organization and moderation. This is a place for issues about marketplace listings (services, training and books), case studies, front page promotion requests, planet drupal feeds, etc.
This is the main project for all Drupal documentation. To learn more about helping with documentation efforts, start with the Contribute to documentation section of the handbook.
Customizations used on drupal.org itself. It is not meant to be useful to other sites, except as an educational example of the kinds of modifications you can make via a site-specific module.
More information about the state of planned work on Drupal.org can be found on the Drupal.org Roadmap.
Bluecheese is the redesigned Drupal.org theme ... This theme is only for use on official Drupal.org sites. Do not use this for businesses, local groups and other sites, which should have their own brand and identity. Questions and comments about use of this theme should go to the Drupal Association ...
An issue tracker for everything related to the Drupal.org servers. This includes the Apache and MySQL installation, the Mailman mailing lists, the Git repositories, and the various Drupal installations on the drupal.org domain.
This project is a place for issues to be filed or questions to be asked about the Drupal.org testbots and all the projects related to them. Since there are so many d.o projects involved it seems that most people just need a one-stop shop for issues. They can be moved out from here.
This project has two purposes.
- It is like the drupal.org customizations project to hold site specific code.
- It is like the Infrastructure and Webmasters queues but is specific to groups.drupal.org. So, where people used to enter things into those queues with a Component of groups.drupal.org they can now enter them here.
About the upgrade challenge
With the above in mind, I think there are dozens of reasons which Drupal.org is not using D8 yet, here are just a few of them (in a random order):
- This site is mission critical for quite a few of us, so failure is not an option.
- There must be quite a few contributed modules that are needed, but for which the D8 release isn't there yet. Not sure if they use/need all of them, but I bet these are some of those modules: Project, Organic groups, Migrate (D7 to D8), something for charting, maybe Rules, Flag, etc ... True, modules like Views, etc are in D8 core now, and you can get pretty far already only using Drupal core, but I doubt drupal.org doesn't need some of the modules like the ones above.
- You need the manpower for it, with appropriate skills (D8 is a bit different from D7, right?).
- You need infrastructure to run D8 sites (PHP version, memory, etc).
- I bet there is "some" custom code involved for running Drupal.org, and upgrading custom code has never been easy.
- And there is a need for things like QA-testing, or documentation, etc.
- From my quote above it seems there are other priorities, so upgrading to D8 seems to be like not a priority.
- I bet there are other things that have a higher priority, e.g. "The forums need improving".
- Even if none of the above were blocking it all, think about the purely accounting part of it (I can't imagine the yearly write-offs for the recent D7 upgrade are finished already).
For all these reasons, I think at this very moment the best we can do is "Hope for George to do it" (the one from the keynote from Dries at DrupalCon Amsterdam 2014 when the very first beta1 version of D8 was announced).
PS: One never realizes what has been done, only what remains to be done.