I'll start with the obligatory yet meaningless reason as to why this is my first blog in some time; I got pretty sick last week and I really wasn't able to accomplish much of anything. I am finally catching up on my work now and though this post will be quite outdated and possibly meaningless, I'll write it anyways, gotta keep the spirit alive right?
The readings put open source development is a slightly new light that I hadn't considered before. I have never been a huge supporter of open source development, not that I am particularly against it, but I am just not one to look at a proposed methodology and simply accept it as being a more correct or superior methodology without having given all other methods serious consideration. With that said, open source development does seem to have a very concrete and everlasting place in the software development world. Still I can't see it overtaking all aspect of software development, closed source development has just as concrete a future in the software world.
The cathedral and bazaar paper illustrated where I think open source development really thrives, where people have a common issue they need solved. With Eric's example of his fetchmail project, he made it seem that he was able to muster up such an large number of people to assist him because they wanted to see the product work, they wanted the problems that they had fixed and that they just wanted to feel like they were contributing something useful. Though I have started noticing more and more that open source seems to be more like an intern program, something you do simply to be able to put it on a resume or to start making contacts and gaining experience, be it unpaid. But perhaps my views on this are clouded, on the tail end of college and only seeing this trend in my peers desperately clamoring for employment.
There is one thing I find brilliant about the bazaar method, quite obviously its greatest strength, which is illustrated in Eric's version of Linus' Law; "that given enough eyeballs, all bugs are shallow". Any programmer knows that finding bugs and narrowing down their cause is likely to be your greatest source of wasted time when you are trying to develop something. Having your clients act as testers and developing rapid fixes for the bugs that arise is probably the best way to make the most out of your programmers time, allowing them to develop more things faster. But this does come with one risk, and that is of course taxing your clients patience, which is why the split distribution is such a good idea, having one stable and one unstable build, allowing the clients to revert to the stable versions if they simply want to work with the product without dealing with bugs.
That about raps up my thoughts on these readings. One thing I did find funny about the people in the video was that, you would think that open source figures like them would be open and (to a degree) submissive but in fact they came across as very territorial, especially towards each other.
As a last note, I filled in my wiki page with some basic information about myself, so if you want to know a bit more about me, check out http://zenit.senecac.on.ca/wiki/index.php/User:CloudScorpion.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment