Proposal: Proposals are the wrong approach [prose and long]

classic Classic list List threaded Threaded
42 messages Options
123
Reply | Threaded
Open this post in threaded view
|

Proposal: Proposals are the wrong approach [prose and long]

Andrew Starks
Disclaimer:

I'm not the list's moderator or culture police. These are my thoughts offered up as advice from someone who was has proposed changes to Lua (even when I wasn't actually using it for anything) and learned that life was frustrating when I did so. 

If I'm wrong or out of line, I can handle it and won't be offended if told so, even off list. 

Short version:

Because Lua is not openly developed, "proposals" should be rare. When we feel compelled to offer one, we should seek alternatives, such as an example of something you are trying to do that is hard to get done in Lua (maybe why you are using Lua to solve this problem) and the value of a solution. (what your life would look like if it were easier)

Narrative:

If you are inclined to spark a conversation about "what would be cool," then you can expect a debate about a theoretical universe that does not exist, probably led by an opposing point of view. Those can be fun, sometimes; like arguing about rock and roll and whether or not Rush should be in the Hall of Fame. 

They are not fun when we believe that the outcome of these fantasies have weight. Then we defend them and display all of the behavior that is so easy to fall into when we argue, made worse by the fact that we are not sitting next to each other, IRL.

Perhaps worse, there is the chance that a good idea might be in that conversation, and those conversations are probably the least effective way to actually influence change.

Think of your experience as a developer. When you are making software for someone else, to solve their problem or to make them money or because you're altruistic, nothing is more frustrating than to have them ask for a feature, or worse, for an implementation detail. You want the freedom to do your best work and having constraints put on the *way* that you do it, or even a conversation about how you should do your job, is deflating. It assumes that your are a worker-bee, devoid of sufficient creativity.

On the other hand, context and feedback are food. If you're like me, you want a normative explanation (a story) of the opportunity and a sense of the context in which it is to be integrated. You want to know the value of the opportunity and you want to know how it fits into the mission of the project. That can be inspiring. With that, you can help. 

Few of us are going to have a unique and valuable insight into features that should be in Lua or changes that should be made, especially since our vision of it is personal and Lua is not ours. [unless we fork it, which we are often reminded that we are free to do: LuaJIT, MoonScipt, etc]

What we *do* have is an excellent insight into our own experience. If we can tell that story, then we are truly adding to the conversation. That story can even include ways that you're working around a design choice in Lua or a request for ideas or a hack. 

As a recovering Proposal Factory, these are some things I think before setting down to write about a change to Lua:

* I ask "Why is this so?" Nobody can argue against a question. I'm often delusional about my ability to come up with awesome questions, but I've come to expect a good answer that will change my mind and that is always valuable to me, personally 

* I remind myself that Lua is not open development, where a proposal can be considered and then voted on in some way and then taken up. Therefore, what is the point of my proposal?

* What is behind my frustration? Why do I want this? What's the value?

* Does this fit with what Lua is about? (AKA: Does it fit on a microwave?)

All of this does not suggest that new ideas don't come from new people. Instead, it's about operating within the unique culture of Lua, being effective and being heard. It's also about showing courtesy and respect to those that have likely heard it all before. 

If you have made it to the end of this email, then thank you! I'm in your debt. 

-Andrew
Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Dirk Laurie-2
2014-05-03 19:31 GMT+02:00 Andrew Starks <[hidden email]>:

> I'm not the list's moderator or culture police. These are my thoughts
> offered up as advice from someone who was has proposed changes to Lua (even
> when I wasn't actually using it for anything) and learned that life was
> frustrating when I did so.
>
> If I'm wrong or out of line, I can handle it and won't be offended if told
> so, even off list.
>
> Short version:
>
> Because Lua is not openly developed, "proposals" should be rare. When we
> feel compelled to offer one, we should seek alternatives, such as an example
> of something you are trying to do that is hard to get done in Lua (maybe why
> you are using Lua to solve this problem) and the value of a solution. (what
> your life would look like if it were easier)
>
> Narrative:
>
> If you are inclined to spark a conversation about "what would be cool," then
> you can expect a debate about a theoretical universe that does not exist,
> probably led by an opposing point of view. Those can be fun, sometimes; like
> arguing about rock and roll and whether or not Rush should be in the Hall of
> Fame.
>
> They are not fun when we believe that the outcome of these fantasies have
> weight. Then we defend them and display all of the behavior that is so easy
> to fall into when we argue, made worse by the fact that we are not sitting
> next to each other, IRL.
...
> All of this does not suggest that new ideas don't come from new people.
> Instead, it's about operating within the unique culture of Lua, being
> effective and being heard. It's also about showing courtesy and respect to
> those that have likely heard it all before.

Wow! You have been on this list only since March last year and you have
reached a position of wisdom that most of us take three years to attain.

I fully agree with everything that you say. Which makes it doubly sad that
this post of yours will not change the culture of the list one whit.

Let me explain. The new ideas, vehemently defended, come from newbies:
Lua newbies, mind, not programming newbies. Those people know a lot.
In particular, they usually know and love some other programming language
well: Python, Ruby, Javascript, C++, Ada etc.

They have often read the manual thoroughly, maybe even PiL. But they do
not spend their idle afternoons browsing back postings to the list or surfing
lua-users.org. The ons that join tomorrow won't even see your post.

On Usenet, the typical group has a re-post of its FAQ and list etiquette once
a month. This helps. Anybody who sticks it out for that long sees it.

Here is a suggestion. Make a short and snappy version of your post, give
it a title like

[FAQ] What is good Lua-L etiquette?

and instruct your computer to post it automatically once a month. Then
it should eventually have some impact.

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Karel Tuma-2
In reply to this post by Andrew Starks
Excerpts from Andrew Starks's message of 2014-05-03 19:31:17 +0200:
> tl;dr

You, sir, won my day with your wonderful sense for irony. Walls of text
are the worst, works maybe only with numbered sections RFCs.

In fact, it wouldn't hurt to tag emails [RFC], [RANT] or [PATCH], see:

> As a recovering Proposal Factory, these are some things I think before
> setting down to write about a change to Lua:

Banal and cosmetic changes, which are unfortunately majority
brought up in here, without (at least sketch) implementation is not
proposal at all, but empty [RANT].

So while I claim to be no proposal expert either, here goes my [RANT]:

While Lua is comparably small so there is no need to formalize,
bigger projects had to tame it with standardised community proposal
process (BitTorrent, Python, Bitcoin...). Even then, for simple
stuff, one is given voice only if there is either workable proof
[PATCH]/[ANN] of concept code or legible [RFC] formal spec.

tl;dr:

http://www.youtube.com/watch?v=3kuCI1dv20Y

--k

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Tim Hill
In reply to this post by Dirk Laurie-2

On May 3, 2014, at 11:13 AM, Dirk Laurie <[hidden email]> wrote:

> 2014-05-03 19:31 GMT+02:00 Andrew Starks <[hidden email]>:
>
>> I'm not the list's moderator or culture police. These are my thoughts
>> offered up as advice from someone who was has proposed changes to Lua (even
>> when I wasn't actually using it for anything) and learned that life was
>> frustrating when I did so.
>>
>> If I'm wrong or out of line, I can handle it and won't be offended if told
>> so, even off list.
>>
>> Short version:
>>
>> Because Lua is not openly developed, "proposals" should be rare. When we
>> feel compelled to offer one, we should seek alternatives, such as an example
>> of something you are trying to do that is hard to get done in Lua (maybe why
>> you are using Lua to solve this problem) and the value of a solution. (what
>> your life would look like if it were easier)
>>
>
>
> Here is a suggestion. Make a short and snappy version of your post, give
> it a title like
>
> [FAQ] What is good Lua-L etiquette?
>
> and instruct your computer to post it automatically once a month. Then
> it should eventually have some impact.
>

Slightly bending the thread .. Is this really a way of saying that the community is out-growing a mail-list as a (productive) communications mechanism? At any given time there are (a) announcements of new library etc releases, (b) long threads about suggestions for Lua,  (c) people seeking help with Lua programming and (d) general Lua discussions etc etc. To me, this starts to suggest some sort of web based forum, rather than a heterogenous mail list. Forums are also far easier to search than the archived discussions, which might help with the re-post problems.

Just thinking out loud, not sure if this would help or hinder.

—Tim


Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Petite Abeille

On May 3, 2014, at 8:22 PM, Tim Hill <[hidden email]> wrote:

> web based forum

( Arghhhhh! Sorry. Allergic reaction. )

Lua Web Forums
— Brett Kapilik, 2003
http://lua-users.org/lists/lua-l/2003-02/msg00053.html

We need a forum!
— Michael Cumming, 2004
http://lua-users.org/lists/lua-l/2004-07/msg00005.html

Why doesn't the lua list get replaced with a forum?
— George Petsagourakis, 2006
http://lua-users.org/lists/lua-l/2006-09/msg00486.html

Another meta thread ready to be crucified in the FAQ.



Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Coroutines
In reply to this post by Andrew Starks
On Sat, May 3, 2014 at 10:31 AM, Andrew Starks <[hidden email]> wrote:

IMO, more than half this list is made up of extremely biased and
discouraging 'regulars', who push down any idea that threatens their
view of what Lua should be -- without reading the full proposal and
considering its viability within the stated constraints and use cases.
 I frequently see things buried under a shitstorm of opinions, not
facts -- not data.  I have been accused many times of making proposals
that aren't backed by usage data or patches, but while I've asked for
constructive criticism unfairly I often get blind dislike of added
syntax sugar, or dislike of ideas that can be attributed to other
languages.  As if an idea can be bad simply because it is well known
from Python or Ruby or Perl.  I view most of you as destructive and
stagnatory.

And then there are some who say this is what you must expect.  This is
the type of people you have to win over.  PUC Rio only listens to what
works its way through the negativity of people fighting for minimal
zealotry.  I know most of you aren't running Lua on toasters or
working with C runtimes from the 80s -- you're full of it.

I've seen this done better, but it won't change here.

I'm not a happy person on Saturdays.

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Hisham
In reply to this post by Karel Tuma-2
On 3 May 2014 15:18, Karel Tuma <[hidden email]> wrote:
> bigger projects had to tame it with standardised community proposal
> process (BitTorrent, Python, Bitcoin...).

Because these projects are community-developed. Lua is not
community-developed so there is no multi-step "process" through which
a proposal turns into a feature. The process is that people throw
ideas around and the Lua team may adopt your ideas (but not your code,
for some reason) or not. In particular, they seem unimpressed by
arguments-by-numbers, so any kind of "voting" (including posts saying
"+1 from me", "I agree") doesn't seem to have any impact.

They've repeated time and again that ideas are considered more
carefully if accompanied with a prototype implementation that shows
its feasibility and/or compelling real-life use cases. Still, this
implies no guarantee that they will take it or even give feedback on
it. Well-prepared proposals are likely to raise good discussion on the
list, in any case.

I predict the URL for Andrew's message in the list archives will be
reposted here many times in the future.

-- Hisham

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Tim Hill
In reply to this post by Petite Abeille

On May 3, 2014, at 11:38 AM, Petite Abeille <[hidden email]> wrote:

>
> Why doesn't the lua list get replaced with a forum?
> — George Petsagourakis, 2006
> http://lua-users.org/lists/lua-l/2006-09/msg00486.html
>
> Another meta thread ready to be crucified in the FAQ.
>
>

Yes I was aware this had been suggested before .. and perhaps there is a reason for that.

—Tim


Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Philipp Janda
In reply to this post by Coroutines
Am 03.05.2014 21:27 schröbte Coroutines:
>   I frequently see things buried under a shitstorm of opinions, not
> facts -- not data.  I have been accused many times of making proposals
> that aren't backed by usage data or patches, but while I've asked for
> constructive criticism unfairly I often get blind dislike of added
> syntax sugar, or dislike of ideas that can be attributed to other
> languages.

That's the problem with syntax sugar proposals: The advantages are
purely cosmetic and subjective, and a matter of opinion. If your only
pro argument is "I like", you will get con arguments "I dislike" (and as
Hisham already said, both arguments are not worth much unless you happen
to live in Brazil). Sometimes you will also get objective con arguments
like conflicts with existing syntax, difficulty of implementation,
backwards compatibility, performance concerns, etc., in which case "I
like" is worth even less. It is a lot easier to get a constructive
discussion going if your proposal has non-cosmetic positive effects.


Philipp



Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Coroutines
On Sat, May 3, 2014 at 1:35 PM, Philipp Janda <[hidden email]> wrote:

> That's the problem with syntax sugar proposals: The advantages are purely
> cosmetic and subjective, and a matter of opinion. If your only pro argument
> is "I like", you will get con arguments "I dislike" (and as Hisham already
> said, both arguments are not worth much unless you happen to live in
> Brazil). Sometimes you will also get objective con arguments like conflicts
> with existing syntax, difficulty of implementation, backwards compatibility,
> performance concerns, etc., in which case "I like" is worth even less. It is
> a lot easier to get a constructive discussion going if your proposal has
> non-cosmetic positive effects.

I am not talking strictly about cosmetic/sugary proposals -- at the
same time I think certain sugar does help one think clearer about a
concept or how data is handled:

local a, b, c in some_table

Depending on the sugar, one can more clearly communicate an
idea/process - it's not all "I like" but sometimes "I can".

PS: One proposal of mine got through, in 5.3 userdata can be
associated with userdata through lua_setuservalue() -- eat that, lua-l
:P

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Philipp Janda
Am 03.05.2014 23:20 schröbte Coroutines:

> On Sat, May 3, 2014 at 1:35 PM, Philipp Janda <[hidden email]> wrote:
>
>> That's the problem with syntax sugar proposals: The advantages are purely
>> cosmetic and subjective, and a matter of opinion. If your only pro argument
>> is "I like", you will get con arguments "I dislike" (and as Hisham already
>> said, both arguments are not worth much unless you happen to live in
>> Brazil). Sometimes you will also get objective con arguments like conflicts
>> with existing syntax, difficulty of implementation, backwards compatibility,
>> performance concerns, etc., in which case "I like" is worth even less. It is
>> a lot easier to get a constructive discussion going if your proposal has
>> non-cosmetic positive effects.
>
> I am not talking strictly about cosmetic/sugary proposals -- at the
> same time I think certain sugar does help one think clearer about a
> concept or how data is handled:
>
> local a, b, c in some_table
>
> Depending on the sugar, one can more clearly communicate an
> idea/process - it's not all "I like" but sometimes "I can".

Often "I can" is just an alias for "I like", and you will get "I can
without" as answer -- unless you really can't without. But the process
of taking values out of a table and putting them into local variables is
very possible in Lua.

>
> PS: One proposal of mine got through, in 5.3 userdata can be
> associated with userdata through lua_setuservalue() -- eat that, lua-l
> :P

Probably because it had nothing to do with syntax :-p

Philipp





Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Coroutines
On Sat, May 3, 2014 at 2:30 PM, Philipp Janda <[hidden email]> wrote:

> Often "I can" is just an alias for "I like", and you will get "I can
> without" as answer -- unless you really can't without. But the process of
> taking values out of a table and putting them into local variables is very
> possible in Lua.

local a, b, c in some_table

^ this is both something [some] people like, and something that can
help people think about a problem more clearly/cleanly.  Others could
argue that "local a, b, c = some_table.a, some_table.b, some_table.c"
is better, but it is a valid point that proper sugar can more clearly
convey a concept.  It's not simply about 'like'.

Otherwise we could say that some_table.a, some_table.b, some_table.c
is sugar and that we should instead be dealing with Lua VM opcodes.
Sugar is not always a want, there can be a need.

> Probably because it had nothing to do with syntax :-p

I'm just saying, it was "downvoted" at the time but still made it
upstream.  It was a valid proposal that was instead shot down with
"Why would you need to?" and "You can do it this convoluted way..."  I
think the list suffers from people who refuse to recognize a need, in
favor of minimalism that stifles growth -- in any direction (even
smaller).

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Petite Abeille

On May 3, 2014, at 11:37 PM, Coroutines <[hidden email]> wrote:

>  I think the list suffers from people …

… on the plus side,  Felipe Contreras is not pestering Lua with patch requests. Yet.

"Who the f*heck is Tadayoshi Funaba and why can he reject sensible patches unilaterally?”
Felipe Contreras, 2014-05-02
https://www.ruby-forum.com/topic/4570470

Epic.


Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Philipp Janda
In reply to this post by Coroutines
Am 03.05.2014 23:37 schröbte Coroutines:

> On Sat, May 3, 2014 at 2:30 PM, Philipp Janda <[hidden email]> wrote:
>
>> Probably because it had nothing to do with syntax :-p
>
> I'm just saying, it was "downvoted" at the time but still made it
> upstream.  It was a valid proposal that was instead shot down with
> "Why would you need to?" and "You can do it this convoluted way..."  I
> think the list suffers from people who refuse to recognize a need, in
> favor of minimalism that stifles growth -- in any direction (even
> smaller).

I re-read that thread: You got one proposal to wrap a struct in another
struct, one proposal to use a single weak environment table indexed by
userdata (both purely constructive), one claim that the memory overhead
probably wasn't that bad, and even some mild support ("p.s.: The
function is called `lua_setuservalue` not `lua_setusertable`. Just
saying ...") -- by myself I might add. The rest was about defensive
programming, segmentation faults in modules, and other totally unrelated
stuff. This is hardly "downvoted" or "shot down". So I still think the
recent proposals are to blame instead of the list members.

Philipp



Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Coroutines
In reply to this post by Petite Abeille
On Sat, May 3, 2014 at 2:57 PM, Petite Abeille <[hidden email]> wrote:

> … on the plus side,  Felipe Contreras is not pestering Lua with patch requests. Yet.

That was a very entertaining read :-)

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Coroutines
In reply to this post by Philipp Janda
On Sat, May 3, 2014 at 3:00 PM, Philipp Janda <[hidden email]> wrote:

> I re-read that thread: You got one proposal to wrap a struct in another
> struct, one proposal to use a single weak environment table indexed by
> userdata (both purely constructive), one claim that the memory overhead
> probably wasn't that bad, and even some mild support ("p.s.: The function is
> called `lua_setuservalue` not `lua_setusertable`. Just saying ...") -- by
> myself I might add. The rest was about defensive programming, segmentation
> faults in modules, and other totally unrelated stuff. This is hardly
> "downvoted" or "shot down". So I still think the recent proposals are to
> blame instead of the list members.

At that time it felt like what I wanted to accomplish wasn't a valid
solution, as it could be accomplished "these other ways..." which was
pretty discouraging.  And I'm still right that you should avoid
segfaulting in libraries :P  I always typo setuservalue()*....

Anyway, I'm pretty sure it was the list members.  I flipped a coin.

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Philipp Janda
Am 04.05.2014 00:09 schröbte Coroutines:
> On Sat, May 3, 2014 at 3:00 PM, Philipp Janda <[hidden email]> wrote:
>
> At that time it felt like what I wanted to accomplish wasn't a valid
> solution,

And it wasn't at that time. The ability to use arbitrary uservalue
values will (most likely) come with Lua 5.3.

> as it could be accomplished "these other ways..." which was
> pretty discouraging.

Some (most?) people who propose a change to Lua have an actual problem
they would like solved, and a new feature can only solve it in future
Lua versions (if accepted at all), so suggestions how one can tackle the
problem in current Lua versions might be appreciated.

>   And I'm still right that you should avoid
> segfaulting in libraries :P  I always typo setuservalue()*....
>
> Anyway, I'm pretty sure it was the list members.  I flipped a coin.
>

Philipp




Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Coroutines
On Sat, May 3, 2014 at 3:28 PM, Philipp Janda <[hidden email]> wrote:

> And it wasn't at that time. The ability to use arbitrary uservalue values
> will (most likely) come with Lua 5.3.

Yes, I'm saying I identified that hole (me me meeeee) -- as you could
only associate a table to userdata, and the best you could do to save
on memory would be to share that table among the userdata you create
and make that table weakly-referenced to get from it the userdata you
would have otherwise paired directly.  IIRC, Luiz gave me that
solution, but it was the best you could do and I remember reading a
lot of replies that weren't working within my constraints for the
problem.  The main concern was memory use, as I was creating a lot of
userdata.

Then it led off on that tangent about libraries segfaulting and from
there on I was still right all the time.  Yep, rereading the thread
confirms.

> Some (most?) people who propose a change to Lua have an actual problem they
> would like solved, and a new feature can only solve it in future Lua
> versions (if accepted at all), so suggestions how one can tackle the problem
> in current Lua versions might be appreciated.

I did appreciate the [best] solution I got from (Luiz?) -- it just
took a lot of wading through "Well why would you..." at the beginning.

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Andrew Starks
In reply to this post by Karel Tuma-2


On Saturday, May 3, 2014, Karel Tuma <[hidden email]> wrote:
Excerpts from Andrew Starks's message of 2014-05-03 19:31:17 +0200:
> tl;dr

You, sir, won my day with your wonderful sense for irony. Walls of text
are the worst, works maybe only with numbered sections RFCs.

In order for me to be ironic, I would have needed to be ranting about long emails.

It would have been the height of awesomeness if you had started a rant against people posting walls of text that rail against other posts. 

You sir, missed an opportunity to be recursive. That is not ironic, but it is tragic. 

The rest of your reply was too long, so I didn't bother to read it. :)

-Andrew


In fact, it wouldn't hurt to tag emails [RFC], [RANT] or [PATCH], see:

> As a recovering Proposal Factory, these are some things I think before
> setting down to write about a change to Lua:

Banal and cosmetic changes, which are unfortunately majority
brought up in here, without (at least sketch) implementation is not
proposal at all, but empty [RANT].

So while I claim to be no proposal expert either, here goes my [RANT]:

While Lua is comparably small so there is no need to formalize,
bigger projects had to tame it with standardised community proposal
process (BitTorrent, Python, Bitcoin...). Even then, for simple
stuff, one is given voice only if there is either workable proof
[PATCH]/[ANN] of concept code or legible [RFC] formal spec.

tl;dr:

MOAR CODE, MOAAAAAAAAR CODE

--k

Reply | Threaded
Open this post in threaded view
|

Re: Proposal: Proposals are the wrong approach [prose and long]

Andrew Starks
In reply to this post by Coroutines


On Saturday, May 3, 2014, Coroutines <[hidden email]> wrote:
On Sat, May 3, 2014 at 10:31 AM, Andrew Starks <<a href="javascript:;" onclick="_e(event, &#39;cvml&#39;, &#39;andrew.starks@trms.com&#39;)">andrew.starks@...> wrote:

IMO, more than half this list is made up of extremely biased and
discouraging 'regulars', who push down any idea that threatens their
view of what Lua should be -- without reading the full proposal and
considering its viability within the stated constraints and use cases.
 I frequently see things buried under a shitstorm of opinions, not
facts -- not data.  I have been accused many times of making proposals
that aren't backed by usage data or patches, but while I've asked for
constructive criticism unfairly I often get blind dislike of added
syntax sugar, or dislike of ideas that can be attributed to other
languages.  As if an idea can be bad simply because it is well known
from Python or Ruby or Perl.  I view most of you as destructive and
stagnatory.

This really says it all, right here.  As long as you don't see people as smart, but with a different perspective you are doomed. 

 

And then there are some who say this is what you must expect.  This is
the type of people you have to win over.  PUC Rio only listens to what
works its way through the negativity of people fighting for minimal
zealotry.  I know most of you aren't running Lua on toasters or
working with C runtimes from the 80s -- you're full of it.

I've seen this done better, but it won't change here.

I'm not a happy person on Saturdays.


This makes me so sad. At the risk of being pulled into this, I'll submit this:

If I'm running thousands of Lua threads in a datacenter application, memory and processing "space" starts to look a lot like that of a "toaster." Size, speed and efficiency always matter everywhere sometimes. 

-Andrew 
123