Primary mailing list for the programming language Lua

Asko Kauppi | 1 Jan 2008 10:55

Re: Lua Lanes...alive or dead?


Lua Lanes is not dead.

The lack of updates has been due to my business, dealing with an  
upcoming Lua project. Once that is out, I can look more into the  
existing ones. There are some wishes I've received for Lua Lanes, but  
no patches or other such contributions. They are welcome. Also one  
reason to maintain the "beta" status is that I don't know of much use  
of Lanes, actually.

That can be either good or bad. It either means the module's serving  
its current users well, or it means there are not enough current users.

Anyway. Personally I think we could benefit of a "round table"  
discussion to decrease the number of multithreading solutions, and  
focus people's efforts on one. Naturally, I wouldn't be seeing Lanes  
as one of those stepping aside. :)  This is why I made the comparison  
to other solutions, which unfortunately lacks LuaTask, as has been  
pointed out. I do welcome someone to contribute a Lua Lanes vs.  
LuaTask comparison.

Open source does not mean the author should do it all by himself. Lua  
Lanes is not Lua. Patches are welcome, and also write access to the  
main subversion repository can be granted for persons who've proven  
to be worth it.

Happy new year,    ;)

-asko

(Continue reading)

headers

Wesley Smith | 2 Jan 2008 04:04

Re: Lua Lanes...alive or dead?

> Anyway. Personally I think we could benefit of a "round table"
> discussion to decrease the number of multithreading solutions, and
> focus people's efforts on one.

Hi Asko,
I tried Lanes a while back and while it seemed like a useful thing in
general, I was really put off by all of the gluax stuff in there.  I
understand your desire to maintain backward compatibility with a
minimum of effort, but I found it to be quite an effort to decrypt
what was going on in the code because of gluax.  Of course I
eventually figured it out after pouring through gluax, but I then
found the barrier to pulling Lanes out of the gluax stuff for
modification within my own personal Lua universe as to be quite high
and thus not worth the effort.  I would have been much more motivated
to submit a patch if Lanes was its own entity minus the gluax.

best,
wes

headers

Asko Kauppi | 3 Jan 2008 12:55

Re: Lua Lanes...alive or dead?


I have been thinking about this as well, and the benefits of gluax  
are not that big for Lanes. Maybe I should cut it out sometime in  
2008. At least I am not at all against that.

Just happened to be the easiest way to get it initially done.    
Thanks for your opinion. :)

And about the "more samples" wish presented by Steve, I also agree on  
that.

-asko

Wesley Smith kirjoitti 2.1.2008 kello 5:04:

>> Anyway. Personally I think we could benefit of a "round table"
>> discussion to decrease the number of multithreading solutions, and
>> focus people's efforts on one.
>
>
> Hi Asko,
> I tried Lanes a while back and while it seemed like a useful thing in
> general, I was really put off by all of the gluax stuff in there.  I
> understand your desire to maintain backward compatibility with a
> minimum of effort, but I found it to be quite an effort to decrypt
> what was going on in the code because of gluax.  Of course I
> eventually figured it out after pouring through gluax, but I then
> found the barrier to pulling Lanes out of the gluax stuff for
> modification within my own personal Lua universe as to be quite high
> and thus not worth the effort.  I would have been much more motivated

(Continue reading)