about summary refs log tree commit diff
path: root/services
AgeCommit message (Collapse)AuthorFilesLines
2019-12-20 r/224 feat(cgit-taz): Patch cgit to construct monorepo URLsVincent Ambo2-1/+51
Effectively dropping /depot/ from all URLs.
2019-12-20 r/223 fix(cgit-taz): Drop /cgit.cgi prefix from vhostVincent Ambo1-1/+1
2019-12-20 r/222 feat(cgit-taz): Display README.md for repositoryVincent Ambo1-0/+1
2019-12-20 r/216 fix(sync-gcsr): Fix broken log statementVincent Ambo1-1/+1
2019-12-20 r/214 fix(cgit-taz): Depot checkout is no longer a bare repositoryVincent Ambo1-1/+1
To be compatible with sync-gcsr, the `.git` folder must be referenced directly as there is an actual worktree on the filesystem now.
2019-12-20 r/213 fix(sync-gcsr): Use local worktree & pull changes into itVincent Ambo1-13/+25
Just fetching the remote of a bare repository does not update local tracking branches, which means that changes do not become visible in cgit.
2019-12-20 r/212 chore(cgit-taz): Remove git synchronisation from startup scriptVincent Ambo1-7/+0
This will be moved into a sidecar container using sync-gcsr that updates the repo periodically.
2019-12-20 r/211 feat(services/sync-gcsr): Add git synchronisation helperVincent Ambo2-0/+90
Adds a tiny program that keeps a checkout of a remote git repository in sync with a folder on the local filesystem. This is going to be used to mirror the GCSR repository onto a local disk for cgit serving.
2019-12-18 r/196 fix(cgit-taz): Force gcloud to use specific CA bundleVincent Ambo1-0/+3
It doesn't like reading the bundle from /etc/ssl apparently.
2019-12-18 r/194 refactor(cgit-taz): Use Google Cloud SDK to authenticate repo accessVincent Ambo1-13/+4
Fighting SSH's peculiarities around how hard it tries to not be scriptable is tiresome.
2019-12-18 r/193 fix(cgit-taz): Create users required by SSH in the containerVincent Ambo1-0/+4
This might not actually work because SSH will have issues with key permissions, presumably.
2019-12-18 r/192 fix(cgit-taz): Pin SSH into image via NixVincent Ambo1-0/+1
2019-12-18 r/191 fix(cgit-taz): Clone depot via SSHVincent Ambo1-4/+2
Keys are provided in a Kubernetes secret.
2019-12-18 r/190 fix(cgit-taz): Explicitly configure CA certs for gitVincent Ambo1-1/+3
2019-12-18 r/189 feat(cgit-taz): Clone depot on container launchVincent Ambo1-1/+9
This makes for a pretty static server for now, but we'll get there.
2019-12-18 r/184 feat(cgit-taz): Apply cgit configuration by patching thttpdVincent Ambo1-22/+29
This is just an optimisation because it compiles faster v0v
2019-12-18 r/180 feat(services/cgit-taz): Check in initial cgit configurationVincent Ambo2-0/+64
This is currently just a test to serve examples from my local machine, nothing more fancy than that.
2019-12-16 r/159 chore(third_party): Bump naerskVincent Ambo1-2/+1
2019-12-09 r/108 fix(third_party): Expose various other missing dependenciesVincent Ambo1-2/+2
2019-12-09 r/103 fix: Amend import paths to use whitelisted nixpkgsVincent Ambo1-1/+1
2019-12-09 r/100 refactor: Introduce new layout with nixpkgs in third_partyVincent Ambo3-3/+4
This is not the final layout yet, but makes it so that my top-level attribute set is no longer overlaid into nixpkgs itself. This is useful for other people who are importing my monorepo.
2019-11-15 feat(nix): Filter projects that should be built by CIVincent Ambo1-0/+1
Instead of specifying CI projects manually, this filters them to move the CI configuration into the derivations `meta` attributes.
2019-11-15 chore(nix): Move files around to conform to new read-tree layoutVincent Ambo3-15/+22
Broadly speaking, the following things are included: * there is now a uniform `args` struct that is passed to all derivations, package headers have been changed appropriately * overrides are now loaded from a separate `override` folder just using read-tree.nix * third-party packages have moved into the `third_party` attribute set
2019-10-26 feat(nixcon-demo): Add CLI mode for NixCon demoVincent Ambo1-2/+13
Maybe a bit more sane than trying to do a network based setup.
2019-10-26 feat(services): Add nixcon-demo service with simple web serverVincent Ambo4-0/+921
2019-09-22 r/89 chore(gemma): Mark gemma as broken (again)Vincent Ambo1-1/+3
Something broke in one of the Lisp dependencies (again).
2019-09-22 r/87 chore(tazblog): Remove more unneeded dependenciesVincent Ambo1-4/+4
2019-09-03 fix(tazblog): Explicitly allow substitutes for the blogVincent Ambo1-6/+10
Not entirely sure which part of the setup set this to 'false', but this is potentially the key for why tazblog ends up being rebuilt all the time.
2019-09-02 chore(tazblog): Clean up unneeded filesVincent Ambo5-76/+0
2019-09-02 chore(gemma): Delete old image build configurationVincent Ambo1-19/+0
2019-09-02 refactor(blog): Use callPackage to import derivationVincent Ambo1-3/+3
2019-09-02 r/70 fix(gemma): Fix build process in NixVincent Ambo2-51/+28
2019-08-25 r/65 chore(tazblog): Clean up unused dependenciesVincent Ambo2-28/+17
2019-08-25 r/64 refactor(tazblog): Implement HLint lints in all filesVincent Ambo5-15/+13
2019-08-25 r/63 chore(tazblog): Remove i18n featuresVincent Ambo7-164/+77
The blog has been English only for a few years. Old entries that survived the migration to DNS will still be accessible.
2019-08-25 r/62 chore(tazblog): Remove 'read more' featureVincent Ambo3-15/+7
2019-08-25 r/60 chore(tazblog): Format source files with ormoluVincent Ambo5-151/+187
Ormolu's formatting is quite annoying (it uses a lot of unnecessary vertical space and doesn't align elements), but I can't be bothered to do manual formatting - especially because whatever formatting haskell-mode in Emacs produces seems to depend on an opaque state machine or something.
2019-08-23 r/57 fix(gemma): Almost fix Gemma build by porting an old ElmVincent Ambo1-7/+5
This pulls in an old version of Elm from NixOS 17.09 which can still build the Elm code in Gemma. However, the Common Lisp build is now broken in some other way.
2019-08-23 r/55 refactor(tazblog): Move blog configuration to envvarsVincent Ambo2-22/+24
The port and resource directory are now specified via environment variables and a wrapper script is created by Nix that sets the resource path and so on correctly.
2019-08-23 r/54 fix(tazblog): Remove debug traceVincent Ambo1-2/+1
2019-08-22 r/51 chore(blog): Minor changes to integrate DNS based backendVincent Ambo5-21/+18
2019-08-22 r/50 feat(tazblog): Implement entry fetching from DNSVincent Ambo1-20/+119
Not all error cases are properly handled yet, stay tuned.
2019-08-21 r/49 feat(tazblog): Add a Nix shell expression that includes GHC with depsVincent Ambo1-0/+11
2019-08-21 r/48 refactor(tazblog): Directly instantiate Resolver when launchingVincent Ambo2-7/+25
Caching behaviour is tied to the resolver.
2019-08-21 r/47 chore(tazblog): Introduce dependency on 'dns'Vincent Ambo2-2/+3
2019-08-20 r/46 chore(tazblog): Introduce dependency on 'cache'Vincent Ambo3-19/+23
2019-08-20 r/44 chore(tazblog): Replace BlogDB with stubs for DNS-based storageVincent Ambo10-560/+115
Removes acid-state specific code and the former BlogDB module, in its stead the new BlogStorage module contains stubs for the functions that will be filled in with DNS-based storage. This code is unformatted and will not currently serve a working blog.
2019-08-15 r/31 chore: Remove remaining Bazel-related filesVincent Ambo1-86/+0
2019-08-15 r/29 chore(services/gemma): Mark package as brokenVincent Ambo1-0/+5
2019-07-04 r/27 feat(tazblog): Add Bazel build configurationVincent Ambo1-0/+86
Adds a build configuration using rules_haskell and rules_nixpkgs with all dependencies coming from the centrally configured GHC in the project root.