I admit that I never being too much into dashboard and widgets. The main reason was that they took long to load in my old macs (remember, I started with an ibook 12" G4) so I never bother to explore the widget power. Now, with a brand new MacBook Pro i7, I start using the widget instead of cluttering my status bar of icons.

Unfortunately, it appears widget are on the verge of extinction, since I don't really see lot of active development behind and apparently, one of the reasons, is that developers have troubles to make money out of widgets and users are still not using the dashboard to avoid wasting hardware resources. But I have some widget open and I do not notice any drain in my mac performances… of course, I have few, useful widget, and not zillion of them, all updating their status every millisecond.

So, what widget do I have? Aside the original apple sticky note, translator, dictionary, weather, calculator and calendar I have:

• screenshot plus : capture screenshots;
• gmail : check your gmail mailbox;
• twidget : tweet client;
• eject volume : eject mounted volumes;
• note file : note app with sync capabilities;
• istat pro : monitor the status of your computer;
• mighty monitor : check the battery status of connected devices;
• tea timer : a nice timer for more than just tea;
• dropbox droplet : easy share dropbox files;
• google apps : all google apps are one click away;

What I cannot find is a good widget for Facebook, so to free my status bar from MenuTab Pro (wonderful app, btw).

And what about you? Do you still use widgets and if yes, which ones?

by AndreaMinoia

Chembytes is not dead. I know, I know, I'm not new to such posts, then followed by silence, but this time I back for real.

In these months I was working to a series of papers and had no time to update the wiki, nor the Facebook page. Plus, I couldn't share the new software I developed since papers were far away from publication.

Meanwhile, GROMACS, TINKER, MATERIALS STUDIO, VMD, YASC, and a bunch of other software received important updates: in particular TINKER 6.0.x now dynamically allocates memory and is parallelized. Materials Studio also reached version number 6.0

On the IT side, I ditched Ubuntu and its awful Unity interface, I sold the iPad and my old macbook pro for the new iPad and a more recent second hand macbook. I have also decided to use this space as personal webspace, rather than build a new one. A new project, called Acad-iPad will also start and together with The Mac Corner page, is dedicated to the use of the iPad in Academia.

On the personal side, family start to grow: I got a dog last December. His name is Paco and is a six and half months old english cocker spaniel.

by AndreaMinoia

by AndreaMinoia

I then started to configure the ibook to become my laboratory buddy and, since is a ppc and not an intel, no virtualization was possible. Therefore, I needed to use fink to compile some linux software.

For my little ibook, compile the latest gcc was not a piece of cake: it took 10 hours to have the job done. Meanwhile, I took this photo of the gcc compiling setup, made hoping to avoid fusions :)

The fan was oriented directly on the keyboard and, below the computer, I had this Belkin cooling stand for laptops.

Proudly, I was able to maintain the CPU/GPU/HD/Memory temperatures so low that the ibook fan never went for a spin. Gh!

by AndreaMinoia

by AndreaMinoia

One of these, is doing tutorial. I knew was a great thing to have step-by-step tutorials, but I was way too lazy to do it. Anyway, the result is that the first two tutorials for TINKER are online. It takes quite a lot of time to write a tutorial, in particular for TINKER that is not well documented. Or, at least, informations are spread between the manual and the source files. I have to admit that was a bit of fun hunting informations in between lines and lines of Fortran code, but at the end, I think I have improve my knowledge on TINKER. So, hope the tutorial will helpful for you, as they have been for me.

Another collateral effect of my trip into the code, is that I was able to retrieve a full reference list for TINKER, which is also online.

by AndreaMinoia

The first great news, is that I have start to collect all the scrapnotes i have in Evernote notebooks. The first notebook with 10 tips is already available here, both in evernote and html format.

The choice of Evernote is due to the fact that is free and is a powerful software to store, manage and search notes, while this is difficult to do in a wiki. Let me know if you like this experiment

by AndreaMinoia

On Thursday I went to London to learn about GPU (graphics processing
unit) accelerated calculations. GPUs seem to be the future in high
performance computing they can provide computing power (per gigaflop)
10 times cheaper and using 20 times less energy than conventional CPUs
(=> less electricity and air conditioning running costs). This is
achieved through a highly parallel architecture; each graphics card
can have up to 512 cores and one can put up to 4 GPUs in a single
desktop. Connecting 2000 CPUs together and achieving good scaling
would require huge investments in special high speed interconnects
(e.g. infiniband) that are far too expensive for most universities and
thus only used in supercomputing centres. This is why on Lucky (where
the nodes are connected with gigabit Ethernet) you can only hope for
good scaling using the cores in a single node, so max 8 cores. Using a
desktop with 4 GPUs, you have 2000 GPU cores in one machine using
shared memory, so you can run parallel jobs on 2000 cores without the
need for expensive high speed interconnects.

While not all applications have been rewritten to work on GPUs, there
is a lot of work going on and progress being made. See
http://www.nvidia.com/object/molecular_dynamics.html for details.

MD codes which now support GPU acceleration include:

NAMD 2.7 Beta 2 including CUDA Acceleration
HOOMD: Highly Optimized Object Oriented Molecular Dynamics
MDGPU
LAMMPS GPU Port
GROMACS using OpenMM
ACEMD Bio-molecular Dynamics Package
AMBER CUDA Port

For quantum chemistry there is a Terachem, written especially for GPUs
which offers up to 1000 x speed up using a GPU compared to using a
single CPU using the GAMESS QC package. Unfortunately, the
functionality is limited at the moment; it can only handle S and P
orbitals and can only do basic single point calculations,
optimisations, transition state searches and molecular dynamics. No TD
or CI for the moment see http://www.petachem.com/products.html

However, GAMESS (US) is currently working on a GPU implementation
which should be able to do CI and TD calculations, but it may be some
time before it is ready.

For those of you who use Matlab, there is also a GPU accelerated
version of Matlab see http://www.accelereyes.com/successstories to get
an idea of the speedups which can be achieved.

If anyone wants to buy a GPU personal supercomputer, you can find them
at http://www.armari.com/system.asp?SysID=1351 Once you put a
reasonable amount of RAM on the machine and add a 2050 GPU, it's about
5000 euros for one machine. But remember this can do an AMBER MD
simulation at the same speed as 64 cores on a supercomputer, which
would cost ~ 50000 euros. And you don't need a special air-conditioned
room, you can put it under your desk.

Not bad eh?

by AndreaMinoia

With respect version 2.0, the main differences are:

New kernel, new UNR menu, splitting up from EasyPeasy and switching to Ubuntu Minimal as base, new applications. Leeenux usplash, xsplash, gdm login. New eee-control modes. For eee 701g (2G, 4G, 4G surf - with Intel Celeron): 450MHz, 540MHz, 630MHz, 765MHz, 900MHz, and overclocking to 945MHz. Corrected modes for netbooks with Intel Atom processor.

I've downloaded the standard edition (extended and game editions are also available): the iso is of 450MB and once installed takes 1.3GB of space on the hard disk. Not bad.
The installation is fast and dummy proof, not a single problem. About the installation I can only say to do manually the HD partitioning, so to get rid of the swap partition of 256MB. With 1GB of ram is almost useless, it take a lot of space (I have 4GB…) and increase the number of read/write cycles for my little SSD HD.

First impressions
The boot, once leeenux is installed, is fast: I would say about 30 sec to arrive at the login screen. The new graphical interface loose a bit of appeal with respect the old one, but I have to admit is more clean

The new gui (the background is mine…)

The top bar can be customized with some applets, which allows to use multiple workspaces.

Navigation through the menus and screens is smooth even with the cpu underclocked at 450MHz (by default the clock is set to 630MHz). The new settings for the cpu clock seems to work flawlessly and allows to manage heat (the cpu temperature while I write this page increase from 48-50 @450MHz to 53 degrees Celsius at 900MHz. Also, the frame gets noticeable hotter) and the duration of the battery (1h30 and I'm still 40%… btw, the battery is almost dead) Also, the Eee icons let us turn on and off wifi and card reader, but not the camera (pity).

All the function keys are working, as well as resume/suspend, wi-fi, brightness, external screen and mute/volume with relative OSD. Microphone is also working, while the camera is, for me, in black and white only, both in cheese and skype. Yes, I have tried to change the size/resolution of the camera. Well, not a big deal for me.

Software
As usual, the pre-installed software is a good selection, and can make happy the majority of the users. What is missing, can be easily installed via synaptic.
Since I need a scientific distro, I had to remove some software I was not going to use, in order to install what I needed. in particular I have removed:

• all the games and emulators
• thunderbird
• sunbird
• prism facebook
• rhythmbox
• gnasm video player
• abiword
• gnumeric
• paint

and some other minor packages

What I have installed is:
for work:

• latex
• openoffice (I needed for impress)
• gfortran
• java
• gnuplot
• xmgrace
• imagemagick
• povray
• tachyon
• pyton-numpy
• gromacs
• tinker
• vmd
• marvin beans
• ekee
• openbabel
• skype
• chrome
• prism google-mail

for fun:

• gphoto2
• babelgps

and something more.
This leave me with 909 MB free on 3.7GB (25% free), more than enough to work, considering I have a 8GB SD always in my card reader.

Screenshots
Now, many people find the screen of the 701 too small to work, and that is true, but with leeenux… is a bit less true. Every window is maximized by default so there is plenty of space. Let's see some examples:

Openoffice
From the repository, we can install the version 3.1. Openoffice has a quite messy GUI: lots of icons, menus and sections. I have set the view options to scale the interface to 90% and to use the small icons. The result is below:

Not bad.

Terminal
The terminal is simply perfect:

Browser: Chrome
Already Chrome has a clean interface, that makes it perfect for the small screens like that of the 701, plus it can be used fullscreen (F11):

xmgrace
xmgrace is not well displayed by default, but it is enough to add the option -free to the command used to fire it up (also via the menu), and you will get:

Marvin Beans
Marvin Beans is one of the best suite to work with chemical structures: it has a 2D editor with 3D clean up capabilities, a converter and a 3D viewer. It runs wonderfully in Leeenux:

VMD
VMD is a bit tricky: I would suggest you to unmaximize at least the main gui. in the picture, a caffeine molecules is displayed and the tk console is also active. All in unmaximized mode. Hint: use the option -size 400 400 to start vmd, so that when unmaximized you will have a manageable windows.

gThumb
I use gThumb to create screenshows of images of plots or VMD renderings. Also, it can be used to perform basic photo editing.

evince
You cannot use a computer in the lab if it cannot visualize pdf files. In Leeenux evince is installed by default. Here in action in full screen mode with a paper of mine:

Conclusions
In few words, leeenux is the uber distro for the 701 and, probably for all the netbooks out there.

by AndreaMinoia

Recently, I had a new computer in the lab, with Intel Core I7 CPU. I wanted to test the performances with a heavy Gromacs simulation on 4 cores, leaving hyper-threading (HT) enabled. I have noticed that at a certain point, randomly, Gromacs stopped to write the output files (.xtc, .edr, .log), even though all the four cores were running at 100%.

I have googled a bit and ended up on a discussion on the Gromacs mailing list, where I found two different suggestions: disable the HT and update openmpi.

I didn't want to disable the HT, so I have checked the version of openmpi installed via synaptic under ubuntu 9.1 and it was the 1.3.something. The discussion suggested there was a bug in that version, so I have updated to the last version of openmpi (1.4.1) and everything now is running smoothly. Better to uninstall with synaptic the old version. Of course, you also need to recompile mdrun.

I have also noticed the during the execution, mdrun jumps from core to core, but this should be a "feature" of HT: I don't know how bad this can affect the efficiency (speed) of the calculation. I haven't had time to disable the HT and try again.

Hope this can help.

by AndreaMinoia

New experiment: now you can follow me and the wiki on twitter. The idea is to use this channel for news and other things related to chembytes, hoping it will make it easy follow the wiki.

by AndreaMinoia

• cpu i7 quad core with hyper-threading, 2.80GHz
• 8gb ram
• great video card (forgot the model though)
• HD 150gb + 1TB
• etc. etc.

First things done: formatted Windows 7 professional and installed the last Ubuntu 9.10 64 bit. Then configuration and transfer data from the old pc to the new one.

Oh joy oh joy :D

Some picture of the nice event:

And, there is also the card reader, so I can share easily stuff between the desktop and the eeepc. yehi!

by AndreaMinoia

xmgrace is, like gnuplot, a nice piece of software to create scientific plots. As opposite to gnuplot, xmgrace has a graphical interface and can generate only 2D plots. To the other hands, it allows for quite sofisticate analysis, such as autocorrelation, fourier transformation, interpolations, non linear fitting, regression, convolution and so on.

The user guide is here.

Install xmgrace in linux is quite straightforward, since is included in the repositories of all the major linux distribution, as for gnuplot.

Soon I will write a page dedicated to xmgrace.

by AndreaMinoia

Gnuplot is, along with xmgrace, my program of choice to plot data. I even make a script to plot data on the fly, directly from bash.

From the official website:

Gnuplot is a portable command-line driven graphing utility for linux, OS/2, MS Windows, OSX, VMS, and many other platforms. The source code is copyrighted but freely distributed (i.e., you don't have to pay for it). It was originally created to allow scientists and students to visualize mathematical functions and data interactively, but has grown to support many non-interactive uses such as web scripting. It is also used as a plotting engine by third-party applications like Octave. Gnuplot has been supported and under active development since 1986.

Gnuplot supports many types of plots in either 2D and 3D. It can draw using lines, points, boxes, contours, vector fields, surfaces, and various associated text. It also supports various specialized plot types. Demos here.

Gnuplot supports many different types of output: interactive screen terminals (with mouse and hotkey input), direct output to pen plotters or modern printers, and output to many file formats (eps, fig, jpeg, LaTeX, metafont, pbm, pdf, png, postscript, svg, …). Gnuplot is easily extensible to include new output modes. Recent additions include an interactive terminal based on wxWidgets and the creation of mousable graphs for web display using the HTML5 canvas element.

You can easily install gnuplot in linux, since is in the repository of all the main distro.

I encourage my students to use gnuplot and to help them, I made a cheatsheet.

This cheatsheet is far to be complete, but just time gnuplot in google and you will have tons of results, tutorials, guides, examples… An amazingly useful link for gnuplot is this.

by AndreaMinoia

It is nice, because is small, light and resistant to shocks, humidity, … : perfect for people like me, that commute every day from home to work. The problem was that I never found a decent linux distribution that was really working out of the box, and that allows me to transform the small eeepc in a perfect lab buddy. The best distro I tried out were:

• crunchee, based on ubuntu 8.10
• fluxflux

and.. that's it. Then Asus put an end to the production of the little 701 and linux distro followed the market: netbook with big screens, huge standard hard drives (bye bye solid state hard drives). So, my little friend was sitting, unused, in a lab drawer.

On monday, I was using train+folding bike to go to work, so I left the macbook pro at home: of course, an unexpected visitor came in the lab and I used the eeepc again. With the usual problems. So I surf the web to see if there was some new distro that could have perform better than crunchee. And, joy oh joy, I have found it, the perfect distro:

Based on ubuntu Jaunty, comes with a iso of less than 500MB. It has a nice graphical interface based on gnome, it boots fast, everything really work out of the box (yes, also suspend/resume, functions keys Fn+Fsomething, wireless strong enough, webcam (B/W only for me when using cheese), sound, microphone, … ) and the windows are, by default, maximized and adapted to the small screen.

Quoting from the web site, some good reasons to use Leenux:

• Don't throw away your old Asus eee netbook, because it can gain usability once again with Leeenux.
• Leeenux is a linux distribution made mainly for Asus eee with 7" screen, but other netbooks can use it as well. It is lightweight, and in every new release it uses less resources. Current ISO contains only 460MB. It can be installed on eee with 2GB of SSD, and on 4GB eee you will have planty of room to spare. Installed, it takes up only 1.2GB.
• It is the best distribution for this widely distributed netbooks.
• Everything works Out of Box, unlike Ubuntu UNR.
• It is made using remastersys from easypeasy 1.1, which was made from Ubuntu UNR 8.04. In order to be as much lightweight and fast, in May 2010 it will be made only from Ubuntu UNR 10.04.
• It is FREE! Liberty means a lot, and there is no proprietary software nor drivers included in this distribution. EasyPeasy is not free. Its only free of charge, but contains lots of propriatery software. Why propriatery software is BAD, you can find out on many websites.
• Unlike many other distros which try to be as complex as possible, Leeenux is trying to reduce Bloating. Thus many things are not installed by default, the things for which there is 99.9% you wont need. If you still need them, they can be installed via Synaptic.
• Propriatery software can be installed, if needed.
• It is licenced under GPLv3, so anyone can redistribute or change it if they like.

And, I could not agree more.

With only 1.2gb used, it leaves plenty of room to install scientific software:

1. latex
2. compilers
3. python/perl libraries
4. shells csh, ipython, tkcon
5. Mendeley desktop for organize papers
6. gromacs
7. tinker
8. gnuplot
9. vmd
10. pymol
11. xmgrace
12. chemistry software like chemdraw and openbabel
13. gimp
14. graphical utiles like imagemagik
15. wine (so to use portable apps such as open office etc.)

and much more…

A list of the default applications is here

Now, having 1gb free and all I need installed and fully working, is time to show some screenshots:

The graphical interface (fully customizable)

The terminal: nice and big thanks to the default maximization… come on… I know you use it too ;)

VMD (unmaximized), with the window set to 400x400 pixels

You can see how nice is the feeling with the small screen of the 701: not a problem anymore. To make the long story short:

If you, like me, are looking for give a new life to your old 701, this is the distro for you. And if also you are appreciate the work of the Leeenux team (which is not connected in anyway with me or this site), you may want to offer them a beer.

by AndreaMinoia

So, if someone is out there and reading, you are welcome to join the site and the forum. The only "drawback" is that you have to register for free on wikidot, in order to post.

That's all folks!

by AndreaMinoia

To create a picture of a latex equation, we can, of course, use latex and a text editor in the usual way. Or, we can use dedicated software or on-line tools.

Software

• LaTeXiT (Mac OS X)
  In Mac OS X I use LaTeXiT, which is freeware and allows to customize the resulting image in many ways. The image that is generated can be easily insert in Keynote, Pages, MS office documents, etc.
  You need to have some Latex tools installed, like ghostview. The easiest way to have latex on Mac OS X, is to download and install the MacTex packages.

• Ekee (Linux)
  Similar to LaTeXiT, but for linux, Ekee can be downloaded here or via package manager if you use the proper repository.
  To make it work with my latex installation, I needed to edit the latex template and comment the following packages:
  %\usepackage[T1]{fontenc}
  %\usepackage[latin1]{inputenc}
  %\usepackage[french]{babel}
  To do this, just display the options and click on "Edit LaTeX template". You can save your input in latex, pdf, png and svg. Not bad!
  This is what you will see¹:

Online Resources

• Latex2png
  You can visit the website of latex2png to write online your latex equation and generate a png or gif image. You can vary the resolution of the image from 75 up to 1200 DPI, useful for poster and other publications. Below a example of Latex2png image:

by AndreaMinoia

Ok, too stressed to start to work, lets get a coffee with some nuts biscuits (I keep them for emergencies and desperate cases). A bit of web surfing and in, between one ridicule/sad-but-true news about italian politics and the other, I was about to regain my normal mood.

Then… disaster strikes: a little chat on skype could not have proven me more wrong. A friend of mine linked me a new paper and…

I got scooped!!!! Darn it!!!

Now I have to double sugar coat my results… if you know what I mean…

*I have borrowed (sort of) those images from http://www.phdcomics.com

by AndreaMinoia

by AndreaMinoia

Here some thoughts about scripting Vs manual donkey job. To clarify, is always better to do a script when we have to deal with repetitive tasks, such as file modification?
For instance, I usually have to work with gromacs and modify pdb files, topologies and other files, but I do this with Vi. why? Here my reasons.

Scripting is in general a good thing: allows to do things faster and better. In general. The downside is that to be valid, a script should be a good script, even in is a non-optimize code. And for "good" script I mean a well written bunch of lines of code: indented, commented, and flexible, both regarding the execution and the management of errors. This is an investment of time, which is good if we have always, exactly the same problem to solve, or the same task to do.

You have to calculate a distance between a list of coordinates? go with a script.
You have to do a sequence of bash command? do a script.
You have to read specific data from a well formatted file? go with a script.

You have to do something with data that vary from time to time or work with files that can have slightly different structures? Don't waste time with a script. It is much easier if we are the intelligent script rather than write one. So, we have just to find the right tools to to our job, in a fast and flexible way.

So, my idea is that when a script is starting to get an intricate sequence of if.. else if.. else ..for.. while… try… catch… etc. to try to deal with different type of input or output, better try to do by ourself the real thing, rather than waste weeks programming. For the well known Murphy's law, we will always find a situation in which our nice script of 100000 lines, fails.

So here we are, human scripts. boring, slow and likely to do mistakes. Or not? Likely, not… at least if you use linux and you are not afraid to learn a bit of bash and the supreme Vi. Here a practical example where having a script is totally useless:

Let's suppose I generate a polymer with my script and save it as pdb file. Now, I need to generate the Gromacs topology and I want to use pdb2gmx. Well, the pdb must then be adapted to the data in the Gromacs rtp file: atom names, residue names and residue numbers must be conveniently assigned, according to those in the rtp file. And all this, maintaining the pdb fix format. A script to do this should read the data in the rtp, understand where those data apply to the pdb file and modify the pdb in the proper way, keeping the right format. Rtp and pdb files can vary a lot from case to case and a script that will account for all the possibilities is out of reach. Much better do the things by our own, even if it seems a complicate (and boring) job. Nothing could be more wrong. Open your rtp in a vi terminal, open the pdb file in another and modify it using the substitution commands in vi and vi macros, defined on the fly. With a bit of experience, in 2 minutes a pdb of hundreds thousands of line is properly modified, ready to be used with pdb2gmx. ant 1min30sec will be used by vi… your job was finished after the first 30 sec:)

So, my idea is to do scripts when it is easy or difficult, but cannot be done otherwise in a simpler way. For the remaining cases, let your brain be the intelligent script, and manually use the powerful tools that are out there.

by AndreaMinoia