Why Open Source Software Does Not Succeed
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View Why Open Source Software Does Not Succeed as PDF for free.
More details
- Words: 1,565
- Pages: 16
Open Source Licensing in Mixed Markets, or Why Open Source Software does not Succeed.
Alexia Gaudeul UEA, CCP
February 18, 2008
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
1 / 16
Introduction Why is Open-Source software not more popular?
Too difficult to install and use? 75% of open source software available on Windows, vs. 95% for proprietary software. But OSS more likely to be available on Mac, Unix, Linux.
Lower quality and less features? 50% of max number of features available in OSS, vs. 70% in proprietary software. But in most categories, at least one OSS has as many features as the proprietary software with the most features.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
2 / 16
Introduction Why is Open-Source software not more popular?
Too new? First mover advantage for proprietary software and lack of computer knowledge of consumers. Proprietary software more likely to be either very young (less than 1 year old) or very old (more than 10 years). OSS is distributed more normally around 5 years of existence.
For developers and geeks mainly, not for mainstream users? Higher popularity in software for professionals (database, graphic design). But also high presence in communications software (Instant messenger, Web browser).
No strategic vision and little knowledge of the market? Study of the organizational structure of OSS projects, e.g. LATEX
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
3 / 16
Introduction When and how is open-source software popular?
When does some proprietary software attaches itself to a dominant OSS and serve a niche. When does it stay incompatible with OSS? Is OSS more present in some markets than others? Markets with high network effects? Markets for professional vs. consumer markets)?
Does compatibility in standards help OSS? Does OSS benefit when it is available under the dominant platform, Windows? Does BSD software attain better diffusion or become more used than GPL software? In what implementation, OS or proprietary?
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
4 / 16
A model Development and licensing choices
Two types of consumers Portion M need a specific feature F (mainstream users, non-specialists). They cannot use the software otherwise. The rest does not value feature F (professionals, ‘geeks’).
Two developers, one open-source, the other proprietary qo is the quality of the open source software (exogenous). Develops F at cost c. qp is the quality of the proprietary software (exogenous). Always develops F .
The OS developer can choose between the BSD and the GPL. If BSD, then the proprietary developer can make his product compatible with OSS, else not (simplification).
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
5 / 16
A model Utility for the consumer
If there is incompatibility between the products, then: If OSS does not include feature F , or if the consumer does not value F , her utility from consuming OSS is qo + kno , with no the number of users of OSS. If the consumer values F , her utility from consuming PS is qp + knp − p, with np the number of users of PS and p the price of the proprietary software.
If both products are compatible, then replace no and np with no + np .
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
6 / 16
A model Utility for the developers
Utility for OS developer is qo + kn if she does not develop F , qo + kn − c if she develops F n / n /
is the number of users of OS code if the OS developer has pragmatic intrinsic motivations is the number of users of OS software if the OS developer has social extrinsic motivations.
The utility of the proprietary developer is profit pnp .
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
7 / 16
Monopoly The OS developer will not develop feature F , as anyway all users will use her software. The proprietary developer will price discriminate and serve only consumers of type F if there are many of those. Proprietary development generates higher total welfare than OS development, except most of it goes to the developer...
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
8 / 16
Duopoly Timeline
The OSD chooses her software license, either BSD or GPL. The OSD chooses whether to develop F or not. The PSD chooses whether to adopt the OS standard and be compatible with OSS, or not. The PSD chooses the price of his software and consumers simultaneously choose which software to use.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
9 / 16
Duopoly The decision of the OS leader (1)
Suppose the OS leader chooses the BSD and does not develop feature F. Compatibility will occur as long as k ≤ qo − qp , and if not, as long as network effects are relatively low and there are many consumers of type q −(1−M )q F (k ≤ o 1−M p ). In that case, OSS serves professionals / experienced consumers.
Whenever software is incompatible, the proprietary software serves the whole market.
If the OS leader chooses the GPL and does not develop feature F, then proprietary software is more likely to serve the whole market and leave no room to OSS. Developing an interface gains market share only for k ≤ qo − qp . Otherwise, PS serves the whole market.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
10 / 16
Duopoly The decision of the OS leader (2)
(qo-(1-M)qp)/(1-M)
k
No OS interfaces: * IC, PS serves all OS interfaces: * IC, PS serves all
qo-qp
(qo-(1-M)qp)/(1-M2) No OS interfaces: * GPL: IC, PS serves all * BSD: C, PS serves inexp. OS interfaces: * IC, PS serves all No OS interfaces: * GPL: IC, PS serves inexp. * BSD: C, PS serves inexp. OS interfaces: * IC, PS serves all
No OS interfaces: * GPL: IC, PS serves inexp. * BSD: C, PS serves inexp. OS interfaces: * OS serves all 0
Alexia Gaudeul (UEA, CCP)
M Open Source Licensing in Mixed Markets,
February 18, 2008
11 / 16
Duopoly The decision of the OS leader (3)
The leader with intrinsic motivation is more likely to choose the BSD than the leader with extrinsic motivation This is because she wants to encourage adoption of her standard, and does not care if this means the PS gets the whole market.
The leader with extrinsic motivation is more likely to develop F than the leader with intrinsic motivation. This is because she wants to attract users to her software. She develops F only if this allows her to gain the whole market (push PS out).
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
12 / 16
Duopoly The decision of the OS follower
There is no difference between followers with intrinsic and extrinsic motivations. Indeed, anyway, the proprietary developer will not adopt the OS standard as it comes too late.
The OS follower is indifferent between the BSD and the GPL.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
13 / 16
Main points Compatibility
The OS follower is less likely to adopt the proprietary standard than the proprietary follower is likely to adopt the OS standard. Even if the proprietary developer makes his standard public, it will usually not be taken up, unless this reduces competition.
The OS leader is more likely to choose the BSD than the OS follower is. When the OSS is the leader in its field, one standard, the OS one, may be used by all. When the OSS is a follower in its field, then it is likely to offer a different standard than the previously established proprietary standard.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
14 / 16
Main points Orientation
The orientation of OSS depends on whether the OSD is a leader or a follower and on what her motivations are. She is ‘more likely’ to develop F if she is a follower or if her motivations are extrinsic.
Competition encourages development by the OSD of user oriented features. The OSD develops user oriented features only if that allows her to gain a monopoly on the market and if her software is of higher quality than proprietary software.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
15 / 16
Main points The effect of motivations
OSS developed by OSDs with intrinsic motivations are more likely to be compatible with PS (i.e. see their standard integrated in PS) than OSS developed by OSDs with extrinsic motivation. This may explain the difference between software for professionals (Apache for example) and software for the end user developed as a hobby, to learn, to establish reputation or for ideological reason (Linux, GNU...). The choice of the license is a good indicator of the type of motivation of the OSD. Bruce Perens and the BSD vs. Richard Stallman and the GPL.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
16 / 16
Alexia Gaudeul UEA, CCP
February 18, 2008
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
1 / 16
Introduction Why is Open-Source software not more popular?
Too difficult to install and use? 75% of open source software available on Windows, vs. 95% for proprietary software. But OSS more likely to be available on Mac, Unix, Linux.
Lower quality and less features? 50% of max number of features available in OSS, vs. 70% in proprietary software. But in most categories, at least one OSS has as many features as the proprietary software with the most features.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
2 / 16
Introduction Why is Open-Source software not more popular?
Too new? First mover advantage for proprietary software and lack of computer knowledge of consumers. Proprietary software more likely to be either very young (less than 1 year old) or very old (more than 10 years). OSS is distributed more normally around 5 years of existence.
For developers and geeks mainly, not for mainstream users? Higher popularity in software for professionals (database, graphic design). But also high presence in communications software (Instant messenger, Web browser).
No strategic vision and little knowledge of the market? Study of the organizational structure of OSS projects, e.g. LATEX
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
3 / 16
Introduction When and how is open-source software popular?
When does some proprietary software attaches itself to a dominant OSS and serve a niche. When does it stay incompatible with OSS? Is OSS more present in some markets than others? Markets with high network effects? Markets for professional vs. consumer markets)?
Does compatibility in standards help OSS? Does OSS benefit when it is available under the dominant platform, Windows? Does BSD software attain better diffusion or become more used than GPL software? In what implementation, OS or proprietary?
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
4 / 16
A model Development and licensing choices
Two types of consumers Portion M need a specific feature F (mainstream users, non-specialists). They cannot use the software otherwise. The rest does not value feature F (professionals, ‘geeks’).
Two developers, one open-source, the other proprietary qo is the quality of the open source software (exogenous). Develops F at cost c. qp is the quality of the proprietary software (exogenous). Always develops F .
The OS developer can choose between the BSD and the GPL. If BSD, then the proprietary developer can make his product compatible with OSS, else not (simplification).
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
5 / 16
A model Utility for the consumer
If there is incompatibility between the products, then: If OSS does not include feature F , or if the consumer does not value F , her utility from consuming OSS is qo + kno , with no the number of users of OSS. If the consumer values F , her utility from consuming PS is qp + knp − p, with np the number of users of PS and p the price of the proprietary software.
If both products are compatible, then replace no and np with no + np .
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
6 / 16
A model Utility for the developers
Utility for OS developer is qo + kn if she does not develop F , qo + kn − c if she develops F n / n /
is the number of users of OS code if the OS developer has pragmatic intrinsic motivations is the number of users of OS software if the OS developer has social extrinsic motivations.
The utility of the proprietary developer is profit pnp .
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
7 / 16
Monopoly The OS developer will not develop feature F , as anyway all users will use her software. The proprietary developer will price discriminate and serve only consumers of type F if there are many of those. Proprietary development generates higher total welfare than OS development, except most of it goes to the developer...
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
8 / 16
Duopoly Timeline
The OSD chooses her software license, either BSD or GPL. The OSD chooses whether to develop F or not. The PSD chooses whether to adopt the OS standard and be compatible with OSS, or not. The PSD chooses the price of his software and consumers simultaneously choose which software to use.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
9 / 16
Duopoly The decision of the OS leader (1)
Suppose the OS leader chooses the BSD and does not develop feature F. Compatibility will occur as long as k ≤ qo − qp , and if not, as long as network effects are relatively low and there are many consumers of type q −(1−M )q F (k ≤ o 1−M p ). In that case, OSS serves professionals / experienced consumers.
Whenever software is incompatible, the proprietary software serves the whole market.
If the OS leader chooses the GPL and does not develop feature F, then proprietary software is more likely to serve the whole market and leave no room to OSS. Developing an interface gains market share only for k ≤ qo − qp . Otherwise, PS serves the whole market.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
10 / 16
Duopoly The decision of the OS leader (2)
(qo-(1-M)qp)/(1-M)
k
No OS interfaces: * IC, PS serves all OS interfaces: * IC, PS serves all
qo-qp
(qo-(1-M)qp)/(1-M2) No OS interfaces: * GPL: IC, PS serves all * BSD: C, PS serves inexp. OS interfaces: * IC, PS serves all No OS interfaces: * GPL: IC, PS serves inexp. * BSD: C, PS serves inexp. OS interfaces: * IC, PS serves all
No OS interfaces: * GPL: IC, PS serves inexp. * BSD: C, PS serves inexp. OS interfaces: * OS serves all 0
Alexia Gaudeul (UEA, CCP)
M Open Source Licensing in Mixed Markets,
February 18, 2008
11 / 16
Duopoly The decision of the OS leader (3)
The leader with intrinsic motivation is more likely to choose the BSD than the leader with extrinsic motivation This is because she wants to encourage adoption of her standard, and does not care if this means the PS gets the whole market.
The leader with extrinsic motivation is more likely to develop F than the leader with intrinsic motivation. This is because she wants to attract users to her software. She develops F only if this allows her to gain the whole market (push PS out).
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
12 / 16
Duopoly The decision of the OS follower
There is no difference between followers with intrinsic and extrinsic motivations. Indeed, anyway, the proprietary developer will not adopt the OS standard as it comes too late.
The OS follower is indifferent between the BSD and the GPL.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
13 / 16
Main points Compatibility
The OS follower is less likely to adopt the proprietary standard than the proprietary follower is likely to adopt the OS standard. Even if the proprietary developer makes his standard public, it will usually not be taken up, unless this reduces competition.
The OS leader is more likely to choose the BSD than the OS follower is. When the OSS is the leader in its field, one standard, the OS one, may be used by all. When the OSS is a follower in its field, then it is likely to offer a different standard than the previously established proprietary standard.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
14 / 16
Main points Orientation
The orientation of OSS depends on whether the OSD is a leader or a follower and on what her motivations are. She is ‘more likely’ to develop F if she is a follower or if her motivations are extrinsic.
Competition encourages development by the OSD of user oriented features. The OSD develops user oriented features only if that allows her to gain a monopoly on the market and if her software is of higher quality than proprietary software.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
15 / 16
Main points The effect of motivations
OSS developed by OSDs with intrinsic motivations are more likely to be compatible with PS (i.e. see their standard integrated in PS) than OSS developed by OSDs with extrinsic motivation. This may explain the difference between software for professionals (Apache for example) and software for the end user developed as a hobby, to learn, to establish reputation or for ideological reason (Linux, GNU...). The choice of the license is a good indicator of the type of motivation of the OSD. Bruce Perens and the BSD vs. Richard Stallman and the GPL.
Alexia Gaudeul (UEA, CCP)
Open Source Licensing in Mixed Markets,
February 18, 2008
16 / 16
Related Documents
Why Open Source Software Does Not Succeed
May 2020 8
Open Source Software Information
June 2020 23
Open Source Software
April 2020 22
Open Source Software
June 2020 14
Open Source Software
June 2020 21
Why God Does Not Exist
May 2020 17More Documents from ""
