Recent Updates

Last post
Notes from the biomass will continue at nftb.net. My...
spitshine - 2006-07-16 13:11
Stubborn
OK, you got me. While technically not blogging at the...
spitshine - 2006-07-07 10:55
Greetings from another...
Greetings from another HBS-founder (media-ocean.de)....
freshjive - 2006-06-15 20:06
HBS manifesto will be...
Hi there! I am one of the hard blogging scientsts. We...
020200 - 2006-06-15 18:13
Latter posts - comment...
Things to do when you're not blogging: Taking care...
spitshine - 2006-04-29 18:46

About this blog

About content and author

A few posts of interest

The internet is changing... Powerpoint Karaoke
Quantifying the error...

Link target abbreviations

[de] - Target page is in German
[p] - Paywall - content might not be freely available
[s] - Subscription required
[w] - Wikipedia link
More...

Search

 

Archive

March 2006
Sun
Mon
Tue
Wed
Thu
Fri
Sat
 
 
 
 2 
 4 
 5 
 6 
 8 
 9 
11
12
16
17
18
19
20
21
23
24
25
26
27
28
30
31
 
 

Credits

vi knallgrau GmbH

powered by Antville powered by Helma


Creative Commons License

xml version of this page
xml version of this page (summary)

twoday.net AGB

Scientific open source software

The market for most specialized science software is rather small, consequently the software is usually expensive. Many scientists even expect scientific software to be transparent (or downright free the younger they are) but how would one sustain a company in a niche market?
Dan Gezelter from the OpenScience project is facing the dilemma and summarizes the points that may have been made before but have never been solved.
Giving the software away for free and try to live off consulting is an approach that has been tried by the people involved in MySQL or WordPress with Automattic.com. However, the small market size will make it a very risky business - after all, grad students are always cheaper than any outside consultant, even if they would work for food and lodging.
Stew (guest) - 2006-03-14 10:52

Expecting scientific software to be free

I don't know about scientific software in general (EndNote etc. probably clean up) but I can't think of any current genetics software packages that are complex, user friendly or unique enough to warrant a great deal of expenditure - programs tied to specific pieces of hardware excepted. There always seem to be in-house or comparatively rough around the edges (but free) open source alternative solutions.

Perhaps the most obvious solution to making money from scientific software is "make it good enough for people to be willing to pay for it". Pricing realistically for individual researchers, not just that big pharma bulk licencsing department, might also help.

Deepak (guest) - 2006-03-14 16:02

Perceived value

It depends on what you need. In much of the informatics space, the end user usually has enough savvy to work with open source software (and free software can be a little rought about the edges, since it is free). This is especially true for software to generate data. I still think that the software to analyze and visualize data is not that trivial, which is why analysis packages do cost money. Stew is also right. There are too many packages doing to same thing, which always brings the value down

In the physical modeling space, the software is not trivial to write (even many of the academic packages are not free). How much should they cost? It boils down to the value people get out of it and the business model around it. I do think that the services model will gain ground, at least in industry. I fail to see how it will be that successful in an academic environment, where grad students and postdocs come cheap.

In the end it boils down to this "what am I getting out of it?". I also think that usability expectations have changed a lot over the years, especially as more and more non-domain experts use software and are less willing to jump through hoops.
spitshine - 2006-03-14 18:04

Can you make a good program with just a handful of people?

Widely used software products - say, Microsoft Word or its FOSS equivalents tend to have a long history, a large userbase and many programmers - and interface designers.
Scientific software developmens teams are much smaller despite data that is often more complex. Commercial developments needs to compete with scientific hacks on Pareto distributed ground. I don't think that a better user interface, higher standards and better support deliver the competitive edge in this particular market.

N.B. We should really have this conversation over at the OpenScience site. The second part of series has just been released.

Trackback URL:
http://binf.twoday.net/stories/1688996/modTrackback

Elsewhere...

Status

Online for 4585 days
Last update: 2006-07-16 13:11

Blogs
Conferences
Databases
Journals
Meta
Misc.
Papershow
Patents
PPI
Predictions
Publishing
The young PI
Useful tools
Profil
Logout
Subscribe Weblog