All things must come to an end

Well it looks like this is going to be the last blog post for my CSCI 462 class. The night before the rough draft of the presentation was due me and Andrew stayed up late messing with it and it is well on its way to being finished just a few little touches to throw in. I would say that this class did a very good job of opening my eyes to just what is open source. Since that word is thrown around to describe almost everything new in computer land.  I still hate it. I think that it is a good process for small projects or major ones that have a lot of buzz around them. For most of these projects they need to be a closed group with proper structure to get the job properly done. I understand it place but I don't like it. BOOOO. I say. Booo.

Prezi Time

Last time our group meet we started working on our presentation on Prezi. I made a very basic shell and Perry and Andrew put a lot of work into it over the weekend and it's coming along nicely. I would have done more but it the end of the semester crunch time and I had a term paper due this weekend. It also looks like its time for classmate appraisal time. So far none of my team mates have brought me sacrifices to sway my decisions. We'll have to see how this bodes for them. Muwhahaha

RTFM

So it turns out there was a modification to the cmake cache that was mentioned in the forums but I had dismissed it because I don't remember having to make that change before. Well it turns out that was at least one of my problems. I'm not sure if all the other tricks that I was doing also made it start working but turning off pthreads (what ever that does) was the final change that got it working again. One major gripe is that his info should be apart of the install process and not something that I should have to find out through forums and troubleshooting.

And you though it could only get easier

So I think I may have found one of the culprits thats been messing up my install, Norton Antivirus. So either ITK has been tricked into making one of its files into a virus or Norton is up to what it does best never catching real viruses and only messing up what legit things you are trying to do.

Super Happy Fun Time

So I stayed home today because some how somewhere I messed up my install of ITK & VTK. It has just hit midnight and I'm still fighting with this thing.  It wouldn't be so bad if it didn't take 2 hours of recompiling time every time I made a change. Makes me kinda wish that I had some sort of subversion for my whole computer and not just the git repository.  I do have a backup of my computer but its from mid last month and a lot of stuff has changed since then. So if all else fails I'll resort to that. As far as the other parts of this project they seem to be moving along and next week we plan on getting our presentation together. I suggested using prezi for our presentation but we haven't made a final decision yet.

Big bug progress

There has been some new break throughs on the big bug. I found a way to resize the canvous in Photoshop and that help us get the right different sizes to test our fix. I was already able to get it to load a 500 mb file but have been having trouble finding a program to covert a tif to png. Then I did a search on vtk website to see what other formats they support and found that with some tweaks you can get it to read tiff's. So that's the next step in the project. I also just realized when righting this that I am behind on my bug to submit.

Making progress

I helped Matt submit his bug for approval after our meeting on Tuesday. The good news is that we finally have some bug fixes getting approved and implemented in the main build. I have the next bug to fix. I have not started on it yet but should get going on it later tonight and also get working on our picture issue.  The new problem in fixing this bug is building the code to make sure that we can test the fix since the original poster seems to not be following the bug anymore. I feel if we can not make any head way in the next week that we need to ditch this bug and find another one.

POSSCON Aftermath

Well I'm bummed my letter to win a Samsung Tab didn't make it to them in time. It got sent back with a return to sender stamped on it. I'm also going to
Go back to the privilege post and add peoples names. I did get to talk to mad dog and he was really an approachable Guy with great stories. He talked about a program he was working with that he needed to optimize and he was able to do that by replacing one card in a stack. That one card change shaved off 6 hours of computer time. He also seemed to be one of the few presenters that seemed to really know what they were talking about. Many of the other presenters sound kinda like me and my fellow classmates when presenting our projects. We know enough about what were doing to get the job done but don't really know what's going on under the hood. I also talked to Jim McCracken about 3d fabrication nothing to interesting to add on top of the presentation. The last person I talked to was John Diamond. It sounds like he has taken on a huge task rewritting the game engine and also doing the art work. That also seems to be the down fall of the project it seems that if he could get some more people involved he might end up with a better product but he would lose some of his control.

Live blog POSSCON

For POSSCON I'm going to try and do a live blog from my phone through out the day.

7:00 am
     I woke up and started getting ready for the day and headed out to pick up Ash in the way out of town. Wasn't quite sure what to ware so I went with regular school attire.

8:34 am
     Well I finally made it to Ash's so if you see a yellow streak on 26 that's us.

10:32 am
     Just arrived so it looks like we missed opening comments but we got to talk to the 3D printing guy (Jim McCracken) and the open source gaming guy (John Diamond) before the seminars.

11:40 am
     Decided to change up from seeing "mad dog" and see the 3d printing presentation.

12:02 am
    The 3d printing was pretty cool. I might end up making one for my dad for car projects. For lunch ended up eating with the sugar guy (John Mertic) but turns out that he's not the same sugar guy that the other groups are talking too. It turned out to be SugarCRM that he was representing. So I kinda ended up looking like an ass talking about too different systems and neither of us really understanding why the conversation wasn't making too much sense.

2:00 pm
     Skipped the talk after lunch but decided to go to the big data talk. So far not as interesting as I thought it would be. The presenter (Nathan Marz) somehow figured out how to give a generalized speech about hoodunk with specifics and the audience still walked away wondering what they had just listened too.

3:00 pm
     Went to the starting a new business with open source (David Duggins) and it is pretty informative by not getting bogged down with specific tech babble. The down side to this approach is that people in the crowd saw it as weakness and start asking questions about software that would be used to show how much smarter they were then the speaker.

5:03 pm
    The lady (Mel Chua) that talked about sugar labs had a really different and cool way of realizing how confusing jumping into open source. From the beginning she tried to make things confusing by moving people around and rearrange people based on their skill set. I like this approach when she first started it my first reaction was to run away from the madness but in the end I realized how close it was to the real first experience of jumping into the open source project.

8:04 pm
    Finally back home. Time to clean this post up and then enjoy new phone Thursday.

POSSCON speakers

Our assignment for POSSCON is to chase down several speakers and ask them some questions and see what they have to say. This will actuall be one of the hardest assignment in this class for me. I'm not much of a meet and greet kind of guy. So I will definatly have to come out of my comfort zone to talk to these speakers



Jon 'maddog' Hall is going to be one of the people that I chase down to talk to. From the description from the website sounds like he has done a little bit of everything. I think he would be a good person to ask "So where do I go from here" meaning now that I have a degree what are the next steps that I need to take.

Corey Atmos Donohoe is another speaker that looks like he has an interesting side life to his programing life. He also works with github and I would like to better understand how that system works.

Jehiah Czebotar is the lead engineer for bit.ly. I would be interested to see if that whole idea of shortening url's started out as a fix for a problem that him and friends were having then decided to turn it into a business or if it was a I know how to make money idea.




As a side note I would like to say that I'm pretty proud of my self for being able to solder a transiter back on my mother board after trying to install a gigantor video card went bad.

Spring Break Fun

Over spring break I installed Ubuntu on a stand alone machine to try and get our project running in an Linux environment.  I also wanted to be able to help Perry with his build. The experience of installing this in a windows environment compared to Linux is night and day. The major difference is the use of cmake. In windows it plainly asks you what is the source folder, what is the destination folder and what options you want to change. Which made this part very easy to get going on coding. The Linux version of cmake and its CMakeLists.txt files all assumes that you know all the inter-workings of what you are dealing with. It mentions library output paths and cmake preinstall directorys and other things that sort of sound like they might be good things to change to get it working but it doesn't seem to work like that. Following what instructions that they have on the wiki don't help either. The crazy thing about it is the group that created cmake is the same group that created VTK. You would think that would make this process straight forward.

I messed with it some more thinking I had found some step by step guide but the libraries are still not linked properly for the test project to fully compile. I'll have to talk to Perry to see what he has done and maybe we can  combine are methods and get one good working example.

Getting everybody caught up

Our team got together to get everyone caught up on how to get the active repository, make changes and then submit them for review. Andrew made a fix to one of the source files and has submitted it for review.  I haven't heard back from him yet on how the bug submission went but I'm guessing it should be accepted. We also came up with a timeline to get all of our bugs submitted. On a side note my bug is still in limbo.

Bug submission

Through further discussion with the mailing list it appears that there is supposed to be a script that creates an alias in your linux system and that is why the commands they were giving me didn't work.  It also turns out that they have a dedicated macro for warnings. I guess I should get some extra points for implementing the same name as what they have there macro named.  They have asked for me to modify my code to incorporate that macro instead of the function that I made and then it will be reconsidered. It took awhile to figure out how to comment on my own submission.

For our next stage were going to work on several simple bug fixes to basically get the rest of the team up to speed on how to work on the program. Then were going to start working on a more involved bug or work on knocking out more minor bugs.

TOS 7 Exersises

7.2.2 I used the diff commands to see what had changed between the two files. It appears the -u flag makes the outputted patch file a bit more robust with its info. This is going to be very handy in keeping track of what changes we make just with in our group and also with the main branch of the repository.

7.8 I felt this exercise was the same as the previous one unless he was trying to get you to go through the whole process from start to finish. Which I did do. It helped seeing what it looks like from the other side of the patching process.

7.9 This one was more of the same but it was with real code instead of some made up stuff.  I will look into our current project and see if this is one of the preferred methods of submitting.

Squashing bug the hard way

Well it turns out some times it is better to know the basics before you start using those holding your hand IDE's like eclipse.  So I spent quite some time trying to figure out exactly where a call was coming from.  For instance there is a call stlWriter->write(); Well write() is not in the vtkstlwriter source code it is actually in some other .h file that program does like 5 hops through to get to the final call. Then there was the whole issue of not knowing that when I changed the library's in another directory that I would need to recompile them every time I touched them. I figured since the test program was just using the .h files and not the compiled files I wouldn't have to do that, wrong.  Once I figured that out I finally knew why I couldn't do a simple cout statement.  Then It was a simple if statement that if the user pushes to many points it would print out a warning and also a way to get around it if necessary. I made the warning statement separate so that it can be maintained down the road.

Now that I've got all of this under my belt now I'm not as scared of VTK anymore.

Alumni Symposium

I was going to try and do a live blog cast during the event but I ended up sitting too close to the presenters. I didn't want to appear as though I was texting on my phone the whole time.

I found the event very informative.  I now know that I need to add that I play some musical instruments to my resume.  Also to develop some apps on my own to show programming skills out side of the classroom.

Working on the bug is progressing but slowly. It should come together in time for our submission.

Revised team bug

So after last class our group realized that our bug was a little too simple and that we needed to pick a new one. We all went through the bug list together and found one that we think we can takle.

Freeciv Final Chapter

Well in an attempt to install gcc with support for c++, Ubuntu will no longer boot up. So I'm throwing in the towel on compiling freeciv. The documentation all around was bad. The TOS book made jokes about how confusing that documentation can be and how a capital letter can throw the whole system off. The real joke is how it's own commands don't work with out some massaging. I'm really surprised nobody piped up when you asked about some changes that need to be made to the TOS book or at least have them tweaked to the Ubuntu system.

This is just one of my many problems dealing with Unix base operating systems.  They all claim to be interchangeable but in the end its not. You have to install all these different packages and then there's the underlying shell. Then there's the online "help" that for the most part somebody will try to give some sort of answer but it will be for some other flavor or some package you don't have installed. Then you go to get clarification and the trolls take over the thread and nothing gets accomplished.

A good example of the difference between Unix and windows is our groups project ITK &VTK. I was able to get the source and compile it with in about 3 hours devoted time. Perry has been struggling to get it to compile on Ubuntu for the past 2 weeks.

When ever I start messing with Ubuntu I always think of this quote from Samuel L. Jackson from Jurassic Park:

computer: "unh-uhn-uh, you didn't say the magic word."

Sam: "PLEASE! GODDAMMIT! I hate this hacker crap!"

Freeciv "The Saga"

So I'm still working on getting freeciv to build correctly.

I started out with trying to run it on a Ubuntu on a usb stick.  It turns out that version is does not run will all parts and does not let you install new parts. So I fought with that for awhile then I saw they had install directions for installing it in cgywin. So I got that installed on my windows 7 machine. The problem that came up after that is that since the auto-configure script is stored in a windows file structure and then getting read in a Linux environment it does not properly interpret the chars in the text file. I made attempts to fix the chars but then it would not recognize the proper end of file. So I gave up on that and tried to use virtual pc to load an Ubunto image. Then with its emulated system it's some other architechture then what the images are looking for so I couldn't get either one of those to work. Currently I'm backing up my laptop before do a dual boot. Since alot of people in the class were complaining of it knocking out their windows install.

So we'll see if I can make this work in a full up Ubunto install.

Fingers crossed.

TOS 6 Exercises

6.4 Issue 2 They found a problem with a certain sized image and they found a work around but nobody has tried to fix it yet.  This problem has been around since 8/13/03.  

6.5 I have already made an account back when we were trying to track down some bugs for us to fix.

6.6 For this exercise i was not able to reproduce a bug due most of the bug reporting is done in a way that they think that the person reading it knows every part of the program and knows the exact area that is wrong.

6.7 For this one I did not find any duplicate issues or at least with my current ability to navigate the bug tracker.  I notified one bug submitter that I would be checking out his bug and will be getting back to him to let him know if I was ever able to reproduce it.  The categorizing is somewhat confusing at first. A minor bug to them seems to mean not critical to the program to be fixed instead of it meaning that it will be an easy bug to fix.  All of the bugs that I checked out had the right version numbers and mentioned the right area of the program to be fixed.

Building the mystery program

Well there are two ways to download the source code. The easiest way is to download the zip and be done with it. The other way is the proper way if you plan on contributing to the source code and is much more involved.

The second way you start out downloading git so that you have the version history and the most up to date version. Then sign up for gerrit. With gerrit access you also have to set up a ssh key on your computer that you share with them so that you can have a authentication handshake when making changes to the repository.

Once that is all done your still not ready to compile the program. You have to download another program they make call cmake. It allows you to configure what parts of the itk that you need and then outputs the result to another directory. Then you go to that directory and type "make" then it sprinkles some more magic dust on the code and should give you some .exe files but they don't work on their own. Which leads to another one of our lessons from this.

So after doing all this it helped us realize that itk by it self is not that useful, it needs to be teamed up with one of their other programs. So our project is now going to be itk + vtk which will let us feed itk an image and vtk will display it with some tweaks. Also it turns out we weren't looking at the bug tracker correctly all the bugs that we saw were for there other programs. Itk actually only has one bug reported at this time but vtk has several. We also realized that when they labeled a bug with severity that it was how important it would be to fix that bug not how hard that bug would be to fix. So no easy "love bug" designation for us.

Once you have itk and vtk compiled your still not done. You then have to write your own c++ code to incorporate the different parts of those two and make a cmake.txt file. Then compile that and get a .exe file. Then your finally done and you get to see "hello world" on your screen.

Much Success

When I first started messing with subversion I was having issues getting new files to sync with the system. The command line system was not being my friend at all. So in frustration I decided to do a Google search to see if there was a better way or a GUI based version.  I was able to find TortosieSVN and now I am completely happy with the system.  I was able to track name changes to files and content changes.

I was also able to download the source code for the ITK project still not quite sure what to do with the finished project but at least I'm on the right track for now.

Project Bugs

You can visit our history project page here. The organization started out in 1999. It has grown significantly since then and has many new contributers every day. It has 13 million in public contributions. The first bit of code was submitted on march 2000.  They have 82 unique developers. They have over 1380 different people on the mailing list. This is all by the version 3.4. The current version is 4.0. Here is a link to the pdf where I got all the number stuff. Link

We also have some simple bugs that we should be able to take care of and then we have one major one that we have a guy that would like to mentor us in getting that bug taken care of. I'm looking forward to seeing what we can contribute to this group.

Well it turns out that I need to learn how to read a table of contents.  I thought the first page was the whole part and that each page was a new article.  Now that I've got that figured out here is my addendum to my previous post.  This does seem to go against the logic of the man-month by the progress of a FOSS being sped up by having more contributors.  I think the big difference which was touched on in the paper is that these are people who weren't just hired to fill a slot in a job but people who wanted to work on this project. So they have a desire to put more effort then just enough to not get fired.

I also liked his hesitation on saying that Mozilla was going to be a successful venture.  I think that if he went back to revise the paper one last time he could put that hesitation aside.

The Cathedral and the Bazaar

What I got from this article was that the switching from a very structured system to a multi-cell style system caught not just the author but many people in the business off guard.

One thing that has always bothered me when reading these article from these "professionals" in Computer Science is how they all seem to think that things will never change or change at a very slow and manageable pace.  Even when things were starting to get off the ground in the 50's it was changing very fast if you compare how fast computer programming changed over the past 100 years at that point. It will just continue to pick up speed to where even now there is only a few keystone products out there but everything else can change up within 6 months or less.

Testing out mobile posting

Project begins

So we were able to get the project that we wanted and we have started on getting the source code to compile. We created a Google group to create a email address that would get to all of us but for some reason I can't get on it yet.

We dug around in the groups wiki documents and found a mailing list but no irc chat channels yet.  On interesting find was that they use second life to communicate. It also appears that they use Twitter and Facebook but we need to dig deeper into it to make sure that it is a real form of communication.

Project Ideas

In class today we talked about several different projects that we could use.  We actually found it quite difficult to find a project that made it easy to contribute to their open source project.  We ended up coming up with 5 different projects.

Vufind
Tor
itk
plone

Hopefully nobody runs off with our project ideas before us.

Signing up for services

Well I attempted to sign up to the required services today. Not sure if it all worked out correctly but I know I will find out tomorrow in class.