[Newcompiler] points for math

[Stéphane Ducasse]

Hi andrew

Mathieu told us that yesterday.

For the bootstrapping why don't we save a compiler in bytecode format  
and install it without the compiler.

Stef

> Hi everyone,
> I have hit a brick wall with the Shout stuff. There is no easy way  
> that I can
> see of emulating the existing Shout behavior when parsing  
> incomplete/invalid
> method source. So, I am going to park this for the moment, and, if  
> I have a
> 'Eureka' moment, then I'll come back to it.
>
> So, I am willing to take a look at the bootstrapping problem, if  
> that is ok with
> everyone? I am equally happy to do this either on my own, or with  
> others. If
> anyone has any thoughts on how this could, or should, be done then  
> please share
> your ideas. Anyone want to help?
>
> Cheers,
> Andy
>
> ----- Original Message -----
> From: "Marcus Denker" <denker at iam.unibe.ch>
> To: "The New Compiler, AST, IRBuilder,Closures..."
> <newcompiler at lists.squeakfoundation.org>
> Sent: Saturday, September 16, 2006 10:54 AM
> Subject: Re: [Newcompiler] points for math
>
>
>
> On 15.09.2006, at 15:51, stéphane ducasse wrote:
>
>> Hi guys
>>
>> we are discussing with math and damien
>> here are the todo list
>>
>> - introduce a double linked list on the tokenstream
>> and add some nice interface so that Gutenberg can deal nicely with
>> comments
>>
>> - better error handling
>>
>> - decompiler IR -> RB
>
> Yes, this looks good.  We should try to get the compiler in a shape
> that it can replace
> the old one, the better error handing and decompiler are required for
> that.
>
> Another problem to be solved is bootstrapping: How to re-generate the
> Parser using
> SmaCC in an image without the old parser.
>
>     Marcus
>