At 8:30 -0800 2000.11.09, Paul J. Lucas wrote:
>On Thu, 9 Nov 2000, Chris Nandor wrote:
>
>> You are arguing at one time against confusion,
>
>I'm not arguing anything.  I never started this whole argument.

Yes, you are.  You are arguing that your names should get to remain.  You
have a position and you offer reasoning and evidence to support it.  That
is an argument.


>> but here you are advocating more confusion, by changing the names of
>> established distributions.  Yes, it is a problem that his distribution and
>> module name don't match.
>
>So essentially I'm being asked to pay for his mistake of
>misnaming his module/distribution ...

If that is how you prefer to see it, then yes.  Though I don't see it as
being a price to pay.  I see it as action to take, but the action is not
negative or punitive in any way that I can tell.


>> But many people have come to know and expect it.
>
>... and the module-powers-that-be for not catching his mistake
>way back when and asking him to fix it.

It has actually been discussed before you came along, I believe, but that's
neither here nor there.  The fact remains that there was a preexisting
distribution called HTML-Tree.  Whether or not there is a module named
HTML::Tree in there, a distribution named HTML_Tree is confusing and should
not be placed on the CPAN.  That you would contest this is, to me, odd,
since your arguments against changing the name centered around the
confusion it would cause, but the existing name is de facto confusing.


>> I think the most reasonable solution is to have a dummy HTML::Tree module in
>> his distribution, thereby unifying the name with the distribution.
>
>Reasonable?  Who knows.

People who have been doing this for a long time.


> How about biting the bullet for the
>"right" solution even if it means short-term confusion and
>pain?

There is no "the right" solution.  There are plenty of "right" solutions,
including the one I gave in my last message.


>The "right" solution, IMHO, is to unify the two
>modules/distributions into a single, fast, efficient one.  The
>reason I invented my own and didn't use his in the first place
>was that his API is not as elegant as it could be (IMHO) and
>not very fast.

If Sean does not agree, then that is his business.  Since his was there
first, it is incumbent on you to deal with that by changing the name of
yours, in my opinion.

-- 
Chris Nandor                      [EMAIL PROTECTED]    http://pudge.net/
Open Source Development Network    [EMAIL PROTECTED]     http://osdn.com/

Reply via email to