Lua voices

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
27 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Lua voices

Dibyendu Majumdar
Most of language requests and proposals appear to come from a few vocal individuals. The vast majority of Lua users do not voice an opinion and are probably not even on the list. 

Do you agree?
Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Axel Kittenberger
I for one didn't request anything lately, only commented. Anyway, I find this "majority counting" a strange sentiment. I'd rather believe the better argument counts rather than this kind of "majority bullying" or whatever this is supposed to develop to.

On Wed, Jul 4, 2018 at 4:35 PM, Dibyendu Majumdar <[hidden email]> wrote:
Most of language requests and proposals appear to come from a few vocal individuals. The vast majority of Lua users do not voice an opinion and are probably not even on the list. 

Do you agree?

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Viacheslav Usov
In reply to this post by Dibyendu Majumdar
On Wed, Jul 4, 2018 at 4:35 PM Dibyendu Majumdar <[hidden email]> wrote:

> Most of language requests and proposals appear to come from a few vocal individuals. The vast majority of Lua users do not voice an opinion and are probably not even on the list. 

Agreed.

However, this is not unique to Lua. C++ proposals also come from a much smaller community than all C++ programmers, and votes are those of an even smaller C++ standard committee. And, in the C++ case, a proposal (not coming from a prominent committee member) would have any serious chance of being considered only if it is presented in person to the committee. I think we can find some similarities with the Lua proposal process here :)

Cheers,
V.
Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Dirk Laurie-2
In reply to this post by Dibyendu Majumdar
2018-07-04 16:35 GMT+02:00 Dibyendu Majumdar <[hidden email]>:

> Most of language requests and proposals appear to come from a few vocal
> individuals. The vast majority of Lua users do not voice an opinion and are
> probably not even on the list.
>
> Do you agree?

It is not always a virtue to keep quiet. “The only thing necessary for
the triumph of evil is for good men to do nothing.”

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Matthew Wild
In reply to this post by Dibyendu Majumdar
On 4 July 2018 at 15:35, Dibyendu Majumdar <[hidden email]> wrote:
> Most of language requests and proposals appear to come from a few vocal
> individuals. The vast majority of Lua users do not voice an opinion and are
> probably not even on the list.

Yes, though I think a good many are on the list who choose not to participate.

I maintain a relatively large Lua codebase, and haven't had any more
problems with Lua than I'd have experienced in other languages I've
used. In fact one of the best things about Lua is its flexibility to
make it how you need. We have places in our code where globals are
restricted, or sandboxed, and places where they are actively used
(e.g. config).

The Lua team brought us to what Lua is today, and I trust them to
continue with the good work and ultimately decide themselves what they
want to incorporate into the language, and to decide based on more
than the number of postings to the mailing list.

Nevertheless, good discussion is healthy, as are experiments (such as
the recent undef).

Regards,
Matthew

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Etiene Dalcol
Same here, I rarely voice opinions because I'm satisfied with how the Lua team has made decisions. Most of the things I'd like to see added on the language are very simple syntax sugars that I never considered worth proposing. Also some of the discussions are too low level for me and sometimes I just don't follow. I'm more interested generally in discussions involving developer tools and community aspects of Lua (which I think are sections that are lacking and need work), because I enjoy how the Lua team has handled the language development itself. 

Matthew Wild <[hidden email]> schrieb am Mi., 4. Juli 2018, 16:57:
On 4 July 2018 at 15:35, Dibyendu Majumdar <[hidden email]> wrote:
> Most of language requests and proposals appear to come from a few vocal
> individuals. The vast majority of Lua users do not voice an opinion and are
> probably not even on the list.

Yes, though I think a good many are on the list who choose not to participate.

I maintain a relatively large Lua codebase, and haven't had any more
problems with Lua than I'd have experienced in other languages I've
used. In fact one of the best things about Lua is its flexibility to
make it how you need. We have places in our code where globals are
restricted, or sandboxed, and places where they are actively used
(e.g. config).

The Lua team brought us to what Lua is today, and I trust them to
continue with the good work and ultimately decide themselves what they
want to incorporate into the language, and to decide based on more
than the number of postings to the mailing list.

Nevertheless, good discussion is healthy, as are experiments (such as
the recent undef).

Regards,
Matthew

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

ThePhD
As a side comment on the C++ Proposals thing: there are people who have never attended a C++ Standards Meeting and have pushed over 14 papers into the Standard.

Some of us (particularly myself) attend because the committee has separate priorities and so we have to represent ourselves. Other people present such great and good ideas that they don't have to do this at all, or get other people so excited about it because there's a present and clear need. And some people are just such good writers that their wording is perfectly apt and it just slides right into the Working Draft.

(I'm also writing papers in territory considered a "Minefield", so if I don't go my paper will just get blown up in the fury of debate and never get anywhere. Controversial things take time and dedication... time and dedication I'm not even sure I have, but will try for anyhow!)

On Wed, Jul 4, 2018 at 12:21 PM, Etiene Dalcol <[hidden email]> wrote:
Same here, I rarely voice opinions because I'm satisfied with how the Lua team has made decisions. Most of the things I'd like to see added on the language are very simple syntax sugars that I never considered worth proposing. Also some of the discussions are too low level for me and sometimes I just don't follow. I'm more interested generally in discussions involving developer tools and community aspects of Lua (which I think are sections that are lacking and need work), because I enjoy how the Lua team has handled the language development itself. 


Matthew Wild <[hidden email]> schrieb am Mi., 4. Juli 2018, 16:57:
On 4 July 2018 at 15:35, Dibyendu Majumdar <[hidden email]> wrote:
> Most of language requests and proposals appear to come from a few vocal
> individuals. The vast majority of Lua users do not voice an opinion and are
> probably not even on the list.

Yes, though I think a good many are on the list who choose not to participate.

I maintain a relatively large Lua codebase, and haven't had any more
problems with Lua than I'd have experienced in other languages I've
used. In fact one of the best things about Lua is its flexibility to
make it how you need. We have places in our code where globals are
restricted, or sandboxed, and places where they are actively used
(e.g. config).

The Lua team brought us to what Lua is today, and I trust them to
continue with the good work and ultimately decide themselves what they
want to incorporate into the language, and to decide based on more
than the number of postings to the mailing list.

Nevertheless, good discussion is healthy, as are experiments (such as
the recent undef).

Regards,
Matthew


Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Sergey Zakharchenko
In reply to this post by Dirk Laurie-2

Hello Dirk,

> It is not always a virtue to keep quiet. “The only thing necessary for
> the triumph of evil is for good men to do nothing.”

OK, since you ask for it, I think the only good idea to be drawn from this discussion is that it would be nice if, in Lua 5.2+, there would be stock means of statically detecting use of globals after an explicit local _ENV=nil (possibly a tough job). I'm using a MetaLua-based static analyzer though in Lua 5.1 use of globals can be detected using a simple shell script. So where does this put me? Likely in the "don't care" bin?

As much as I like the quote, I like using right tools for the job, so you could set up an online poll or something to reduce traffic...

Best regards,

--
DoubleF

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Peter Hickman-3
Personally I find the suggested improvements to Lua to be addressing problems that I do not seem to encounter. The perennial "how long is an array" makes me wonder what I am doing wrong to not have encountered this problem

Saying "Well it works for me" is not really a contribution so I stay silent

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Laurent FAILLIE
Hi,

I would like to see in Lua :

1/ As I discussed several times : Thread safe State sharing. For the moment, or you have to "send" bytecode among threads or use a lock mechanism which seem to penalize by a large amount performances. I would like to have an option at compile time enabling thread safe Lua, having out of the box transparent semaphore locking on global objects. Something efficient, easy :)

2/ yes, I'm missing compactness of C compoundeds operator like += *= :? and so on ... something suggested it seems several times on this list but seems always disdained by gurus.

Bye
Le mercredi 4 juillet 2018 à 22:31:28 UTC+2, Peter Hickman <[hidden email]> a écrit :


Personally I find the suggested improvements to Lua to be addressing problems that I do not seem to encounter. The perennial "how long is an array" makes me wonder what I am doing wrong to not have encountered this problem

Saying "Well it works for me" is not really a contribution so I stay silent

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Sean Conner
It was thus said that the Great Laurent FAILLIE once stated:
> Hi,
> I would like to see in Lua :

> 1/ As I discussed several times : Thread safe State sharing. For the
> moment, or you have to "send" bytecode among threads or use a lock
> mechanism which seem to penalize by a large amount performances. I would
> like to have an option at compile time enabling thread safe Lua, having
> out of the box transparent semaphore locking on global objects. Something
> efficient, easy :)

  You can recompile Lua with lua_lock() and lua_unlock() defined, but then
you get Python performance because of the "global interpeter lock" you just
introduced.

  Second, locking "global objects" isn't enough:

        do -- new local scope
          local data = {}
          spawn_thread(foo,data) -- one system thread using data
          spawn_thread(bar,data) -- a second system thread using data
        end

  You really need the fine grained locking that lua_lock() and lua_unlock()
provide.

  Easiest answer---just say to no system threads 8-P

> 2/ yes, I'm missing compactness of C compoundeds operator like += *= :?
> and so on ... something suggested it seems several times on this list but
> seems always disdained by gurus.

  Question to answer:  is

        x += b

the same as

        x = x + b

at the metatable level?  That is, will (assuming at least one object has a
metatable with the __add() function defined):

        x += b

do

        mt = getmetatable(x) or getmetatable(b)
        x = mt.__add(x,b)

or is this a "different" operator?  I only ask becuase Python has

        __add__() x = x + b
        __iadd__() x += b

  But I think the reason that Lua doesn't support this is that it
complicates the grammer or parser.

  -spc


Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Dibyendu Majumdar
In reply to this post by Axel Kittenberger
On 4 July 2018 at 15:44, Axel Kittenberger <[hidden email]> wrote:
> I for one didn't request anything lately, only commented. Anyway, I find
> this "majority counting" a strange sentiment. I'd rather believe the better
> argument counts rather than this kind of "majority bullying" or whatever
> this is supposed to develop to.
>

I would have thought it is more a case of minority bullying :-)

Actually the question I was asking is that the majority of Lua users
(in my estimate) do not complain or ask for new features. So it is
hard to know whether any proposed new feature is really beneficial or
not, as only a few voices (and always the same voices too!) are heard.

If you believe that some people are more capable than others in
deciding language features ... I agree. But to be honest, only the Lua
team would qualify as the few capable of language design. The rest of
us: do we have anything to show that would qualify us as great
language designers?

Regards
Dibyendu

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Sean Conner
It was thus said that the Great Dibyendu Majumdar once stated:
>
> But to be honest, only the Lua
> team would qualify as the few capable of language design. The rest of
> us: do we have anything to show that would qualify us as great
> language designers?

  What makes a language designer "great"?

  I really hate the notion that only a select few people are capable of
designing a language.  I don't think anyone here will claim that Bill Gates
is a "great" language designer, yet his BASIC language was arguably one of
the most successful languages in use (Apple ][, Atari, Commodore, Tandy,
IBM, all used Microsoft BASIC).  C has problems (many problems) and if the
consensus on Hacker News [1] or Lobsters [2] are any indication, anyone
using C should be arrested (at best), so that leaves out K & R as "great"
language designers.  I personally despise the philosophy of Java and Go [3]
so I don't consider Gosling or Pike to be "great" language designers [4] (I
would also throw Guido van Rossum, creator of Python, in with this
group---he too, does not trust programmers).

  Larry Wall [5] was a *linguist*, not a computer scientist and yet
people *love* Perl (or loved---it's considered "legacy" these days).  And a
lack of programming experience never stopped Rasmus Lerdorf [6] from
creating PHP.

  Anyone can create a programming language.  The trick is getting people to
*use* it.

  -spc (*I* even created a language while in college!  It's not that hard)

[1] http://news.ycombinator.com/

[2] http://lobste.rs/

[3] Don't trust the programmer with any advanced features; they're too
        dumb to understand and the advanced features will be abused.

[4] The have large corporations pushing their agenda unto us.

[5] Creator of Perl.

[6] http://www.azquotes.com/author/47278-Rasmus_Lerdorf

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Andrew Starks-2
I think it’s a great sign that we see the same few tired issues get trotted out ad nauseam. It means that there are very few real problems left to solve at the language level and people are left battling their own preconceived notion of what perfection looks like, convinced that if only their proposal for globals or lists or case statements or... would be adopted, some nebulous goal would be achieved Lua would finally be... what?

We are all the hero of whatever narrative we’re creating and it’s easy to be convinced of our own brilliance. And fortunes have been built on Lua as it is. So come with data and a real problem to solve or know that your just venting a bunch of noise.

Voices are fine, if they start with a good story about a problem that’s hard to solve the way Lua is. I think it’s ignorant to start off with statements like, “global by default is bad...” Better to come with your hat in your hand and ask, “Why?” first.

-Andrew
On Wed, Jul 4, 2018 at 18:37 Sean Conner <[hidden email]> wrote:
It was thus said that the Great Dibyendu Majumdar once stated:
>
> But to be honest, only the Lua
> team would qualify as the few capable of language design. The rest of
> us: do we have anything to show that would qualify us as great
> language designers?

  What makes a language designer "great"?

  I really hate the notion that only a select few people are capable of
designing a language.  I don't think anyone here will claim that Bill Gates
is a "great" language designer, yet his BASIC language was arguably one of
the most successful languages in use (Apple ][, Atari, Commodore, Tandy,
IBM, all used Microsoft BASIC).  C has problems (many problems) and if the
consensus on Hacker News [1] or Lobsters [2] are any indication, anyone
using C should be arrested (at best), so that leaves out K & R as "great"
language designers.  I personally despise the philosophy of Java and Go [3]
so I don't consider Gosling or Pike to be "great" language designers [4] (I
would also throw Guido van Rossum, creator of Python, in with this
group---he too, does not trust programmers).

  Larry Wall [5] was a *linguist*, not a computer scientist and yet
people *love* Perl (or loved---it's considered "legacy" these days).  And a
lack of programming experience never stopped Rasmus Lerdorf [6] from
creating PHP.

  Anyone can create a programming language.  The trick is getting people to
*use* it.

  -spc (*I* even created a language while in college!  It's not that hard)

[1]     http://news.ycombinator.com/

[2]     http://lobste.rs/

[3]     Don't trust the programmer with any advanced features; they're too
        dumb to understand and the advanced features will be abused.

[4]     The have large corporations pushing their agenda unto us.

[5]     Creator of Perl.

[6]     http://www.azquotes.com/author/47278-Rasmus_Lerdorf

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Jay Mithani
Many Lua users picked up the language due to the power of LuaJIT, which (I believe) has limited or no support for the more recent Lua versions. Perhaps this is why some users are reluctant to participate in these discussions?
Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Russell Haley
In reply to this post by Dibyendu Majumdar


On Wed, Jul 4, 2018 at 7:35 AM, Dibyendu Majumdar <[hidden email]> wrote:
Most of language requests and proposals appear to come from a few vocal individuals.
I think this is true of many things in life and the Lua team has picked a wise course of action regarding contributions. Who was quoting Bruce Lee lately? 
 
The vast majority of Lua users do not voice an opinion and are probably not even on the list. 

Do you agree?
I speculate that the vast majority of Lua users are writing scripts for games. There are literally millions of kids playing games that are scripted in Lua. 
I was using the Roblox Studio IDE and it's beautiful. Really slick autocomplete. Writing scripts in Roblox and the questions kids are asking on the various forums are in a different category than the questions I see on this list. Those kids aren't interested in language semantics (though I am now understanding what Soni is onto with Cratera), they need help with the game API and technical issues.

Most games are still using a 5.1 API. Just to mix posts, Cloud Wu pointed out the possible memory footprint improvements for large servers. I can't help but wonder if GC and performance improvements were directed at coercing new games to use 5.4. For some time I have wondered why I don't see more game and game engine people posting on this forum but perhaps the 5.1/5.x schism is why? I assume that these large companies are in contact with the Lua team and the engine implementers' would be well known in the Lua community *somewhere*. 

Personally I just really like Lua. Its refreshing coming home from a day job slinging Java or C# and writing some nice little scripts that do the same thing with no threads and 1/10th the memory.  I have to admit to myself thought that I spend more time opining on the mailing list and trying to build some weird library than I do scripting. But that's why I'm on the list, to find those fun libraries and learn about the language semantics.  Each to their own. :)

Russ
Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Axel Kittenberger
Most games are still using a 5.1 API. Just to mix posts, Cloud Wu pointed out the possible memory footprint improvements for large servers. I can't help but wonder if GC and performance improvements were directed at coercing new games to use 5.4. For some time I have wondered why I don't see more game and game engine people posting on this forum but perhaps the 5.1/5.x schism is why? I assume that these large companies are in contact with the Lua team and the engine implementers' would be well known in the Lua community *somewhere*. 

AFAIK most game implementations stick to LuaJit and thats why the 5.1-ism.
 
Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Marc Balmer
In reply to this post by Dibyendu Majumdar


Am 04.07.2018 um 16:35 schrieb Dibyendu Majumdar <[hidden email]>:

Most of language requests and proposals appear to come from a few vocal individuals. The vast majority of Lua users do not voice an opinion and are probably not even on the list. 

Do you agree?

Yes.

Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Marc Balmer
In reply to this post by Jay Mithani


> Am 05.07.2018 um 06:19 schrieb Jay Mithani <[hidden email]>:
>
> Many Lua users picked up the language due to the power of LuaJIT, which (I believe) has limited or no support for the more recent Lua versions. Perhaps this is why some users are reluctant to participate in these discussions?

Maybe your statement is just not true and not so many Lua users picked up the language due to the power of LuaJIT.  Maybe LuaJIT isn't even that powerful.  Maybe it is not needed at all, because you can write code in C and mix it with Lua very well.


Reply | Threaded
Open this post in threaded view
|

Re: Lua voices

Sam Putman


On Thu, Jul 5, 2018 at 8:51 AM, Marc Balmer <[hidden email]> wrote:


> Am 05.07.2018 um 06:19 schrieb Jay Mithani <[hidden email]>:
>
> Many Lua users picked up the language due to the power of LuaJIT, which (I believe) has limited or no support for the more recent Lua versions. Perhaps this is why some users are reluctant to participate in these discussions?

Maybe your statement is just not true and not so many Lua users picked up the language due to the power of LuaJIT.  Maybe LuaJIT isn't even that powerful.  Maybe it is not needed at all, because you can write code in C and mix it with Lua very well.



I did. It's that powerful. Needed? Is any language needed, really?
12