I have recently decided to try Ubuntu (a South African term meaning “humanity towards others”), which is an open-source operating system built on Linux but similar in interface to traditional Windows or Mac operating systems. My question is whether a scientist could get by using only free software. Here is a list of programs that I think more or less covers the basics. If you have suggestions/experiences/comments to share, please let us know.
Operating system:
Office suite (Word processor, spreadsheet, presentations):
Web browser:
(Also free, but not worth the price: Internet Explorer)
Email and personal information manager:
- Gmail and Google Calendar
- Evolution (Ubuntu)
- Kontact (Kubuntu)
- Thunderbird
Reference manager:
PDF reader/creator:
(A PDF creator and editor is included in OpenOffice.org)
Statistics software:
- R (with R Commander or RKWard for easier use)
- SOFA
- PSPP
Graphing software:
Image viewing/editing software:
Audio editing software:
Lab management software:
Collaboration software:
Scheduling:
I’m getting along well using just free software. I’ve often had less trouble installing scientific tools and libraries on linux b/c of package managers like apt and rpm. Here are software some suggestions:
Image viewing: gthumb, feh
Image editing: inkscape
PDF viewer: okular
References: bibtex + Mendeley + Dropbox. I like Mendeley, but I’ve used up my alloted storage space. I’ve solved this by putting all my PDFs in Dropbox, and syncing Mendeley w/ Dropbox. Now I can access PDFs anywhere, even on my iPhone. I also have Mendeley update a bibtex file.
Document production: latex + bibtex + emacs
Version control/Collaboration: git
Terminal: yakuake
perry(Quote)
First roadblock — installed Ubuntu on an old Toshiba Satellite 4090XDVD laptop, and can only get 800×600 resolution. Surfing the Ubuntu forums offers some solutions, but none that can be easily implemented by someone new to Linux. Why can’t it just be an option like in, dare I say, Windows?
T. Ryan Gregory(Quote)
Chemical structure drawing?
David Bradley(Quote)
Well, I call myself a scientist and I’ve used free software exclusively at work and at home for 10 years or more, so I’d say the answer is “yes”.
If you’re new to Linux, it helps a lot to have a local friend or colleague who uses it (i.e. a real person) to get you started. There is an enormous amount of help and documentation on the Web, but it can be overwhelming at first. After a while, you get the hang of how to skim through it, discarding the useless stuff and picking out what you need. In a way, you need to know what questions to ask before you can find the right answers.
Your screen resolution may be related to the age of the laptop. If there is no suitable video driver available, Linux will resort to a “generic” driver and 800×600 might be the best you can get. If there were an option available, you’d see it – just like “in Windows”. It would be under System->Preferences->Display.
My other tip – learn the shell. For me, the joy of Linux comes not so much with whizz-bang standalone software packages, but in the way you can glue together many small tools, each of which do one thing well, to achieve an impressive result.
Neil(Quote)
Tried this:
http://art.ubuntuforums.org/showpost.php?p=8852781&postcount=21
No effect.
Tried this:
http://art.ubuntuforums.org/showpost.php?p=8233593&postcount=2
and broke it (flashing login problem)
T. Ryan Gregory(Quote)
Video used to be configured using the file /etc/X11/xorg.conf. However, I assume that you are trying the latest Ubuntu (9.10) – which no longer uses that file.
Advice that I’ve read for your model of laptop is to reduce color depth from 24 bit to 16 – maybe that is an option somewhere in the System menu?
Neil(Quote)
A new version of Ubuntu is coming out this month. Sometimes simply installing a newer version of Linux which often have better hardware support can be enough to solve the problem.
Keith(Quote)
I guess the roadblock I encountered with trying to do something similar is collaborators. Most biologists use Word + Endnote for manuscripts and refuse to try anything else. They aren’t going to install Linux (or even Open Office on Windows) just to please you. Well, some of the bioinformaticans are Linux geeks, but experimentalists, no way.
Jonathan Badger(Quote)
Of course, the best things in life are free!
Roger Harris(Quote)
It’s possible… at least that what’s I’m doing right now and I’m using about the same thing.
Operating system:
Ubuntu for personal computers (especially the new version 10.04)
CentOS or ScientificLinux (for servers)
Kubuntu is arguably based on the most promising desktop environment (KDE), but it’s still quite immature compared to Ubuntu.
Word processor & presentations:
Kile/LaTeX
OpenOffice.org
Notes
TomBoy : a very simple program to take notes. Really, really great.
Web browser:
Firefox
Chromium
Reference manager:
KBibTeX (for LaTeX)
PDF reader/creator:
Okular
Programming/Math/Statistics software:
Eclipse (by IBM) for C++/Java programming
IDLE for Python (using SciPy. SymPy and MatPlotLib to replace MatLab)
SOFA (written in Python BTW)
R
Non-open source software I need
Skype
Non-free software I need
Intel C++ compiler (perhaps)
Philippe Desjardins-Proulx(Quote)
Actually, the title is wrongly phrased — depending on what you are doing, FS might be your primary or even only choice, and not a replacement of a matter of “can that be done”. I am a biologist, but I am working with hight throughput data , with sequence data, statistics etc. All that spells: R, perl/python/whatever, some sort of httpd to provide services, a lot of software contributed by the scientific community (like phylip or blast), intelligent command line etc. Using Ubuntu is the simplest way of achieving a usable bioinformatic platform with all the necessary goodies installed. MacOS is fine, if you install apt-get (but in the end you end up using FS anyways), and Windows is out of question.
Thanks for the Mendeley tip, however. I am now immersed in the Windows/experimentalist combo world, so using BibTex and LaTex just doesn’t cut it :-)
z.
ztrewq(Quote)
This would refer to individual applications, which you’re right are often only possible with free programs. My question is, can you get by without using any non-free programs at all?
T. Ryan Gregory(Quote)
ztrewq(Quote)
Those who do numerical simulations can use Octave as a free alternative to Matlab, but perhaps even better is Python with NumPy and SciPy (as used by Los Alamos to simulate nuclear reactions so it must be good). Also, one can use Sage as a free alternative to Maple and Mathematica, although I have to say that it is not quite as elegant as Mathematica at the moment. I’m an Emacs guy for text editing, but there are numerous free text editors that are good for coding or LaTeXing. Kile has already been mentioned for LaTeXing and I like Komodo Edit for coding. Also, Jabref is a good alternative for reference management if you don’t want all the social features of Mendeley and use BibTeX as your primary reference format. A lot of people like Zotero for reference management, but I find that firefox extensions really slow down my browser so I avoid them. I am sad to say that I still use Xfig for drawing diagrams, even though I know there are much better programs available, but I have been using unices for a long time and am a bit stuck in my ways. Finally, no list would be complete without mentioning VLC for its ability to play just about any video file (as well as its ability to skip the mandatory pre-reel ads and trailers on rental DVDs that other players force you to watch).
In my opinion there isn’t a really good free alternative to Keynote (or even Powerpoint if you must) for doing presentations. Of course, there is the crappy presentation software that comes with Open Office, but I’ve always found it a battle to use. There are several free web-based presentation tools that are quite good (Prezi being worth a mention for its novelty), although I always find web apps to be clunkier than their desktop equivalents. If you don’t mind writing in markup languages then you have a choice of LaTeX+Beamer, which produces decent pdf presentations that are good for math-heavy talks, or S3, which produces HTML presentations viewable in any web browser. I’ve also seen people give presentations that are automatically generated from a wiki, but I do not know what the relevant software/plugins are.
Matt Leifer(Quote)
To browsers, add Safari and probably too many others!
There is more of the LaTeX-oriented collection of software, e.g. BibDesk, Lyx, and so on.
Plenty of free programming software, if that’s your thing.
My own private feeling on the overall matter is that while it can be done (working only with free software that is), it’s best suited to geeks who can work around any issues that arise, etc. Commercial companies take a lot of effort to gear their software to a typical end-user, whereas a lot of free software tends to need a bit of knowledge (generalising wildly here!).
I’m an independent computational biologist, who mainly works contracting to research groups and companies. I mainly use free software, with a few exceptions for convenience or where it fits with the wider trend. Personally while I could probably use only free software, it is more practical to pay for things that increase productivity. There is a point where you are better paying a small fee and spending your time on earning rather than fiddling with settings, etc. I know FS purists might bite my head off for writing that, but there is the practical reality, esp. when it’s your own income that’s counting. Life’s like that. My servers are Linux-based and most of the number-crunching is done there, but my desktop is Mac OS X as it gives me Unix + a commercial desktop platform (one that can host Linux / Windows via virtualisation).
David Bradley, I’m sure there is free software for this. I work at the protein-sized end of the spectrum. Molecular graphics software lets you build models as well as display them. (e.g. PyMol, UCSC Chimera, etc.) There will also be software specifically for small-molecule chemists as opposed to molecular biologists, but I’m not familiar with them.
Grant(Quote)
I’m surprised no one’s mentioned Zotero for reference management. Mendeley is doing some great things, but for now their software is free-as-in-beer, not as in speech.
Jason Priem(Quote)
Dude, you’re missing ImageJ for microscopy work/image analysis! Some of us are practically married to it…
http://rsbweb.nih.gov/ij/
Very useful list though — thanks!
Psi Wavefunction(Quote)
As a biologist regularly dealing with confocal image stacks, I’d add NIH ImageJ.
Tim(Quote)
> Graphing software: SciDAVis, Graph
I use also <a href=”http://magicplot.com” title=”MagicPlot Student”>.
And why didn’t you mentioned gnuplot?! :)
Alex(Quote)