Gjorgji sent you a private message :)

Re: fixed point type

Re: fixed point type

ok

and what about adding the infinite integers ?
it in your plans, isn't it ?

On Tue, Sep 23, 2008 at 7:46 PM, C H Forsyth <forsyth@...> wrote:
>>or to compile calls to C-library.
>
> we wanted them where the JIT could see them.
>

fixed point type

Hello

It seems the fixed point type was added to Limbo after the
documentation had been written.
Could anybody provide an example how to use it ?

Is there a regression test suite?

Hello

Is there a set of regression test for Inferno ?

Compiling it by different compiler for different platforms without a
test suite can produce long-lived bugs.
I have tried to compile Inferno for openmoko (ARM/Linux) using gcc-4.3
and found a few bugs there, although Inferno and most of applications
can start on that platform.
For example,
conversion between int and big types (as well as '%b' in sys->sprint)
does not work correctly because GCC assume stack is 8 byte aligned,
registers used by jitted code are not saved in stack so jit does not
work at all,
etc
I guess there are a lot of bugs remainded and some test suite may help
to find them.

Process explosion

 Hi guys,
 I came home today to a mailbox full of cron jobs failing because the
"couldn't fork". Sure enough, netbsd has a conservative process limit
(160), and I'm running pretty close to it - primarily because of the
130(!) emu processes.
 I only have a single emu instance running, and within that I'm
serving some instances of ircfs. ps within emu gives 45 processes
which look believable - I don't think there's anything funny going on
on that side of the fence.
 So, my questions are:
1) is this behaviour expected?
2) is there a way to limit the process growth?
 I've upped my limit for the moment, I'll see if the number of procs
keeps growing or not...
-sqweek

Polymorphic pick adts

Hey all,

I stated getting limbo failures when I tried add a pick structure to a
previously declared polymorphic adt. From a cursory look at the
compiler it looks like pick adt's can't be polymorphic. Am I right?

Noah

RE: EMU 'usesec' issue

Would it be possible for someone to compile the updated Inferno emu.exe for Nt (with 'usesec' fix) and post
it to the inferno-bin project at Google Code?

I think it would be nice if the inferno-bin project had weekly/monthly builds of the Inferno emulators that
are actively developed. (Maybe I should ask the question: How often is the code for the Inferno emulators updated?)

----------------------------------------
> From: forsyth@...
> Date: Sun, 7 Sep 2008 15:50:54 +0100
> To: inferno-list@...
> Subject: Re: [inferno-list] EMU 'usesec' issue
> 
> i've updated inferno-os.  i suspect i closed the item before because i misunderstood
> caerwyn's comment at the time about it having been "fixed in -r181" because i failed
> to follow that that link.

_________________________________________________________________
Stay up to date on your PC, the Web, and your mobile phone with Windows Live.
http://clk.atdmt.com/MRT/go/msnnkwxp1020093185mrt/direct/01/

EMU 'usesec' issue

I would like to try and use Inferno to develop some helpful applications for work.

Unfortunately, my computer at work has the 'usesec' issue. (http://code.google.com/p/acme-sac/issues/detail?id=19)

Questions:

   -- What is the purpose of 'usesec'?

   -- Is 'fixing' the problem just as easy as re-compiling the kernel with 'usesec' disabled? (What issues
might arise with 'usesec' disabled?)
_________________________________________________________________
Stay up to date on your PC, the Web, and your mobile phone with Windows Live.
http://clk.atdmt.com/MRT/go/msnnkwxp1020093185mrt/direct/01/

Memory Requirements

What are the limitations that require 1M of memory in the inferno
system?  Is this a hard limit somewhere or is it an estimation?

Ian

Re: unsubscribe

Was that done because of the Microsoft patent?

On Tue, Sep 2, 2008 at 8:02 AM, David Aiello, sympatico
<daiello@...> wrote:
>
> Pls & thx
>
> David
>
>
> If you can look into the seeds of time, and say which grain will grow and
> which will not, speak to me.  - Shakespeare
>
>
>
>
>
> On 2-Sep-08, at 4:50 AM, Mechiel Lukkien <mechiel@...> wrote:
>
>> hi all,
>>
>> last week, i was wondering why page up/down was not working (on the
>> openbsd port).  it seems the code to handle page up/down events is
>> commented out, in emu/port/win-x11a.c.  the patch below enables it, and
>> works for me.  i don't know why that code was commented out.  there was
>> nothing in inferno-os svn about it.  i don't know the difference between
>> XK_Page_Down & XK_KP_Page_Down, i'm assuming handling both is correct
>> (the other switch-cases are similar).  a similar change might be useful
>> to enable Home & Insert key processing?
>>
>> does anyone see problems with this? or know why the keys were disabled?
>>
>> best regards,
>> mechiel
>>
>>
>> the patch:
>>
>> Index: win-x11a.c
>> ===================================================================
>> --- win-x11a.c    (revision 299)
>> +++ win-x11a.c    (working copy)
>> @@ -1084,18 +1084,18 @@
>>       case XK_KP_Right:
>>           k = Right;
>>           break;
>> -//        case XK_Page_Down:
>> -//        case XK_KP_Page_Down:
>> -//            k = Kpgdown;
>> -//            break;
>> +        case XK_Page_Down:
>> +        case XK_KP_Page_Down:
>> +            k = Pgdown;
>> +            break;
>>       case XK_End:
>>       case XK_KP_End:
>>           k = End;
>>           break;
>> -//        case XK_Page_Up:
>> -//        case XK_KP_Page_Up:
>> -//            k = Kpgup;
>> -//            break;
>> +        case XK_Page_Up:
>> +        case XK_KP_Page_Up:
>> +            k = Pgup;
>> +            break;
>> //        case XK_Insert:
>> //        case XK_KP_Insert:
>> //            k = Kins;
>