Hoopla aside, I was directed to the following article (click) which has sufficiently motivated me to push my linkedin account to 100%, make my fb profile a little more professional (i.e. remove the suspicious gas-mask) and actually start tweeting, or microblogging or updating my status or what-have-you. But, woe-is-me who has very little interest in this to begin with to have to update all these social-media sites so regularly. Can't I just have a single program where I can follow the updates and when I make an update it will syndicate it to all the social media sites I subscribe to? Well I went searching and settled on the first thing I found (a lot like how I ended up with my wife) ping.fm which is a website that accomplishes this syndication very easily. However, thanks to publicizing my interest in this syndication business, I was turned on to Tweetdeck, which is a superior tool for posting and managing multiple features to all these subscription sites. I also am really impressed that they are completely cross-platform (Michaelsoft Binbows, Macintosh Apples, and Linux) and I have it installed on my Kubuntu box at home, my wife's ubuntu netbook remix (though there I had to separately install Adobe AIR first) and my "business machine" running 7. Anyway, check out tweetdeck at tweetdeck.com
if [ -z "$1" ] then echo $0: usage: $0 filename exit 1 fi
while [ ! -z "$1" ] do # do stuff to $1 shift doneexit 0
Selection.Font.Hidden = True
A bigger problem with this job was losing all the toshiba drivers including those that run the network adapter. I spent a few minutes searching and finally came across this forum thread:
Which pointed me here:
This was critical as I was S.O.L. with most of the hardware on this machine without these!
So now I just need to get any other drivers deemed useful, deal with the activation, SP2 and SP3 upgrades, WGA issues and all that jazz. (yes windoze is a pain).
After the couple hours needed for all that, I'll take 30 min. to do a Ubuntu 10.04 LTS (lucid lynx) install just in case the other system goes to shit!
p.s. I figd out activation (FFX Madman), however it may have interfered with installation of MSOffice 2007.
I accomplished a network upgrade to 10.04 (lucid) in about an hour, however I had to troubleshoot a few bugs, two of which were related. The login screen lagged a very long time and my USB drive would not automount on boot like it used to. Searching the web, I found a fix and after deactivating the floppy drive (I don't even have one) in the BIOS settings this was corrected.
After installing my standard (and more) suite of software, I made my "Studio" menu and fired up JACK, Qsynth, and Qtractor to test things out. Unfortunately, I was unable to record midi from my keyboard! After racking my brain a bit, I decided to uninstall Qtractor and build it from source. Although this is always a non-trivial task (although getting easier), it fixed the problem. I also installed LinuxSampler, and the front-end GUIs Qsampler, and Jsampler (see earlier post 3.10) from the CVS repos.
I had hoped that the LinuxSampler engine would eventually be my "go-to" for running multiple samples on different channels simultaneously. However, I cannot figure out why the GIG files (or any samples for that matter) sound lousy, bad, terrible, fuzzy, etc (keywords!). Really annoyed after having built Gigedit, LinuxSampler, and Qsampler from source. I decided to play with Qsynth (.sf2 soundfont sample handler GUI for FluidSynth).
After some digging and tinkering I believe I have a solution. Use Qsynth with Qtractor and Hydrogen (on top of JACK of course) to build all my MIDI files in real-time with a MIDI keyboard and then record into Ardour. The breakthrough was figuring out how to play multiple samples (.sf2) from Qsynth simultaneously. Here is how:
Qsynth loads audio "engines" and you can set up several at once. However, you only need one!
In setup I select alsa_seq as the midi driver, and jack as the audio driver. Then under "Soundfonts" tab select each of the soundfonts (.sf2) files you want to have available for your Qtractor piece. Then edit each entries "offset" so they are different (0, 1, 2, etc). This sets the "stack" with different banks. Save and leave "setup." Then select "Channels" from the main Qsynth interface. Right-click the channel of interest and select "edit." Then you can select different "banks" and assign different soundfont samples to different channels! Now give your neglected 2 year-old a high five!
Now I am curious if I can make the webcam run in Ubuntu Linux (while I'm waiting around for a network upgrade to 10.04 LTS). I ran lsusb and found the driver info:
Ricoh Co., Ltd Visual Communication Camera VGP-VCC6 [R5U870]
Now I think I may be able to get some headway according to this site:
back for an update...
Installation of 10.04 took ~2 hours over a wireless connection (a bit slow, but I had other things to do anyway). It will take some time to see what kinds of advantages, upgrades, etc it really adds, but I expect to be fully satisfied.
After that I installed the following files from the above URL.:
Then I tried "cheese webcam booth" software and verified the motioneye webcam works!
How to Create an HTML Download Link
Self Extracting Zip File (1) (2)
I also found a website that used a xspf player that included a "shuffle" feature. I thought this was neat and had to have it! I found one with instructions here and here, extracted and uploaded the .fla, .as, and .swf files and presto-blamo it worked.
Regarding my presentation, I ended up exporting my OOO3 Impress presentation to a PDF and running a slide show with evince! Very fast, hardy solution. Of course I was not able to run my x-ray crystal videos embedded or from a macro. But I had them both queued up in totem and running in the background during the presentation so I could easily pull them up without losing momentum.
So now I just have about a weeks worth of corrections to make to my dissertation and I will submit the final version to the graduate school. I plan to attend a hooding ceremony at commencement on May 22.
New Discovery - It looks like I should be able to embed an AVI file directly into my Impress Slide. From OOO go to Insert -> Movie & Sound and select the .avi file. May need to install a codec (libavcodec?) first...verification needed. Also to convert the MPEG from PyMOL to an AVI format install ffmpeg and then run
> ffmpeg -i file.mpeg file.avi
A video toolbar comes up when you click on the embedded box where you can select "repeat." The AVI video will automatically start playing when you advance to the slide and continue until you move on.
Disclaimer: Although writing a chemistry dissertation with open-source software is readily accomplished, it is probably unwise to make a switch from MS Windows after already starting your writing. It is recommended that you make the decision a year-ish early and write at least one hefty report to acclimate :)
OpenOffice.org 3.1 Office Suite:
OOO3 Writer - Word Processor
OOO3 Draw - Drawing Program - I used this to prepare all of my Figures and Schemes saving each one separately with a unique alpha-numeric identifier for easy renumbering at the end.
OOO3 Calc - Spreadsheet - Used to generate tables and graphs
JabRef 4.2 - Java-Based Reference Manager - Fast and flexible, however there is a bit of a learning curve (maybe a few hours) associated with using the "style file," a template that formats your bibliography. I know there are several others out there, but JabRef was easy enough that I could get started with it right away and it is very robust. Documentation on their web site is good, and you will find there the critical plugin for syncing to OOO3.
More significantly just follow the CVS instructions from here.
Note that to install from the cvs source you must first install:
cvs, automake, g++, and respective dependencies, all of which are available through synaptic.
In addition either libsndfile or libaudiofile with the 'dev' files must also be installed.
These are fun for the kids, however, GCompris may be for 4+ not 2+ like advertised!
Gcompris 8.4.13, and dependencies:
gnucap + gnuchess + python + pysqlite + pycairo + pygobject +pygtk + SDL + SDL mixer +tuxpaint
which you can get here
Also needed for TuxPaint: SDL_image-1.2.7-i486.pet + SDL_Pango-0.1.2-i486.pet + SDL_ttf-2.0.9-i486.pet + libpaper-1.1.23+nmu1-i486.pet from here as well :)
Note all the other goodies from the same site: Klavaro, Enigma, Childsplay, TuxMath.
iPod Touch-2G 8GB (firmware: v. 3.0)
Desktop PC (home-built) with Pentium IV 2.8 GHz, 2GB DDR RAM, Audigy Soundblaster Platinum with mounted breakout-box (operating system: Ubuntu-9.10 with linux-rt kernel v. 18.104.22.168.10)
- iPod -
Xewton Music Studio
Flare (virtual turn-table)
Now with just a little left to do and possibly not enough time to do it, I want to document a few ideas to add to the project:
- add theremin/scratch? to nOe-i06
- add noise.io synth/field-samples? to nOe-i09
- use field-samples for nOe-i10
- find noise cancellation software (audacity plugin?) for nOe-i08
- need artwork and liner-notes
At some point, I decided to upgrade to UbuStu-9.04 as I had been enjoying the usability of Ubuntu-9.04 (Jaunty) and 9.10 (Karmic) at school. Sadly, I had some major issues with this install that forced me to move to Ubuntu-9.10. I knew I would probably try UbuStu-9.10 when it was released with the hope that the earlier incompatibility would be corrected. It seemed to work okay, however I could not get it to work with my Ralink Wirless card. Also, I was frustrated that the UbuStu developers didn't include some of the slick upgrades (namely the Ubuntu Software Center) that is steadily improving the usability of Ubuntu. I had no way of accessing the web without "ruining my life" so I moved on to 9.10 and decided to go on a significant "shopping spree." I installed loads of science, education and music apps from the Software Center, as well as the kernel-rt from synaptic.
The most significant challenge was getting the JACK audio server to run and I eventually found that I had to add my user to the "audio" and "pulse" groups under System -> Administration -> Users and Groups. The follwing URLs were helpful for me:
and most significantly:
I still get the following error from Rosegarden however:
System timer resolution is too low
Rosegarden was unable to find a high-resolution timing source for MIDI performance.
You may be able to solve this problem by loading the RTC timer kernel module. To do this, try running sudo modprobe snd-rtctimer in a terminal window and then restarting Rosegarden.
Alternatively, check whether your Linux distributor provides a multimedia-optimized kernel. See http://rosegarden.wiki.sourceforge.net/Low+latency+kernels for notes about this.
To date, I have sold just a few items:
3 cell phones, and a graphics card.
And have earned a grand total of $180 which has nicely offset my recent purchase of a netbook for my wife (asus eee). She loves it, and I must say, I am quite impressed with how much power can come in a little package for pretty cheap ($250).
I'm sure I'll have Ebay updates in the near future!
Probably my favorite puplet for general use would be Nearly Office Pup (NOP). Although I find it runs a little slower on my hp pavilion ze5600 laptop, that is probably due to a number of useful changes to the base Puppy, specifically xfce as a window manager and Opera for the browser. However, it is these two additions that make it a slick, user friendly system. Once adding the openoffice.sfs to the boot-up options, NOP is a clear winner for a desktop environment that even Grandma could use.
My second favorite puplet as an all-purpose desktop would be TeenPup-legacy2009. Although about as "fat" as they come, this distro has a tremendous amount of applications with a "windows-like" feel and usability. I would keep it off of most old laptops, but it is a great choice to get the most out of an older desktop.
Perhaps the most cutting edge of the puplets is UPup which is the current "pet" project of Barry Kauler (Puppy Linux originator). It is listed as "experimental" and in fact I remember finding some basic applications were broken. Simply put, UPup is a Puppy platform which readily installs debian (.deb) packages from Ubuntu repos. This makes it one of the most extensible, flexible of any puplets for a linux newbie like myself. My guess is this will be the future of Puppy Linux.
At school, my French colleague and two of the Asians have left. Also my Vermont friend is gone to go teach ski racing to teens, but plans to come back to write his thesis in April.
Much of my time has been split between writing my dissertation and playing with the new $40,000 computer cluster. It was assembled and configured by a gifted 22 y.o. UNH undergrad who did a really nice job. It consists of a head node and 6 compute nodes running OpenSuSE, each with 24gb memory and the newest, fastest intel nahelem procs in a dual-quad core configuration. The nodes are networked with gigabit up/down cables. That's about the extent of what I can remember. Of course it has a slew of useful computational chemistry packages.
Although I don't have any training in CS, I managed to piece together a useful python script based on a similar one that I used as a template...though I added a number of useful bits for my own preference. Here it is:
print "*************************************************************\n**See g03jobLOG in Home Directory for Log of Jobs Submitted**\n************
*************************************************\n\nEnter the Following Information for Your Gaussian Calculation:\n"
xyzFile = raw_input("Enter .xyz File Name (omit extension) > ")
uniqueID = raw_input("Enter Unique Identifier (no spaces; optional) > ")
g03root = raw_input("Enter Root Section (#) > ")
charge = raw_input("Enter Charge > ")
multiplicity = raw_input("Enter Multiplicity > ")
title = raw_input("Enter Title; optional > ")
queue = raw_input("Enter Queue (short (s) or long (l)) > ")
if uniqueID == "":
UID = ""
UID = "_"+uniqueID
if queue == "s":
SGEQ = "short"
if queue == "l":
SGEQ = "long"
if title == "":
title = xyzFile+" / "+uniqueID
f = open(""+xyzFile+UID+".com", 'w')
f.write("# "+g03root+" \n")
f.write(" "+title+" \n")
f.write(charge+" "+multiplicity+" \n")
inp = open(""+xyzFile+".xyz", "r")
outp = open(""+xyzFile+UID+".com", "a")
for line in inp:
lines = open(""+xyzFile+UID+".com", "r")
list = lines.readlines()
fout = open(""+xyzFile+UID+".com", "w")
print "Gaussian .com file ("+xyzFile+UID+".com) generated ... submitting"
jobName = "g03job"
f = open(jobName, 'w')
f.write("#$ -S /bin/csh \n")
f.write("#$ -N "+xyzFile+UID+" \n")
f.write("#$ -q "+SGEQ+".q \n")
f.write("#$ -cwd \n")
f.write("source /opt/g03/environment.csh \n")
f.write("g03l "+xyzFile+UID+".com \n")
#command to submit
task = "qsub "+jobName
subJob = commands.getstatusoutput(task)
log = "/home/"+os.environ["USER"]+"/Computation/g03jobLOG"
cwd = os.getcwd()
now = datetime.datetime.now()
if subJob == 0:
f = open(log, 'a')
f.write("Job Directory: "+str(cwd)+"\n")
f.write("Submission Time: "+str(now)+"\n")
print("...Failed : "+subJob)
To Summarize - this script conveniently takes user input and edits an XYZ coordinate file to generate the appropriate Gaussian .com input file. It then writes a jobscript , submits it to the SGE queue, and logs (appends) the date/time and directory in a single file.
At home, I have enjoyed running ubuntu 9.10 and hope to do some recording with it soon.