Weekends are here and we are back working on the genome browser project that is sure to fail. We will write detailed notes on our thoughts as we go along. That way you can tell your children and grandchildren, why starting another genome browser project is a stupid idea 
Our goals and expectations are very modest at this point. We like the client side to use HTML5-rendering with AJAXified data import from server side. The program should self-install out of the box. We plan to release it under MIT-type license so that anyone can continue to develop it further. Lastly, we like its development to continue without any support from government grants.
Many thanks to Istvan for pointing us to chromozoom. It is a beautiful browser. When we saw it, we were nearly ready to abandon our efforts. We thought we would borrow the client side (HTML+Javascript), add our server-side modules and get done in days. Then we found out -
ChromoZoom is free for academic, nonprofit, and personal use. The source code is licensed under the GNU Affero General Public License v3…..ChromoZoom is not free for commercial use. For commercial licensing, please contact the Roth laboratory.
We do not understand the insanity of these academics. The code of chromozoom is littered with MIT-licensed functions and libraries (jQuery UI, Tokyo Tyrant, underscore.js, TipTip, ImageMagick, LodePNG and many others). JQuery UI, among them, is essential for development of the code. Why take a bunch of open-license (MIT) libraries, make some improvement and lock up the product from further development? GPL used to be cool some time in the past but not any more. None of the modern frameworks and libraries in the computer world come under GPL.
We are also wondering whether author of Chromozoom was the same Theodore Pak mentioned here. It is puzzling why Harvard was so busy catching minor infractions, when real crooks continued to do bigger damage to their institution.
Distractions aside, Chromozoom is beautiful, and we hope to reproduce its functionality. The main program is a 3000 line Javascript file with three JQuery UI widgets. The functions are here and we do not know how long it will take us to produce something as good.
Another beautiful HTML5 program is Sebastien’s Ray Cloud Browser. It is more a de Bruijn browser than genome browser, but a designer can get inspiration by going through his code. His code is GPL-licensed, but we have no complaint, because he developed it fully without using any external library.
Compare the above two programs with Gbrowse, which is the biggest piece of garbage being used by bioinformaticians today. PERL-CGI was cool in 1997, but how many internet companies still use PERL? Here is the number, and here is number for PHP. Even those companies using PERL for web-server moved on to frameworks like Catalyst, Dancer or Microlicious. Without government-backing, Growse would not have survived, and every additional year of support for it takes us back, not forward. Jbrowse is much better with its dojo-based client side, but on the server side, it is hostage to the same backdated PERL modules as GBrowse.
Chromozoom, in contrast, uses Jim Kent’s programs on the server side. From cursory reading, it seems to us that only the UCSC browser has restrictive licenses, whereas the other C codes by Jim Kent can be developed freely.
To summarize, we did not make any progress today apart from learning what others did in their genome browser projects.
If you’re not careful, Mick is going to call you a hipster macademic…
I for one believe that there is no such thing as free software – there is only software that appears to be free because you don’t see the manner you pay for it – and people rush to complain when act of payment is made visible.
For just about every major open source software that you list up there commercial entities actually support the cost of developing them and pass the cost back to us. Same with scientific funding – the software may be free but it sure cost a lot to develop and we all pay for – and it usually costs an order of magnitude more. In fact that’s why most bioinformatics software sucks – there is no continuing support to develop the code once the temporary funding runs out/or never materializes – leading to big piles of wasted effort.
If only paying for software would be more widely accepted it would solve many of the problems we currently have in science. Yet when someone attempts to find some alternative means of supporting their work others start pointing fingers. I think there is nothing wrong with licensing any software under any licensing scheme an owner sees fit. If you don’t like it then do it yourself. Note that the software is still open you can learn and understand how it works and what it does. That is still a major step in the right direction.
I think the right solution would be for you to evaluate how much it will cost you to replicate the work that the authors above did – then offer them something comparable and see if they are willing to relicense it for your use. They may or may not.
Also note that when you look at all frameworks the second most popular platform is ASP.NET from Microsoft. That choice together with Shockwave, Dreamweaver and Frontpage and their variations add up to more than 60%. So that popularity index perhaps is not indicative of the qualities you seem to want to demonstrate.
Now look at the second, third and next most popular frameworks – in a list like that being first is not necessarily a virtue
http://trends.builtwith.com/framework
hehe, good point
I don’t get your dislike of GPL. It basically just says that if you modify and distribute a derivative of a GPL program, you have to distribute the source code too. I don’t get what’s so bad about that.
Concerning ChromoZoom, I’m not sure if that “not free for commercial use” is actually compatible with the Affero GPL. As far as I know (and remember) Affero GPL is basically that if you make a website using the program, you have to distribute the source if you modify it. If you don’t modify the source, you don’t have to distribute it. I don’t think there’s anything in the Affero GPL that stops a commercial business from making a website and providing a version of ChromoZoom (as long as they distribute the source code).
Since the Roth laboratory made the program, they can decide the license, but in this case they should have chosen another license that does not contradict their message.
Ole, Anyone is free to choose what license form he wants to distribute his program under. Sebastien, for example, distributes his Ray Cloud Browser under GPL. The code is entirely written by him, and we respect his choice of license.
The software world is moving to MIT type of license from GPL type of license, because MIT type of license is less restrictive. ChromoZoom is using a large number of such MIT licensed libraries, and then packaging the whole thing with GPL. Of course legally they are free to do whatever they want to do, but I do not think it is very nice.
Let me give an (imperfect) analogy. Suppose you borrow a large chunk of my code to write a bioinformatics paper and write a paper with it without giving any credit to me, even though my code is a big part of your work. I do not mind, because I shared my code with you without restriction. Next you also make a rule that anyone borrowing part of your code to do derivative work needs to include you as an author in all their paper. That means if I take your code+mine and make something even better, I will have to include you as an author in my paper. So, you created additional layer of restriction, and that is what I am not happy with.