Debian Science Umbrella for scientific packages or dustbin for - - PowerPoint PPT Presentation

▶

Jan 24, 2024 330 likes •1.19k views

Debian Science Umbrella for scientific packages or dustbin for scientific code Andreas Tille DebConf 17 Montreal, 6th August 2017 Andreas Tille ( DebConf 17 ) Debian Science Montreal, 6th August 2017 Overview Current status 1 Better

SLIDE 1

Debian Science

Umbrella for scientific packages or dustbin for scientific code Andreas Tille

DebConf 17

Montreal, 6th August 2017

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 2

Overview

1

Current status

2

Better science support in Blends

3

Problems and suggested solutions

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 3

Overview

1

Current status

2

Better science support in Blends

3

Problems and suggested solutions

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 4

Overview

1

Current status

2

Better science support in Blends

3

Problems and suggested solutions

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 5

History of Debian Science

DebConf 5: "Using Debian for science research" Helen Faulkner (video) Started simply with a mailing list Competing packaging teams evolved due to lack of coordination Today merged to Debian Science packaging team Debian Science is using Blends framework since 2008

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 6

History of Debian Science

DebConf 5: "Using Debian for science research" Helen Faulkner (video) Started simply with a mailing list Competing packaging teams evolved due to lack of coordination Today merged to Debian Science packaging team Debian Science is using Blends framework since 2008

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 7

History of Debian Science

DebConf 5: "Using Debian for science research" Helen Faulkner (video) Started simply with a mailing list Competing packaging teams evolved due to lack of coordination Today merged to Debian Science packaging team Debian Science is using Blends framework since 2008

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 8

History of Debian Science

DebConf 5: "Using Debian for science research" Helen Faulkner (video) Started simply with a mailing list Competing packaging teams evolved due to lack of coordination Today merged to Debian Science packaging team Debian Science is using Blends framework since 2008

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 9

History of Debian Science

DebConf 5: "Using Debian for science research" Helen Faulkner (video) Started simply with a mailing list Competing packaging teams evolved due to lack of coordination Today merged to Debian Science packaging team Debian Science is using Blends framework since 2008

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 10

Debian Pure Blends

Specific adaptations to certain workfields Reside completely inside Debian (no derivative) Form packaging team around specific topic Advertising and QA of packages in web sentinel

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 11

Debian Pure Blends

Specific adaptations to certain workfields Reside completely inside Debian (no derivative) Form packaging team around specific topic Advertising and QA of packages in web sentinel

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 12

Debian Pure Blends

Specific adaptations to certain workfields Reside completely inside Debian (no derivative) Form packaging team around specific topic Advertising and QA of packages in web sentinel

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 13

Debian Pure Blends

Specific adaptations to certain workfields Reside completely inside Debian (no derivative) Form packaging team around specific topic Advertising and QA of packages in web sentinel

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 14

Umbrella for potetial specific Blends

"Science" in itself is to wide area Enable spin-offs of more specific Blends Sustainable Blend needs a critical mass of

packages maintainers doing the actual work

Blends framework should help to create that critical mass

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 15

Umbrella for potetial specific Blends

"Science" in itself is to wide area Enable spin-offs of more specific Blends Sustainable Blend needs a critical mass of

packages maintainers doing the actual work

Blends framework should help to create that critical mass

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 16

Umbrella for potetial specific Blends

"Science" in itself is to wide area Enable spin-offs of more specific Blends Sustainable Blend needs a critical mass of

packages maintainers doing the actual work

Blends framework should help to create that critical mass

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 17

Umbrella for potetial specific Blends

"Science" in itself is to wide area Enable spin-offs of more specific Blends Sustainable Blend needs a critical mass of

packages maintainers doing the actual work

Blends framework should help to create that critical mass

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 18

Umbrella for potetial specific Blends

"Science" in itself is to wide area Enable spin-offs of more specific Blends Sustainable Blend needs a critical mass of

packages maintainers doing the actual work

Blends framework should help to create that critical mass

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 19

Umbrella for potetial specific Blends

"Science" in itself is to wide area Enable spin-offs of more specific Blends Sustainable Blend needs a critical mass of

packages maintainers doing the actual work

Blends framework should help to create that critical mass

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 20

Specific scientific Blends

Debian Med Strong focus on Microbiology Debian GIS Geographical Information Systems DebiChem Chemistry Debian Astro Astronomy

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 21

Specific scientific Blends

Debian Med Strong focus on Microbiology Debian GIS Geographical Information Systems DebiChem Chemistry Debian Astro Astronomy

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 22

Specific scientific Blends

Debian Med Strong focus on Microbiology Debian GIS Geographical Information Systems DebiChem Chemistry Debian Astro Astronomy

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 23

Specific scientific Blends

Debian Med Strong focus on Microbiology Debian GIS Geographical Information Systems DebiChem Chemistry Debian Astro Astronomy

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 24

Why not more?

Idea of Debian Pure Blends needs some time to penetrate Advantages are not widely known In my talk @ DebConf 13 Asheesh Laroia (in video at about minute 38): "We should try hard to run around asking users and developers: Is there a topic you care about create a Blend today." Nobody likes to do the grunt work Idea for of the umbrella for spin-offs should be better advertised

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 25

Why not more?

Idea of Debian Pure Blends needs some time to penetrate Advantages are not widely known In my talk @ DebConf 13 Asheesh Laroia (in video at about minute 38): "We should try hard to run around asking users and developers: Is there a topic you care about create a Blend today." Nobody likes to do the grunt work Idea for of the umbrella for spin-offs should be better advertised

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 26

Why not more?

Idea of Debian Pure Blends needs some time to penetrate Advantages are not widely known In my talk @ DebConf 13 Asheesh Laroia (in video at about minute 38): "We should try hard to run around asking users and developers: Is there a topic you care about create a Blend today." Nobody likes to do the grunt work Idea for of the umbrella for spin-offs should be better advertised

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 27

Why not more?

Idea of Debian Pure Blends needs some time to penetrate Advantages are not widely known In my talk @ DebConf 13 Asheesh Laroia (in video at about minute 38): "We should try hard to run around asking users and developers: Is there a topic you care about create a Blend today." Nobody likes to do the grunt work Idea for of the umbrella for spin-offs should be better advertised

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 28

Why not more?

Idea of Debian Pure Blends needs some time to penetrate Advantages are not widely known In my talk @ DebConf 13 Asheesh Laroia (in video at about minute 38): "We should try hard to run around asking users and developers: Is there a topic you care about create a Blend today." Nobody likes to do the grunt work Idea for of the umbrella for spin-offs should be better advertised

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 29

Debian Science Blend

Blends provide straightforward access to package pool No need for searching interesting packages Engagement for packaging more scientific Free Software Not yet implemented but possible

Further add-ons like user menus, preconfiguration, . . . Installer with adapted package selection

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 30

Debian Science Blend

Blends provide straightforward access to package pool No need for searching interesting packages Engagement for packaging more scientific Free Software Not yet implemented but possible

Further add-ons like user menus, preconfiguration, . . . Installer with adapted package selection

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 31

Debian Science Blend

Blends provide straightforward access to package pool No need for searching interesting packages Engagement for packaging more scientific Free Software Not yet implemented but possible

Further add-ons like user menus, preconfiguration, . . . Installer with adapted package selection

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 32

Debian Science Blend

Blends provide straightforward access to package pool No need for searching interesting packages Engagement for packaging more scientific Free Software Not yet implemented but possible

Further add-ons like user menus, preconfiguration, . . . Installer with adapted package selection

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 33

Debian Science Blend

Blends provide straightforward access to package pool No need for searching interesting packages Engagement for packaging more scientific Free Software Not yet implemented but possible

Further add-ons like user menus, preconfiguration, . . . Installer with adapted package selection

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 34

Debian Science Blend

Blends provide straightforward access to package pool No need for searching interesting packages Engagement for packaging more scientific Free Software Not yet implemented but possible

Further add-ons like user menus, preconfiguration, . . . Installer with adapted package selection

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 35

Web sentinel providing

Tasks web pages (featuring citations!) Specific bugs overview Status in releases including backports and Ubuntu (=Thermometer) Maintainer stats from teammetrics

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 36

Web sentinel providing

Tasks web pages (featuring citations!) Specific bugs overview Status in releases including backports and Ubuntu (=Thermometer) Maintainer stats from teammetrics

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 37

Web sentinel providing

Tasks web pages (featuring citations!) Specific bugs overview Status in releases including backports and Ubuntu (=Thermometer) Maintainer stats from teammetrics

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 38

Web sentinel providing

Tasks web pages (featuring citations!) Specific bugs overview Status in releases including backports and Ubuntu (=Thermometer) Maintainer stats from teammetrics

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 39

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 40

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 41

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 42

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 43

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 44

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 45

Rise attractivity for upstream

The difference between a single maintainer and a Blend is like approaching upstream as

a. "random" person or
b. member of Debian Science

Advertise upstream publications Link to registration Provide popcon stats ➜ Giving credit to upstream

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 46

Science specific metapackages

Brain-computer interface (science-bci) Biology (science-biology) Chemistry (science-chemistry) Economics (science-economics) Electronics (science-electronics) Electrophysiology (science-electrophysiology) Engineering (science-engineering) Financial engineering (science-financial) Geography (science-geography) Geometry (science-geometry) High Energy Physics (science-highenergy-physics) Linguistics (science-linguistics)

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 47

Science specific metapackages (continued)

Logic (science-logic) Machine Learning (science-machine-learning) Mathematics (science-mathematics) Meteorology (science-meteorology) Nanoscale Physics (science-nanoscale-physics) Cognitive Neuroscience (neuroscience-cognitive) Modeling of neural systems (science-neuroscience-modeling) Numerical Computation (science-numericalcomputation) Physics (science-physics) Psychophysics (science-psychophysics) Robotics (science-robotics) Social (science-social)

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 48

Common science utilities

Data acquisition (science-dataacquisition) Device (science-devices) Distributed Computing (science-distributedcomputing) Image analysis (science-imageanalysis) Presentation (science-presentation) Simulations (science-simulations) Statistics (science-statistics) Tools (science-tools) Typesetting (science-typesetting) Viewing (science-viewing) Workflow (science-workflow)

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 49

Development

Data acquisition development (science-dataacquisition-dev) Engineering Development (science-engineering-dev) High Energy Physics Development (science-highenergy-physics-dev) Image analysis development (imageanalysis-dev) Mathematics development (science-mathematics-dev) Meteorology development (science-meteorology-dev) Nanoscale Physics Development (science-nanoscale-physics-dev) Physics Development (science-physics-dev) Robotics Development (science-robotics-dev) Viewing Development (science-viewing-dev)

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 50

Developers not aware of Debian Science

Single maintainers ITPs of scientific software ➜ Read ITPs and redirect to Debian Science team Single maintainers orphaned packages featuring RC bugs ➜ Ping maintainer, may be team hijack package Developers don’t care about tasks ➜ UDD query what packages are not mentioned in tasks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 51

Developers not aware of Debian Science

Single maintainers ITPs of scientific software ➜ Read ITPs and redirect to Debian Science team Single maintainers orphaned packages featuring RC bugs ➜ Ping maintainer, may be team hijack package Developers don’t care about tasks ➜ UDD query what packages are not mentioned in tasks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 52

Developers not aware of Debian Science

Single maintainers ITPs of scientific software ➜ Read ITPs and redirect to Debian Science team Single maintainers orphaned packages featuring RC bugs ➜ Ping maintainer, may be team hijack package Developers don’t care about tasks ➜ UDD query what packages are not mentioned in tasks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 53

Developers not aware of Debian Science

Single maintainers ITPs of scientific software ➜ Read ITPs and redirect to Debian Science team Single maintainers orphaned packages featuring RC bugs ➜ Ping maintainer, may be team hijack package Developers don’t care about tasks ➜ UDD query what packages are not mentioned in tasks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 54

Developers not aware of Debian Science

Single maintainers ITPs of scientific software ➜ Read ITPs and redirect to Debian Science team Single maintainers orphaned packages featuring RC bugs ➜ Ping maintainer, may be team hijack package Developers don’t care about tasks ➜ UDD query what packages are not mentioned in tasks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 55

Developers not aware of Debian Science

Single maintainers ITPs of scientific software ➜ Read ITPs and redirect to Debian Science team Single maintainers orphaned packages featuring RC bugs ➜ Ping maintainer, may be team hijack package Developers don’t care about tasks ➜ UDD query what packages are not mentioned in tasks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 56

Bad maintainer per package relation

1 2 3 4 5 6 7 8 9 11 100 200 300 400 Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 57

Compare the graph with the one from pkg-perl team

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 19 200 400 600 800 Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 58

Short term contribution of scientists

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try hard to get more than one Uploader

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 59

Short term contribution of scientists

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try hard to get more than one Uploader

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 60

Short term contribution of scientists

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try hard to get more than one Uploader

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 61

Short term contribution of scientists

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try hard to get more than one Uploader

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 62

Team orphaned packages

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try to salvage un(der)-maintained packages ➜ Team orphaned packages are easier to update ➜ Team hijacks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 63

Team orphaned packages

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try to salvage un(der)-maintained packages ➜ Team orphaned packages are easier to update ➜ Team hijacks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 64

Team orphaned packages

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try to salvage un(der)-maintained packages ➜ Team orphaned packages are easier to update ➜ Team hijacks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 65

Team orphaned packages

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try to salvage un(der)-maintained packages ➜ Team orphaned packages are easier to update ➜ Team hijacks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 66

Team orphaned packages

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try to salvage un(der)-maintained packages ➜ Team orphaned packages are easier to update ➜ Team hijacks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 67

Team orphaned packages

Scientific work is driven by projects with limited time frame Quite specific software frequently packaged by single maintainer (example: paw+cernlib) Software frequently hard to package (old libraries, FORTRAN, etc.) ➜ Try to salvage un(der)-maintained packages ➜ Team orphaned packages are easier to update ➜ Team hijacks

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 68

Strengthening team by sponsoring

Sponsering of Blends (SoB) Blends concept remains widely unknown amongst newcomers (but also amongst DDs) Newcomers might desperately seek for sponsors and simply do not know how to find one Kill two birds with one stone: Get the package sponsored after proving that you understood the Blends techniques

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 69

Strengthening team by sponsoring

Sponsering of Blends (SoB) Blends concept remains widely unknown amongst newcomers (but also amongst DDs) Newcomers might desperately seek for sponsors and simply do not know how to find one Kill two birds with one stone: Get the package sponsored after proving that you understood the Blends techniques

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 70

Strengthening team by sponsoring

Sponsering of Blends (SoB) Blends concept remains widely unknown amongst newcomers (but also amongst DDs) Newcomers might desperately seek for sponsors and simply do not know how to find one Kill two birds with one stone: Get the package sponsored after proving that you understood the Blends techniques

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 71

Strengthening team by sponsoring

Sponsering of Blends (SoB) Blends concept remains widely unknown amongst newcomers (but also amongst DDs) Newcomers might desperately seek for sponsors and simply do not know how to find one Kill two birds with one stone: Get the package sponsored after proving that you understood the Blends techniques

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 72

SoB results

SoB was heavily used by Debian GIS team - now sponsees are DDs/DMs themselves It helped to drastically reducing waiting time for very active Debian Sciene members Watching Wiki page and take mostly immediate action Lurking on relevant mailing lists as well on debian-mentors to catch maintainers in need of a sponsor If you notice a candidate for SoB please redirect them to me

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 73

SoB results

SoB was heavily used by Debian GIS team - now sponsees are DDs/DMs themselves It helped to drastically reducing waiting time for very active Debian Sciene members Watching Wiki page and take mostly immediate action Lurking on relevant mailing lists as well on debian-mentors to catch maintainers in need of a sponsor If you notice a candidate for SoB please redirect them to me

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 74

SoB results

SoB was heavily used by Debian GIS team - now sponsees are DDs/DMs themselves It helped to drastically reducing waiting time for very active Debian Sciene members Watching Wiki page and take mostly immediate action Lurking on relevant mailing lists as well on debian-mentors to catch maintainers in need of a sponsor If you notice a candidate for SoB please redirect them to me

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 75

SoB results

SoB was heavily used by Debian GIS team - now sponsees are DDs/DMs themselves It helped to drastically reducing waiting time for very active Debian Sciene members Watching Wiki page and take mostly immediate action Lurking on relevant mailing lists as well on debian-mentors to catch maintainers in need of a sponsor If you notice a candidate for SoB please redirect them to me

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 76

SoB results

SoB was heavily used by Debian GIS team - now sponsees are DDs/DMs themselves It helped to drastically reducing waiting time for very active Debian Sciene members Watching Wiki page and take mostly immediate action Lurking on relevant mailing lists as well on debian-mentors to catch maintainers in need of a sponsor If you notice a candidate for SoB please redirect them to me

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 77

Is Debian Science doing bad in QA?

Debian Science team was blamed about not caring about bugs High number of packages leaves wrong impression Team policy permits anybody to do team uploads ➜ If you are really bored by a bug in Debian Science simply use dch -team and fix it. Barrier is lower than NMU

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 78

Is Debian Science doing bad in QA?

Debian Science team was blamed about not caring about bugs High number of packages leaves wrong impression Team policy permits anybody to do team uploads ➜ If you are really bored by a bug in Debian Science simply use dch -team and fix it. Barrier is lower than NMU

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 79

Is Debian Science doing bad in QA?

Debian Science team was blamed about not caring about bugs High number of packages leaves wrong impression Team policy permits anybody to do team uploads ➜ If you are really bored by a bug in Debian Science simply use dch -team and fix it. Barrier is lower than NMU

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 80

Is Debian Science doing bad in QA?

Debian Science team was blamed about not caring about bugs High number of packages leaves wrong impression Team policy permits anybody to do team uploads ➜ If you are really bored by a bug in Debian Science simply use dch -team and fix it. Barrier is lower than NMU

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 81

Is Debian Science doing bad in QA?

Debian Science team was blamed about not caring about bugs High number of packages leaves wrong impression Team policy permits anybody to do team uploads ➜ If you are really bored by a bug in Debian Science simply use dch -team and fix it. Barrier is lower than NMU

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 82

What to do next?

If you are maintaining some scientific software, please contact Debian Science on the mailing list and join the team. If you intend to package new scientific software do it right inside the team. If you found some DDs maintaining similar software like you leave the umbrella and create your own Blend.

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 83

What to do next?

If you are maintaining some scientific software, please contact Debian Science on the mailing list and join the team. If you intend to package new scientific software do it right inside the team. If you found some DDs maintaining similar software like you leave the umbrella and create your own Blend.

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 84

What to do next?

If you are maintaining some scientific software, please contact Debian Science on the mailing list and join the team. If you intend to package new scientific software do it right inside the team. If you found some DDs maintaining similar software like you leave the umbrella and create your own Blend.

Andreas Tille (DebConf 17) Debian Science Montreal, 6th August 2017

SLIDE 85