On Sat, 18 Jan 2003, _brian_d_foy wrote:
> the only thing i see there is that Tim suggested a name. i do not
> see any decision or plurality.
<G> When I applied for the Curses::* modules I was told that the absence of an
objection implied acceptance. I took the advice for the name, and went with
it. At this point I do have many users that have both contributed to it and
requested more features. I think an installed user base deserves
consideration.
> i'm was not suggesting that you dump it from CPAN, but that we try to
> work together to reduce parallel development and duplication of effort.
> i typically suggest that when i see someone submit a module that is
> very similar to another one. however, i think you have made it clear
> that you do not want to do that.
Let's make something clear: I didn't know about your module until after you
told me, which was some time after I took Tim's suggestion for a name (which
flies a bit in the face of your implication that I'm closed to outside input),
and proceeded with the module in the first place. Since that time, this
module has gain some acceptance and use from the wider community. I believe
that an installed user base deserves some consideration.
> i don't think Parse is the right name---it only describes part of
> your module, since you also write the structure back to a file.
> indeed, you could use your module for its intended task without
> ever parsing anything. furthermore, this sort of task does not fit
> in with most of the other modules in Parse::*.
Again, I whole-heartedly disagree. Spewing structures into a file is the
simple part. The bulk of the code has to do with *parsing*. Furthermore, I
don't see this demand being made of Michael Fowler's Parse::PerlConfig, so
leaving mine where it is is hardly a violation of any rules of consistency, is
it? Our modules were originally written in parallel to provide the same
functionality with the same API. Mine has evolved a bit since then, but it's
core mission remains the same.
> the task seems much more suited to either the Config:: or
> ConfigReader:: namespaces, and that is where i think you should
> register it.
So, so far I have names that two members have suggested: yours is only one
voice, and it was absent when the original subject was broached. Do you make
the final determination, or is there any kind of democratic process here? For
that matter, do I have any influence, being the code monkey? I, personally,
back the suggestion of your cohort.
No offense, brian, but if we had had this discussion the *first* time I
brought it up, it would have saved me a lot of what is starting to look like
wasted effort. I use that module, in that namespace, pretty much everywhere
these days. And I'm not the only one. I'm not relishing having to dig up
every script on every server I admin and updating it for a name change, or,
should you do what you suggested earlier and just simulate similar
functionality in your module, have to rework the code for a different API.
I agree that it appears that there has been some lapses in communication since
I first submitted it. I *thought* I was clear to proceed, and I think that we
have in many ways gone too far to just wipe that slate clear.
I'm not going to keep arguing my case. If this is your show and I have no say
in this, then tell me explicitly what my options are, or what you are (or
aren't) going to do, with both my module and yours.
--Arthur Corliss
Bolverk's Lair -- http://arthur.corlissfamily.org/
Digital Mages -- http://www.digitalmages.com/
"Live Free or Die, the Only Way to Live" -- NH State Motto
