#send in 1.9

[Trans]

On Aug 8, 9:11 am, James Edward Gray II <[email protected]>
wrote:

James,

Like most Prag fans I've been dipping into Erlang a little lately.
One thing I noticed about their world was that they have one mailing
list that was just for discussing proposals to change the language.
I thought that was neat idea when I saw it. For some reason it feels
less overwhelming to me.

That is what I have been suggesting for years now -- a dedicated
mailing list and a wiki for the community to jointly develop RCRs on.
It also wouldn't hurt if there were some sort of stated protocols,
about what happens if an RCR gets substantial support by the
community.

T.

 

[Robert Dober]

On Aug 8, 9:11 am, James Edward Gray II <[email protected]>
wrote:

James,


That is what I have been suggesting for years now -- a dedicated
mailing list and a wiki for the community to jointly develop RCRs on.
It also wouldn't hurt if there were some sort of stated protocols,
about what happens if an RCR gets substantial support by the
community.

T.

I guess UR right a Wiki would just do great
That at least I have learned from this discussion ...
Robert

 

[Phrogz]

I guess UR right a Wiki would just do great
That at least I have learned from this discussion ...
Robert

Interesting - I disagree. I very much like the idea of a single
champion with his/her own personal vision for an RCR. I'd not like to
put forth a well-thought-out design, and have some random chump come
in and decide that he personally thinks it should be different and
migrate my RCR to something different. Sure, I could vigilantly watch
my page and fight with the misguided fool in version control, but that
would turn me off more than the current system.

I like the idea of a single mailing list. I like the idea of a single
discussion thread (nested or not) per RCR, where modifications to the
proposal may be suggested. I don't like the idea of letting everyone
put their grubby little hands all over my spec.

 

[Trans]

You're right in that there's nothing really wrong with this
discussion. My personal feeling is that Trans's solution to most
situations is to change Ruby to fit his world view. I feel like it
should be the other way around most of the time.

I admit that in the past I haven't been as self-critiquing of some of
my thoughts as perhaps I should, prior to posting them to this list.
But that's been quite some time ago. Nowadays, my posts in this regard
and fewer and a bit more thoughtful. Nonetheless, I still post more
than others on these matters b/c it interests me. Unfortunately there
has never been a dedicated forum for these discussions. I always get
flack on ruby-dev: "this is the not the place for this discussion",
and I get flack here for being too much noise. I just have to accept
that I am an individual, with my own approach to things, and often
enough, for whatever reason, they're "controversial". Nonetheless, I
like to think my individualism adds, on the whole, positively to this
collective.

T.

 

[Trans]

Your prose got too flowery for me here. I looked up haptic, as I
suspect you intended, and I still have no clue what you said.

Sorry, by "haptic" I meant the "hands-on" nature of the system. Maybe
"ergonomic" would have been a better word. In any case, I simply mean
that the system is not inviting b/c it is not conducive to the
process.

T.

 

[Trans]

Interesting - I disagree. I very much like the idea of a single
champion with his/her own personal vision for an RCR. I'd not like to
put forth a well-thought-out design, and have some random chump come
in and decide that he personally thinks it should be different and
migrate my RCR to something different. Sure, I could vigilantly watch
my page and fight with the misguided fool in version control, but that
would turn me off more than the current system.

I like the idea of a single mailing list. I like the idea of a single
discussion thread (nested or not) per RCR, where modifications to the
proposal may be suggested. I don't like the idea of letting everyone
put their grubby little hands all over my spec.

Oh, I agree with you. There would have to still "owners" of an RCR. If
the honor system wasn't enough --and I guess ultimately it's not, then
the admins of an RCR would have to grant permissions for others to
work on it.

T.

 

[Gregory Brown]

Interesting - I disagree. I very much like the idea of a single
champion with his/her own personal vision for an RCR. I'd not like to
put forth a well-thought-out design, and have some random chump come
in and decide that he personally thinks it should be different and
migrate my RCR to something different. Sure, I could vigilantly watch
my page and fight with the misguided fool in version control, but that
would turn me off more than the current system.

I like the idea of a single mailing list. I like the idea of a single
discussion thread (nested or not) per RCR, where modifications to the
proposal may be suggested. I don't like the idea of letting everyone
put their grubby little hands all over my spec.

+1, right on.

 

[Robert Dober]

Interesting - I disagree. I very much like the idea of a single
champion with his/her own personal vision for an RCR. I'd not like to
put forth a well-thought-out design, and have some random chump come
in and decide that he personally thinks it should be different and
migrate my RCR to something different. Sure, I could vigilantly watch
my page and fight with the misguided fool in version control, but that
would turn me off more than the current system.

I like the idea of a single mailing list. I like the idea of a single
discussion thread (nested or not) per RCR, where modifications to the
proposal may be suggested. I don't like the idea of letting everyone
put their grubby little hands all over my spec.

Sure but it depends on the community, if we have the same community
than here I feel that a Wiki is a good choice and that the problems
you are afraid of are unlikely to happen.

And it can be exactly as what you liked in the thread based
discussion, the RCR will be writeable by the author and some admins
only, and everybody else will just join the discussion page of the
RCR, what'd you think about that format, I feel it would be one
possible approach.

Cheers
Robert

 

[Robert Dober]

+1, right on.

funny you say that, because you really seemed to be annoyed with this
behavior, or are you thinking about ruby-core, please kindly evaluate
the format I suggested in my reply to Phrogz.

 

[Robert Dober]

I apologize. I should have been nicer.

You're right in that there's nothing really wrong with this
discussion. My personal feeling is that Trans's solution to most
situations is to change Ruby to fit his world view. I feel like it
should be the other way around most of the time.

His regular change requests have led me to consider them more noise
than signal, though I shouldn't have taken that out on this thread.
Again, I apologize.

Getting back on topic: I feel as I have already stated that send()
and funcall() are on the right sides of the equation. send() sends
messages to an object and I feel that should be treated as a normal
method call, ignoring the private stuff.

funcall() was selected because Matz sometimes refers to receiverless
method calls as a "function style" syntax. We have at least
module_function() in the language today as another sign of this.

Like David Black, I don't care for the name and would prefer send!
(). The bang is suppose to indicate a dangerous alternative and
using a send()-like tool to bypass method visibility feels dangerous
to me. You better know what you are doing.

Regardless though, I can't build any rational for reversing them,
beyond backwards compatibility with the current send(). While that's
a noble goal, 1.9 is known to break compatibility when needed and if
that leads to a better thought-out API, I'm for it.

funcall() makes zero sense on the other side of the equation, so now
we need a name change too. To me, that's one of the signs that this
suggestion is on the wrong path.

That's just my opinion though. I could be wrong and I definitely
don't make these decisions.

James Edward Gray II

Wow now things make sense again, thanks for your clear and brave
words. I guess that even Tom will appreciate them, but he'd rather say
that himself.

Ok I did just not share that POV, I feel that Tom is a dreamer,
dreaming out loudly, for me the signal/noise ration was good but I am
a dreamer too .

Please note that he gave in very fast, I really think he did no
lobbying. But I understand now why you(plural form) are annoyed with
this post.

May I make a suggestion, let us be a little more liberal about
expressing wishes and visions on this list and OTH post to ruby-core
when we are serious, than of course on ruby-core this discussion would
have been annoying,
While on this list people not interested could have ignored it without
being afraid that something is going to change the language.
Well with being less afraid at least

Just a thought of course.

Robert

 

[Gregory Brown]

Please note that he gave in very fast, I really think he did no
lobbying. But I understand now why you(plural form) are annoyed with
this post.

James said lob. Not lobby.

As in "Lob a big old slab of Chunky Bacon over to those Cartoon Foxes.
They look hungry!"

 

[Robert Dober]

James said lob. Not lobby.

Sorry James, I thought it was a neologism for lobby, funny these languages...

As in "Lob a big old slab of Chunky Bacon over to those Cartoon Foxes.
They look hungry!"

Nice example indeed in the Ruby Context.