skalarproduktraum / phd-thesis Goto Github PK

View Code? Open in Web Editor NEW

0.0 8.0 0.0 158 MB

TeX 99.52% Makefile 0.33% Shell 0.16%

phd-thesis's Introduction

ulrik's official phd thesis repository

Requirements

pandoc
pandoc-citeproc
pandoc-latex-environment for using LaTeX environments in Pandoc Markdown documents
bibtool, available for most Linux distributions via their package manager, and on Homebrew
a working TeX distribution that includes XeLaTeX and BiBTeX, e.g. TeXLive

Building

Put chapters in wanted order in chapters.list, can include both TeX and Markdown files.
Bibliography is in bibliography.bib, and will be filtered by bibtool to get rid of fields that often cause issues (e.g. abstract, annote, url,...)
Run make, rest happens magically.

phd-thesis's People

Contributors

Watchers

phd-thesis's Issues

Fix bibliography

The bibliography has some errors, mainly missing publisher, etc.

Warning--I didn't find a database entry for "hdf52017"
Warning--empty publisher in Ashtiani:2010dw
Warning--empty publisher in Bayer:2002ds
Warning--empty journal in Baştuğ:201645e
Warning--empty journal in Brugues:2012fx
Warning--empty journal in Bryson:19942f4
Warning--empty year in Castellina:vg
Warning--empty year in Castellina:vg
Warning--empty publisher in Catmull:1974cf
Warning--empty year in Cheeseman:ia
Warning--empty year in Cheeseman:ia
Warning--empty journal in Clift:2018ncs
Warning--empty school in Crassin:2011uo
Warning--empty journal in Darken:1997odt
Warning--empty publisher in Duchowski:2017ii
Warning--empty journal in Erazo:2015d6c
Warning--empty journal in Fernandes:2016cvrs
Warning--empty publisher in Gray:1878ahb
Warning--empty journal in Guignard:bi
Warning--empty year in Guignard:bi
Warning--empty year in Guignard:bi
Warning--empty publisher in Niemz:2007laa
Warning--empty booktitle in Hatscher:2017bi
Warning--empty school in Huisken:5iIUZiJj
Warning--empty year in Huisken:5iIUZiJj
Warning--empty year in Huisken:5iIUZiJj
Warning--empty publisher in TaylorII:2001bq
Warning--empty year in jacob:1993
Warning--empty year in jacob:1993
Warning--empty publisher in Jacob:1990hz
Warning--empty publisher in Jerald:201667e
Warning--empty journal in John:1995cp
Warning--empty booktitle in Kassner:2014kh
Warning--empty booktitle in Klamka:2015ka
Warning--empty journal in Laine:EffectiveSVO
Warning--empty year in Laine:EffectiveSVO
Warning--empty year in Laine:EffectiveSVO
Warning--empty publisher in Lattner:2004vw
Warning--empty journal in Mania:2004ps
Warning--empty school in Oswald:2010pr
Warning--empty booktitle in Pfeuffer:2017jk
Warning--empty booktitle in Pohl:2016fy
Warning--empty publisher in Raskin:2000thi
Warning--empty journal in Reipschläger:2018945
Warning--empty journal in Rice:2014341
Warning--empty publisher in Snowden:2012wu
Warning--empty journal in Theis:2018tw
Warning--empty booktitle in Wang:2016fz
Warning--empty year in Workshop:uc
Warning--empty year in Workshop:uc
Warning--empty publisher in Zhang:2008ex

Ch: Design and Architecture

Needs something about the overall design concept, can't jump directly into nodes.

Nodes

A diagram of a scenery rendered scene with annotations (e.g. corresponding to your cell, nucleus, ER example), even if you just hand code primitive geometries geometries.

The Hub

A diagram/flowchart of the architecture might be worthwhile.

Renderers

Would it be worth discussing the how of hypothetically including of a future ray tracing renderer or sth? This might help demonstrate that you have anticipated future issues, and could even be useful for some future researcher building on scenery.

Rendering with X

I wonder how the initialization routines would look side by side, especially if you aligned where steps match. Would this help emphasize how different OpenGL and Vulkan are? If so, that might help support an argument that scenery has solved key complexity issues of handling multiple rendering backends.

Overall comments

Essentially all of the content can stay. The organization is confusing, and this is a bad place to have confusing organization (since part of an architecture is the organization of elements). It might be the layout since the only way hierarchical structure is conveyed is through different font size in the headings.

External hardware is should be more clearly separated.

Everything up to UBOs feels like it is close to the right organization (with an added overview at the beginning), but after that it feels like a laundry list of features or a sort of appendix of features. I think the content about those features should present, but it needs to be organized and I need to understand why I care about these features.

I see some potential subsections (and numbering might really help you here, because font size of headings is definitely not doing the trick):
Overall architecture (nodes, hub, input, renderer-in-brief)
Rendering (renderers, pipelines, push mode, UBOs, workflow, shaders, instancing, volumes)
External hardware (the ones already listed)

After writing ^^^ I'm now thinking about

Part Scenery:

scenery
Design and architecture (nodes, hub, input, renderer-in-brief, more on scene graphs?)
Rendering (renderers, pipelines, push mode, UBOs, workflow, shaders, instancing, volumes)
External hardware (the ones already listed)
Developing with scenery (REPL, scripts, walk through scenery tests similarly to the following case studies but briefly [maybe at most 1.5 pages incl screenshot per scenery test where you dont show annotated code but you explain the concept of the test and what features it is testing])

The scenery chapter can then focus more on scenery's guts than ij stuff