aboutsummaryrefslogtreecommitdiff
path: root/README
diff options
context:
space:
mode:
authorKaiden Fey <kookie@spacekookie.de>2020-10-25 02:40:33 +0100
committerKaiden Fey <kookie@spacekookie.de>2020-10-25 02:40:33 +0100
commit392444d21101ce7b637f2e5a385490605f93ccf1 (patch)
tree75247f7a3b49a17c70a31106b04e09f72848f98a /README
parenta0dca8186bdef76e09e9c388d7f85839e85ce8db (diff)
Big project update
Diffstat (limited to 'README')
-rw-r--r--README77
1 files changed, 3 insertions, 74 deletions
diff --git a/README b/README
index ed972cd..2834983 100644
--- a/README
+++ b/README
@@ -3,82 +3,11 @@
. ( o O)
\_ ` _, _
-.___'.) ( ,-'
- '-.O.'-..-.. octopus git web frontend
+ '-.O.'-..-.. octopus git mono-repo explorer
./\/\/ | \_.-._
;
._/
-A git web frontend that wants to hug your code.
-
-
-Why?
-----
-
-This is a very good first question, and one that I think is important
-to answer before getting more into the project. Whenever I brought up
-this project during the creation of it, most people would react with
-"have you tried...", followed by some git web software, such as
-gitlab, gittea, and many many more. But there is a reason why I stuck
-with octopus and the design ideas I had for it.
-
-Fundamentally it's about decentralisation. The internet is a pretty
-big place (allegedly), but a lot of the services that people use are
-very centralised by a single company or even server. If this company
-or server goes away, so does valuable knowledge. There are many ways
-to create more decentralised systems, and one aspect of this for me,
-is code repos.
-
-Git is by it's nature decentralised, meaning that it doesn't require
-an "upstream" or canonical server to function. Theoretically you
-could use git to exchange code patches with people without the
-internet. Yet, a lot of people's perception of git is one that is
-dictated by the workflows on github and gitlab: centralised into a
-single service. Many git web frontends mirror this workflow, because
-after all, it is what people want and love.
-
-However, it has some flaws (which would take too long to elaborate
-here), and for octopus I had something else in mind. Instead of
-replicating the same mistakes, I took much more inspiration from cgit,
-which is used and loved by many today. octopus is a re-imagining of
-cgit, trying to improve the UX and maintainability where possible,
-adding new features, but mostly staying true to it's core: a simple
-git web service, that doesn't lock you into a vendor, or server.
-
-
-Configuration
--------------
-
-octopus is easy to configure and run, even on systems that don't give
-you priviledged system access. The main server binary is configured
-primarily with a few environment variables, and an app config.
-
-+ OCTOPUS_CONFIG should contain the path of the main configuration
-
-+ OCTOPUS_SSL_KEY can optionally be set to the path of a certificate key
-
-+ OCTOPUS_SSL_CERT can optionally be set to the path of a certificate
-
-
-The main configuration file is written in yaml, and outlines things
-like the application domain, repo paths, and added modules. Because
-octopus is vory modular, you can only depend on certain features.
-
-
----
-app_path: git.octopus.example
-port: 8080
-handle_ssl: false
-cache_path: /var/cache/octopus
-repo_path: /var/lib/octopus/repos
-repo_discovery: false # Disables automatic config loading from repo_path
- # and instead let's you set a static set of repos
- # in the section below
-repos:
- - octopus:
- description: "🐙 It's a water animal"
- category: "/"
- - libkookie:
- description: "My personal nix expressions"
- category: "/nix"
-
+A visualiser for mono-repos, with per project pages, histories, and
+committer data. \ No newline at end of file