[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: (Off Topic ) Open Source: The Model Is Broken ??



Tom Horsley wrote:
>>> "Open-source code is generally great code, not requiring much support."
>>>
>>> What universe does Stuart Cohen live in?
>>>     
>>>       
>> So, what is your view?   Generally crappy code in need of constant
>> support? 
>>     
>
> Uh, oh. You didn't want to get me started on this... :-)
>
> Well, seeing as how in this new F10 release alone we have the DNS
> resolver library code broken for lots of people, the NetworkManager
> being used to replace network even though it screams "not ready for
> prime-time and not backward compatible" and GDM completely
> rewritten, leaving out vast chunks of functionality, I wouldn't
> exactly call it "not requiring much support".
>   
Do you think it is valid to use F-anything as a point of reference for
"support"? 

Remember, the target audience of the article is "business". 
> But my biggest issue with the open source model is the utter lack of
> any documentation for anything. And if, God Forbid, something
> should, over time, become well documented through mechanisms like
> google searches and wiki pages, and "dummies" books, that seems to
> be some kind of catalyst for the developers, triggering a frantic
> need to utterly rewrite something that was perfectly OK, just to
> make sure it retains its traditional level of obfuscation :-).
>   
You point out a long standing problem.  Developers are generally poor
documenters.   Good documenters generally don't come cheap.  I feel the
world is populated by more good programmers/developers than good
documenters.  Or, at least, there are more programmers/developers who
earn enough from their day jobs to contribute than there are documenters.

When I hear folks lamenting the lack of documentation I often wonder
what percentage of them dedicate their time to a documentation project.

While you did end your sentence with a smiley, we all know that rewrites
are prompted by other things.  Sometimes it is as simple as the
developer realizing they should have done it differently to make it more
efficient.  At other times it is a request for a new feature and the
developer finding out it can't be done reasonably in the original
framework.  I suppose you were trying to construct an non-sequitur in
suggesting that out dating the documentation would be a driving force in
rewriting software.  I let Wiley handle those....  :-)*
*
> Even worse, the lack of documentation forms a kind of positive
> feedback loop, increasing the feeling that things need to be
> rewritten, not because they really need it, but because it is
> easier to rewrite than to understand how to modify the existing
> code.
>   
Sorry, I don't follow that logic either. 

I'm not sure if you are talking about the end-user documentation or the
documentation that should exist within the source code.  I feel these
are vastly different things. 

Certainly there have been projects have been adopted by people other
than the original developers and there have been times where the code
wasn't documented to the point where the new staff felt it was easier to
rewrite the code to provide the same functionality.

Would documentation be something the author of the BW article be talking
about when he speaks of "adding value"?  And, isn't the target of the
article the business end-user?

-- 
O.K.! Speak with a PHILADELPHIA ACCENT!! Send out for CHINESE FOOD! Hop
a JET! Mei-Mei Greshko greshko com

Attachment: signature.asc
Description: OpenPGP digital signature


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]