Friday, February 6, 2015

Digest for comp.lang.c++@googlegroups.com - 25 updates in 4 topics

Christopher Pisz <nospam@notanaddress.com>: Feb 05 05:45PM -0600

On 2/5/2015 5:14 PM, Jens Thoms Toerring wrote:
> But to achieve that I definitely needed to use types for which
> I knew exactly what sizes they had, not some "could be that or
> alsosomething else" types.
 
I'm making the issue because I had the displeasure of debugging this
very problem for months. People on this newsgroup are not understanding
the problem.
 
The problem is not that you use them or that I don't. It isn't that you
program drivers and I program business apps on Windows. The problem is
what happens when two people work together where one uses it and one
does not, or similarly when you write your interface with it and my code
does not use it, but needs to call your interface or vica versa.
 
There has to be a boundary somewhere. I already asked Ian, but I'll ask
again. What are you going to do at that boundary? What is someone else
going to do?
 
> other month make it known that there's another supported archi-
> tecture. If that "bunch of hobbyists" can do it can it really
> be that hard?
 
It isn't hard for the developer, it's hard for his dependencies.
I mean really, all we have to do is change 32 bit pointers to 64 bit
pointers, check places where we are doing bitwise operations, and carry
on. That's been my experience.
 
It hasn't as much been a problem for my development as it has been
running software in general.
 
I mean, for example, Visual Studio itself is 32 bit, launches 32 bit
tools to debug your 64 bit projects, then it gives ridiculous error
messages about "bad image format" leading a person to think something is
wrong with his code, when its a silly msvc tool linking to your 64 bit
dll....
 
Or the fact that I can't find 64 bit versions of my favorite VSTs when I
want to write some Drum and Bass. :)
 
 
> stand how one could do anything that's architecture-independent
> and thus probably assume that it's rocket science. But it really
> isn't, it's mostly revolves around a few things to keep in mind.
 
You're right. I've been in my Windows bubble. The employers and their
customers are the factor there. It certainly is hard for me to
understand anything *nix. I can't even wrap my head around why people
want to use text editors on the command line and have to memorize 125
different keyboard shortcuts to get the same thing done I can with a
mouse click. But I can accept that is their strange world and it must be
there for a good reason.
 
I did play with Red Hat for 2 years and got a small taste, but it
certainly wasn't fun.
 
> littered with unnecessary types and casts and whatever is a
> PITA, but with your "avoid everything but the most basic
> types" approach you're throwing out the baby with the bath.
 
Then we can agree. Don't use typedefed types without reason and I won't
delete them until I get your reason from you.
 
But the question about what to do at the boundry still remains. I am
sincere in seeking the answer to this. None of the posts in this threads
are addressing that issue.
 
Do you want to stictly rely on compiler warnings about truncation?
What about the reverse case, where programmer X assigned your smaller
type to his larger, but someone wanted to check bit #x for some special
meaning.
 
You can blame developer A or B, but the code still won't work as expected.
 
> those ugly Windows API types and instead get people to use just the
> fewer relevant and easier to grok ones from <cstdint>;-)
 
> Best regards, Jens
 
The windows types poop is largely why I have issue with it. This idea of
using these typedef types seems like it is the same idea they had when
they screwed everything up. It has made life hell for years and pretty
much everyone thinks it was a horrible thing to do. I try, like I
believe any good programmer should, to tuck anything with Windows
specific types away and hide that from the main source.
 
I can however rely on the WORD or BYTE or whatever Windows type poo poo
the Windows API wants me to use, being a certain size, that MSDN will
tell me, and safely assign it to my corresponding C++ primitive. But
sure, I am working on Windows only, and expect my software to run on x86
only, for now.
Robert Wessel <robertwessel2@yahoo.com>: Feb 05 07:32PM -0600

On Thu, 5 Feb 2015 23:59:26 +0100, Martijn Lievaart
 
>Well, a BYTE is 8 bits on the Windows implementations I'm familiar with.
>An unsigned char can be anything, as long as it is 8 bits or more.
 
>I personally prefer uint8_t if I need 8 bits.
 
 
I will be astonished if unsigned char is ever not an 8 bit, two's
complement type on Windows.
Ian Collins <ian-news@hotmail.com>: Feb 06 02:32PM +1300

Christopher Pisz wrote:
 
> There has to be a boundary somewhere. I already asked Ian, but I'll ask
> again. What are you going to do at that boundary? What is someone else
> going to do?
 
I already answered - the platform library interfaces use fixed width and
other typedef types. System data structures use fixed width types. The
standard library uses typedef types.
--
Ian Collins
Robert Wessel <robertwessel2@yahoo.com>: Feb 05 07:43PM -0600

On 5 Feb 2015 23:14:44 GMT, jt@toerring.de (Jens Thoms Toerring)
wrote:
 
>recent ARM port (which also seems to be a dead end) I am aware
>of was for the Alpha processor around 1998 or so, and that ended
>rather abruptly.
 
 
NT shipped on day one on MIPS and x86. Alpha and PPC were added,
although PPC was pretty short-lived, and MIPS died soon thereafter.
Windows was also ported to IPF fairly quickly, and that died only
recently. Something called "Windows" has been on ARM for quite a
while in the Windows CE branch of things, the "real" Windows on ARM is
more recent.
 
 
>So, everything done on Windows for a long time
>was restricted to a single architecture.
 
 
Most Windows software developers (and users) have ignored everything
but x86-32, x86-64 and ARM. So in practice, yes, Windows was a
roughly one platform OS for a long time.
 
And certainly anything in the Win16 or Win9x/Me branches of the world
was x86 only. Of the Win16 versions, only Win3.11 actually required a
386, Win3.1 would run on a 286 (and 386+), and Win3.0 in real mode
(and 286 and 386+ modes). Earlier versions were approximately real
mode, although some came with a DOS extender that required a 386.
Martijn Lievaart <m@rtij.nl.invlalid>: Feb 06 08:35AM +0100

On Thu, 05 Feb 2015 19:32:09 -0600, Robert Wessel wrote:
 
 
>>I personally prefer uint8_t if I need 8 bits.
 
> I will be astonished if unsigned char is ever not an 8 bit, two's
> complement type on Windows.
 
Why depend on that when it's completely unnecessary to do so?
 
M4
 
-- Haven't you heard about Win4DSP yet :-) --
David Brown <david.brown@hesbynett.no>: Feb 06 08:56AM +0100

On 06/02/15 02:32, Robert Wessel wrote:
 
>> I personally prefer uint8_t if I need 8 bits.
 
> I will be astonished if unsigned char is ever not an 8 bit, two's
> complement type on Windows.
 
You would not be alone in your astonishment. However, code - and to a
greater extent, habits - move around. It is better to use clear and
accurate coding rather than sloppy style, even if you know the sloppy
style works for the program you are working on at the time. Maybe your
code will be re-used later, or you will work on a different system at a
later time. "unsigned char" is usually the same as uint8_t, but the
principle applies more widely to other assumptions that are more often
wrong (such as assuming the signedness of plain char, or the size of
long ints).
 
Note that if you are using a Windows API that is defined using type BYTE
for a specifically 8-bit unsigned integer, then you have a choice of
good practices - you can use BYTE to be consistent with the OS and API,
or uint8_t to be consistent with the language standards and to be
explicit about your type requirements.
David Brown <david.brown@hesbynett.no>: Feb 06 09:14AM +0100

On 05/02/15 20:29, Christopher Pisz wrote:
 
> Well you aren't most programmers. You are Ian Collins, guru #2 on the
> newsgroup, and one could have some faith that you know what you are
> doing.
 
Then you would do well to listen to him and his advice, rather than
arguing that because the average windows programmers codes badly, you
should do so yourself.
 
> That isn't the case with the OP and code like what the OP wrote
> is what I get to fix all day. Where you look at it and say to yourself,
> "What in the hell?!"
 
The OP is a newbie who was confused and asking for help (I mean no
disrespect to the OP - we all start at the beginning). No one - except
teachers and helpful newsgroup posters - should be fixing code written
like that, because code written like that never makes it far. You
picked on a single, irrelevant aspect of his code - the use of "uint8_t"
- which you totally misunderstood, blamed on "C" programmers, and
declared it to be a great evil and the root of all coding problems.
 
The OP's code snippet had many errors and misunderstandings - the use of
uint8_t was not one of them.
 
> aren't fixed? Or when perhaps third party code calls your code?
> The whole point is now negated, no? There is a boundry somewhere, what
> do you do at that boundry?
 
Boundaries should /always/ use fixed size and clearly defined types. If
they don't, they are not boundaries.
 
If a customer pays me to interface with code that doesn't have a clear
interface, then they must first pay me to create that clear interface.
 
 
> You'd have to ensure the entire execution path, every developer,
> everywhere is using these same types, 100% contained, no?
 
In the embedded programming world, yes, that is the case. But then,
embedded programmers write code that has to work, and keep working - we
don't get to pop up error boxes or install updates and service packs.
David Brown <david.brown@hesbynett.no>: Feb 06 09:25AM +0100

On 05/02/15 20:41, Christopher Pisz wrote:
> a new target architecture and far far more to consider then the size of
> primitive types. We're still sorting out the move from x86 to x64 how
> many years later?
 
No, typical windows programmers usually write apps for the particular
version of windows plus service packs plus fonts, dlls, etc., that they
happen to have on their machines. Then they test on a few related
systems, hope that it works for a wider audience, pray that it works on
the next windows release, are overjoyed if it works on the next
architecture, and are stunned if it can be recompiled natively on other
windows architectures.
 
In the *nix world, code regularly gets recompiled on different
architectures with different cpus, different primitive type sizes,
different endiannesses, different OS's - all with little extra effort
because the programmer knows the difference between pointless
assumptions such as primitive types, and useful assumptions such as
posix compliance.
Christian Gollwitzer <auriocus@gmx.de>: Feb 06 09:40AM +0100

Am 05.02.15 um 20:29 schrieb Christopher Pisz:
> aren't fixed? Or when perhaps third party code calls your code?
> The whole point is now negated, no? There is a boundry somewhere, what
> do you do at that boundry?
 
There are several different types of boundaries that you may hit. Nobody
here has given the advice to store a HWND in an uint64_t, even if that
would work on all current Windows platforms. A HWND is an opaque type to
the programmer and could just as well be a void* or a struct. But since
there is no point in adding to HWNDs together, storing them in a file or
outputting them to the screen (debuggers aside), noone would ever feel
the need to convert them into an integer.
 
OTOH, there are boundaries across different computers and media. For
instance, reading binary file formats (image files, MATLAB as explained
above, HDF, SQLite), communicating over the network, calling CORBA,
loading a plugin to your program, all that requires you to specify the
interface in an unambigous way and often to the level of bits. Well yes,
there are libraries to do most of this stuff, but still they could
return a type to you at runtime. For instance, in TIFF you can store
1,2,4,8 bit signed and unsigned integers or 32bit and 64 bit floats (aka
double), maybe even other types. You can ask the libtiff library to
convert this to 8bit RGB for you, but if you want to do something else
then just previewing the image, you need to cope with that variety. And
having int16_t & friends comes in very handy here.
 
The real problem IMHO are type casts. If you don't have a very good
reason for a cast, you're doing something wrong.
 
Christian
David Brown <david.brown@hesbynett.no>: Feb 06 09:43AM +0100

On 06/02/15 02:43, Robert Wessel wrote:
>> rather abruptly.
 
> NT shipped on day one on MIPS and x86. Alpha and PPC were added,
> although PPC was pretty short-lived, and MIPS died soon thereafter.
 
I believe that on MIPS, Alpha and PPC, the windows ports were all 32-bit
- despite all three cpus having 64-bit versions as standard at the time.
The ports also used the cpus in little-endian mode, although the PPC at
least is more efficient in big-endian mode.
 
And while the Windows OS itself worked on these systems, applications
and drivers were almost non-existent, because windows programmers would
not or could not think about portability. Most were struggling with the
move to 32-bit x86 at the time, nearly a decade after 32-bit x86 had
become standard on PC's.
 
> Windows was also ported to IPF fairly quickly, and that died only
> recently.
 
The problems with Windows on the "Itanic" are also partly due to the
outstandingly bad processor - MS and Windows programmers are not to
blame for the chip being orders of magnitude hotter, costlier and slower
than other cpus.
 
> Something called "Windows" has been on ARM for quite a
> while in the Windows CE branch of things, the "real" Windows on ARM is
> more recent.
 
And Windows on ARM ("Windows RT") has just been killed off too. Though
if ARM servers take off, maybe we will see a Windows server version for ARM.
 
WinCE supported a number of other architectures too, not just ARM - but
it has never been successful or popular, and been a continuous financial
drain for MS.
 
Then there is Windows Phone, which runs on ARM - and has ambitions to
hopefully become a meaningful minor player in the mobile market.
 
Average windows developers do not make code for any non-x86 systems,
however.
 
David Brown <david.brown@hesbynett.no>: Feb 06 10:01AM +0100

On 06/02/15 00:45, Christopher Pisz wrote:
 
> I'm making the issue because I had the displeasure of debugging this
> very problem for months. People on this newsgroup are not understanding
> the problem.
 
I expect that most people in this newsgroup don't have to deal with
crappy Windows programmers writing crappy code on a daily basis - and
you genuinely have my sympathies for having to work with this sort of
thing. I have seldom had to deal with Windows C++ code, but I have
sometimes had to work with other people's badly-written code (including
the problems caused by their use of primitive types and assumptions
about their sizes), and know how unpleasant it can be.
 
> I mean really, all we have to do is change 32 bit pointers to 64 bit
> pointers, check places where we are doing bitwise operations, and carry
> on. That's been my experience.
 
If you had been coding properly, using types like uint32_t, uintptr_t,
applying sizeof() to types, etc., then there is /nothing/ to change.
That is far better than going through the code and hoping you spot any
undocumented unwarranted assumptions about sizes.
 
> messages about "bad image format" leading a person to think something is
> wrong with his code, when its a silly msvc tool linking to your 64 bit
> dll....
 
Again, you have my sympathies for having to work with such crap tools.
In the *nix world, it is common for tools to work the same whether they
are compiled as 32-bit or 64-bit versions, and for development tools to
happily work with code for either architecture size (or
cross-development for completely different targets).
 
> different keyboard shortcuts to get the same thing done I can with a
> mouse click. But I can accept that is their strange world and it must be
> there for a good reason.
 
Um, you do know that *nix users have been using gui's for three decades
at least? Experienced *nix users often do things from the command line
- when it is easier and faster to do so. (Most Windows users don't,
because the command line "shell" in windows is so limited and there are
no command-line utilities to speak of. More serious Windows users make
use of msys or cygwin ports of *nix command line utilities.)
 
I thought it was really funny that Windows 10 is making a big thing
about their fantastic new "invention" of multiple virtual workspaces -
something *nix users have been using heavily for 25 years.
 
But this is perhaps straying from the topic of the thread...
 
> type to his larger, but someone wanted to check bit #x for some special
> meaning.
 
> You can blame developer A or B, but the code still won't work as expected.
 
Blame the developer who failed to use proper sized types - and make sure
the price tag on your bill reflects that.
 
> much everyone thinks it was a horrible thing to do. I try, like I
> believe any good programmer should, to tuck anything with Windows
> specific types away and hide that from the main source.
 
So because MS made a dogs breakfast out of their types, you think that
/all/ types are bad?
 
Robert Wessel <robertwessel2@yahoo.com>: Feb 06 03:28AM -0600

On Fri, 06 Feb 2015 09:43:37 +0100, David Brown
>- despite all three cpus having 64-bit versions as standard at the time.
> The ports also used the cpus in little-endian mode, although the PPC at
>least is more efficient in big-endian mode.
 
 
All of the released MIPS, PPC and Alpha versions were 32 bit. The
R4000, with the MIPS III 64-bit ISA, didn't arrive until 1992, which
would likely have far too late to be an issue for the initial WinNT
release in 1993.
 
Almost all of the 64-bit port was actually done on Alpha, although the
platform died* before that was ever released, and IPF became the first
64-bit release of Windows.
 
 
*For Windows - Alpha lived on for quite a while with other OSs after
DEC and MS parted ways.
 
 
>not or could not think about portability. Most were struggling with the
>move to 32-bit x86 at the time, nearly a decade after 32-bit x86 had
>become standard on PC's.
 
 
Yep. And the problem was circular. The platforms were not that
compelling, a bit maybe in raw performance at first, but not really in
price/performance, and they lacked software, so very few customers
existed. So nobody bothered to port, even though it wasn't that bad
(we made it half way through a couple of ports, it was really not too
bad). Then x86 pretty much closed the performance gap, while
maintaining a solid cost advantage, and there simply was no point.
 
In some ways MS doesn't get enough credit for its effort to be on
multiple platforms, it's not really MS's fault if all the platforms
mostly sucked compared to x86 for one reason or another. And MS did
make an effort. For the most part the MIPS, PPC, Alpha and IPF
releases were full featured, although the later IPF releases did pare
that back a fair bit. They did fall short of drivers and some
applications (I don't know if a full version of Office was ever
available as a non-x86 build, but partial ones were, and the x86
versions ran on IPF).
 
 
>outstandingly bad processor - MS and Windows programmers are not to
>blame for the chip being orders of magnitude hotter, costlier and slower
>than other cpus.
 
 
Well, IPF (at least ignoring the execrable Merced), was never great,
but it never was "orders of magnitude" hotter or slower, and usually
not nearly that much more expensive either, at least for machines of
comparable size. IOW, an 8S IPF box was not nearly that much more
expensive than an 8S x86 box, but you basically had no workstations or
small servers to choose from.
 
There was a while when SQL Server on IPF was a nice little niche
market for MS. Between all the extra memory (remember that for years
IPF was the only 64-bit version of Windows), and the general
availability* of fairly large, high RAS, boxes (even if the individual
CPUs were not really competitive with x86s), made it a pretty decent
high end platform for SQL Server. In fact SQL Server was more-or-less
the last supported application for Windows on IPF.
 
 
*Even if mostly from a single vendor.
Robert Wessel <robertwessel2@yahoo.com>: Feb 06 03:30AM -0600

On Fri, 06 Feb 2015 08:56:44 +0100, David Brown
>principle applies more widely to other assumptions that are more often
>wrong (such as assuming the signedness of plain char, or the size of
>long ints).
 
 
I certainly don't disagree, but I thought the question was in the
context of an application you know will always run on Windows.
David Brown <david.brown@hesbynett.no>: Feb 06 11:33AM +0100

On 06/02/15 10:30, Robert Wessel wrote:
>> long ints).
 
> I certainly don't disagree, but I thought the question was in the
> context of an application you know will always run on Windows.
 
That was the context - Christopher seems convinced that everything he
ever does will only ever be in the context of 32-bit Windows on x86.
However, there are a few points here:
 
1. Even in the context of Windows, things can change with different
architectures. Some type sizes changed with 64-bit windows - and it is
not impossible that Windows on ARM will become a realistic choice one
day (though type sizes are likely to match Windows on x86, there may be
other subtle changes).
 
2. Even as a Windows programmer, you don't know that your code will only
ever run on that platform. Some parts of the code will clearly be
Windows-specific, but other parts could be reused.
 
3. Even as a Windows programmer, you don't know that you will always be
a Windows programmer. You may change to working on different systems,
and it is better to learn good habits early rather than having to
unlearn the bad habits common in the Windows world.
 
4. Even as a Windows programmer, you may have to interact with
well-written code from other places, and deal with properly typed code.
It's good to be familiar with the standards of the language you claim
to use professionally.
scott@slp53.sl.home (Scott Lurndal): Feb 06 03:40PM


>> We're still sorting out the move from x86 to x64 how
>> many years later?
 
>Bingo!
 
That may be true for consumer grade stuff in the Windows space,
but Linux has been happily x86_64 for a decade now.
Christopher Pisz <nospam@notanaddress.com>: Feb 06 10:01AM -0600

On 2/5/2015 7:32 PM, Ian Collins wrote:
 
> I already answered - the platform library interfaces use fixed width and
> other typedef types. System data structures use fixed width types. The
> standard library uses typedef types.
 
 
Ok, if I understand you correctly, then you are using these fixed width
types in code that the operating system calls. You are leaving it to the
operating system to deal with it. On the other side of the operating
system, I am calling libraries that, in turn, call the operating system,
and I am using int rather than int32_t justifiably. Is that correct?
 
If I do use int32_t in a public interface on my side of the operating
system, then haven't I created another boundary between types that
someone somewhere has to deal with if their code uses the regular
primitive types like int? So, is it your opinion that I shouldn't or
that I should?
 
Let's say in fictional theory Windows has some API with a function
 
void DoSomeSecurityThing(int ID);
 
and I need to call it from my function
 
void MySecurityThing(int32_t userid)
{
// Call OS here
}
 
What is it that the posters in this thread are proposing be done?
 
Since we are talking fiction here, let's say the same situation occured
in *nix. Are we saying that it wouldn't and that *nix would give you
APIs using int32_t instead?
 
I've never seen any library give me functions, structures, or constants
using those types. Are we saying it is because I live in a Windows
bubble, or because I am working on the opposite side of the operating
system from people who do?
Christopher Pisz <nospam@notanaddress.com>: Feb 06 10:04AM -0600

On 2/6/2015 9:40 AM, Scott Lurndal wrote:
 
>> Bingo!
 
> That may be true for consumer grade stuff in the Windows space,
> but Linux has been happily x86_64 for a decade now.
 
I dunno how we define "happily has been"
 
I seem to recall less than 5 years ago, I tried to install 64 bit
drivers on a 64bit install of Red Hat Linux and reading all manner of,
"this doesn't work yet" articles, much like the problems on other 64bit
operating systems had to go through.
 
I guess we could say Windows XP 64bit worked too if all you wanted to do
was install it.
scott@slp53.sl.home (Scott Lurndal): Feb 06 04:54PM

>drivers on a 64bit install of Red Hat Linux and reading all manner of,
>"this doesn't work yet" articles, much like the problems on other 64bit
>operating systems had to go through.
 
"I seem to recall"?
 
I've been using, and developing, 64-bit linux distributions since
2004. There have been no issues vis-a-vis I/O drivers in
this time other than the standard issues surrounding proprietary
binary kernel drivers [**](which are architecture independent). Also
did quite a bit of 64-bit Unix OS work on IRIX in the 90's which
along with the Alpha work, set the stage for 64-bit x86 linux
in the 2000's.
 
Between the quality of the programmatic API (which I was involved
in the definition of during the decade of the 90's[*]) and the
binary processor-specific ABI (psABI), and the general use of
the correct types in code designed to execute on multiple
architectures (as most unix/linux code has been); running 64-bit
applications (and legacy 32-bit applications) has been pretty much
trouble free.
 
[*] 88Open, X/Open, Unix International and the IEEE POSIX.
[**] 32-bit PCI device issues notwithstanding (e.g. SWIOTLB),
which were simliar to the old ISA issues vis-a-vis DMA.
scott@slp53.sl.home (Scott Lurndal): Feb 06 04:56PM

>operating system to deal with it. On the other side of the operating
>system, I am calling libraries that, in turn, call the operating system,
>and I am using int rather than int32_t justifiably. Is that correct?
 
It's really simple. Use the defined interface type. If the library
routine header file requires an int, pass an int. If it requires an
unsigned long long, pass an unsigned long long. If it requires an
uint8_t, pass a uint8_t.
 
That said, int is too often used when the domain doesn't include
negative numbers.
Peter Olcott <OCR4Screen>: Feb 06 06:57AM -0600

The terminology and syntax shown below may have errors and/or not
consistently conform to standard conventions.
 
enum three_value_logic{FALSE, TRUE, ERROR}; // C++
{0,1,2} single digit on final tape of Turing Machine.
 
The halting property of (p,i) is true iff (if and only if) p is a
program that
halts on its input and false otherwise.
 
01 phd_3, p and i in Sigma*
02 <ThereExists> phd_3 in Partial_Halt_Deciders such that
03 <ForEach> p in Programs and <ForEach> i in Inputs
04 (phd_3(p,i) == halts(p,i)) <xor>
05 ((halts(p,i) != (phd_2(p,i) == FALSE)) <and>
06 (halts(p,i) != (phd_2(p,i) == TRUE)))
 
phd_3 correctly divides Program x Inputs into these two subsets:
1) On line 04 phd_3(p,i) would return either TRUE or FALSE corresponding
to halts(p,i).
This would correspond to correctly deciding the halting property for a
subset of
Program x Inputs.
 
2) On lines 05 and 06 phd_3(p,i) would return ERROR.
This would correspond to the impossibility of correctly deciding the
halting property
for a subset of Program x Inputs. // These lines have not yet been
reviewed by others
 
Automata and Computability Dexter Kozen (1997) Page 245
(Rice's Theorem) Every nontrivial property of the r.e. sets is undecidable.
 
The intent of the specification above is to divide up Program x Inputs
into the subset
of string pairs (p,i) where phd_3(p,i) can correctly decide the halting
property and
the subset of strings pairs (p,i) where correctly deciding halting is
impossible for phd_3(p,i).
 
The counter-example to Rice's Theorem arises because phd_3 correctly decides
whether or not it can correctly decide the halting property for every
element of
Program x Inputs.
Peter Olcott <OCR4Screen>: Feb 06 08:34AM -0600

On 2/6/2015 8:01 AM, Ben Bacarisse wrote:
>> 05 ((halts(p,i) != (phd_3(p,i) == FALSE)) <and>
>> 06 (halts(p,i) != (phd_3(p,i) == TRUE)))
> phd_2 is unbound. Probably a typo for phd_3.
 
Yes
 
> Partial_Halt_Deciders is
> an undefined set.
 
I used the abbreviated notation that you specified.
Previously I always stated:
Partial_Halt_Deciders = Sigma*
Programs = Sigma*
Inputs = Sigma*
 
> machines m that don't halt on the input (p,i). (A common solution is to
> use an undefined value, often called "bottom" _|_ to denote the
> situation.)
 
phd_3 is specified (in the notes) to be a total Turing Machine because it
always returns exactly one of {FALSE, TRUE, ERROR}.
 
Perhaps there is some way to say this in the formal specification?
 
>> Program x Inputs.
> phd_3 is not a machine or function. It is the name used in a quantified
> formula.
 
I am still not clear on the exact difference between a quantified
formula and the
specification of elements of a set. It seems to me that I have specified
at least
one element of Sigma*.
 
It would seem that this element must also be either the encoding of a
machine
or the encoding of a software function because a random string could not
correctly divide the set of all possible finite string pairs into the
subset that it
correctly decides and the subset where correctly deciding is impossible
for it.
 
Because of this it would seem that phd_3 would specify at least one literal
string that is the encoding of a Turing Machine or a software function.
Peter Olcott <OCR4Screen>: Feb 06 08:41AM -0600

On 2/6/2015 8:04 AM, Ben Bacarisse wrote:
>> phd_3 is not a machine or function. It is the name used in a quantified
>> formula.
> That's wrong. It is not in a forall. You need to prove that it exists.
 
I do not know what you mean about the forall.
 
Partial_Halt_Deciders, Programs, Inputs = Sigma*
<ThereExists> phd_3 in Partial_Halt_Deciders such that
<ForEach> p in Programs and <ForEach> i in Inputs
(phd_3(p,i) == halts(p,i)) <xor>
((halts(p,i) != (phd_3(p,i) == FALSE)) <and>
(halts(p,i) != (phd_3(p,i) == TRUE)))
 
Maybe now at least I have a clear basis for someone else to
provide a counter-example showing that phd_3 does not exist?
Martin Shobe <martin.shobe@yahoo.com>: Feb 06 09:43AM -0600

On 2/6/2015 8:41 AM, Peter Olcott wrote:
> (halts(p,i) != (phd_3(p,i) == TRUE)))
 
> Maybe now at least I have a clear basis for someone else to
> provide a counter-example showing that phd_3 does not exist?
 
Go learn some logic. The burden of proof is on you to show that it does
exist. Good luck with that. Especially as we can construct, for each
phd_3, a machine, c, in the following way.
 
C(p, i)
{
if (phd_3(p,i) == false)
{
return;
}
 
while (1);
}
 
Now, halts(C, i) is true if and only if phd_3(C, i) is false.
 
Therefore, if phd_3(C, i) returns TRUE halts(C, i) is false. Therefore,
(phd_3(p,i) == halts(p,i)) <xor> ((halts(p,i) != (phd_3(p,i) ==
FALSE)) <and> (halts(p,i) != (phd_3(p,i) == TRUE)))
 
(TRUE == FALSE) <xor> ((FALSE != (TRUE == FALSE)) <and> (FALSE !=
(TRUE == TRUE)))
 
FALSE <xor> ((FALSE != FALSE) <and> (FALSE != TRUE))
 
FALSE <xor> (FALSE <and> TRUE)
 
FALSE <xor> FALSE
 
FALSE
 
So, if phd_3 meets the requirements, phd_3 does not return TRUE.
 
If phd_3(C, i) returns FALSE, halts(C, i) is true. Therefore,
(phd_3(p,i) == halts(p,i)) <xor> ((halts(p,i) != (phd_3(p,i) ==
FALSE)) <and> (halts(p,i) != (phd_3(p,i) == TRUE)))
 
(FALSE == TRUE) <xor> ((TRUE != (FALSE == FALSE)) <and> (TRUE !=
(FALSE == TRUE)))
 
FALSE <xor> ((TRUE != TRUE) <and> (TRUE != FALSE))
 
FALSE <xor> (FALSE <and> TRUE)
 
FALSE <xor> FALSE
 
FALSE
 
So, if phd_3 meets the requirements, phd_3 does not return FALSE.
 
If phd_3(C, i) returns anything else (we'll call it X), halts(C, i) is
false. Therefore,
(phd_3(p,i) == halts(p,i)) <xor> ((halts(p,i) != (phd_3(p,i) ==
FALSE)) <and> (halts(p,i) != (phd_3(p,i) == TRUE)))
 
(X == FALSE) <xor> ((FALSE != (X == FALSE)) <and> (FALSE != (X ==
TRUE)))
 
FALSE <xor> ((FALSE != FALSE) <and> (FALSE != FALSE))
 
FALSE <xor> (FALSE <and> FALSE)
 
FALSE <xor> FALSE
 
FALSE
 
So, if phd_3 meets the requirements, phd_3 does not return anything
else.
 
If phd_3 doesn't return anything, it doesn't meet the requirements.
 
Therefore, phd_3 doesn't meet the requirements. Since phd_3 was an
arbitrary machine, There are no machines that meet the requirements.
 
Martin Shobe
bleachbot <bleachbot@httrack.com>: Feb 06 03:04PM +0100

"CHE SIA REVOLUCIOOON!!!" <johnnybureddu@gmx.com>: Feb 06 06:04AM -0800

PEZZO DI MERDA CRIMINALISSIMO DAVIDE SERRA DI ALGEBRIS E TWITTER, HA BECCATO L'INSIDER SULLE BANCHE POPOLARI DAL MEGALAVA CASH MAFIOSO PIER LUIGI BOSCHI DI BANCA ETRURIA. SON TUTTI FASCIOCAMORRISTI, TIPO PAOLO BARRAI DI WMO, CHE SI FINGONO DEL PD X...
 
 
STO CRIMINALISSIMO ESCREMENTO DI DAVIDE SERRA DI ALGEBRIS E TWITTER HA RICEVUTO L'INSIDER TRADING SULLE BANCHE POPOLARI DA NOTO LAVA SOLDI MAFIOSI PIER LUIGI BOSCHI DI AREZZO E BANCA ETRURIA. HA PURE RAGLIATO CHE COMPRA BANCHE POPOLARI DAL MARZO 2014. SPUTTANANDOSI ANCORA DI PIU', COME UN IMBECILLE, PEDERASTA SODOMIZZA BAMBINI E MEGA COCAINOMANE, QUALE DA SEMPRE E'. IL SUO POR-CO-RROTTISSIMO MATTEO RENZI (CHE VIA "SS", SPINTA E STECCHE DI SILVIO BERLUSCONI, HA SCIPPATO SEGRETERIA PD E PALAZZO CHIGI, NEL SECONDO CASO COL FEBBRAIO 2014), GLI HA PASSATO, ATTRAVERSO "A ZOCCOLONA BERLUSCONICCHIA, STECCATISSIMA E CELLULITOSA" MARIA ELENA BOSCHI http://www.dagospia.com/img/foto/08-2014/maria-elena-boschi-bikini-rosa-in-spiaggia-a-marina-di-pietrasanta-581617_tn.jpg E IL VERME MEGA LAVA SOLDI MAFIOSI PIER LUIGI BOSCHI DI BANCA ETRURIA, L'INSIDER SULLE BANCHE POPOLARI. E PER QUESTO, STO ESCREMENTO HITLERIANO DI DAVIDE SERRA DI TWITTER E ALGEBRIS, SI E' MESSO A COMPRARE BANCHE POPOLARI DAL MARZO 2014. UN MESE DOPO (ULLALA CHE COINCIDENZA, ULLALA). SEMPRE INSIDER E'. TRATTASI DI MANDRIA DI PORCI FASCIOCAMORRISTI, TIPO, ANCHE, NOTO AVANZO DI GALERA PAOLO BARRAI (DI CRIMINALISSIME WMO, BSI ITALIA SRL DI VIA SOCRATE 26 MILANO E BLOG "MERDATO"LIBERO), CHE SI FINGONO DEL PD, X... DISTRUGGERLO, INFILTRARLO A MORTE, RENDERLO DIARREA BERLUSCONICCHIA! VOGLIAMO UNA ACCESISSIMA E VINCENTISSIMA REVOLUCIOOOOOON! VOGLIAMO IL CANCROMICIDA DEL MONDO INTERO, SILVIO BERLUSCONI, FALLITO ED IN GALERA! SUBITO! PLS, DOTTOR SERGIO MATTARELLA, CI DIA UNA MANO. IN ONORE A SUO FRATELLO UCCISO DALLA MAFIA ( MAFIA CHE QUANDO SI METTE LA FASCISTISSIMA CRAVATTA DOLCE E GABBANA, SIGNIFICA SILVIO BERLUSCONI E DAVIDE SERRA). IN ONORE AD ETERNI GIOVANNI FALCONE E PAOLO BORSELLINO, FATTI SPAPPOLARE, SICURISSIMAMENTE, DA SILVIO BERLUSCONI, VIA, A SUA VOLTA, BERLUSCONIANISSIMA COSA NOSTRA! ED OLTRE A VOLER SILVIO BERLUSCONI FALLITO ED IN GALERA, VOGLIAMO VEDERE IL SUO PICCIOTTO INCRAVATTATO, IL FACCENDIERE DI BERLUSCONAZISTI, PADANAZISTI, E CRIMINALITA' ORGANIZZATE DI MEZZO MONDO, PAOLO BARRAI DI MALAVITOSA WMO, PURE, IN GALERA! VOGLIAMO IL NUOVO GIANCARLO LANDE, IL NUOVO BERNARD MADOFF, IL NUOVO MICHELE SINDONA, VERME CRIMINALISSIMO DAVIDE SERRA DI TWITTER ED ALGEBRIS, FALLITO, E PER LO MENO, PER QUALCHE MESE, IN GALERA! CHE SIA ETICISSSIMA E VINCENTISSIMA REVOLUCIOOOOON!
 
COME DA OTTIMO SITO INFORMARE X RESISTERE:
http://www.informarexresistere.fr/2015/01/27/qualcuno-sapeva-in-anticipo-che-il-governo-avrebbe-varato-un-provvedimento-sulle-banche-popolari-enormi-speculazioni/

COME DA CORRIERE DELLA SERA, DI, OTTIMAMENTE, ANTIRENZUSCONIANO FERRUCCIO DE BORTOLI, DA NON TOCCARE E A TUTTI I COSTI:
http://www.corriere.it/economia/15_gennaio_24/quei-movimenti-un-po-sospetti-popolari-f59ffb1c-a3a5-11e4-808e-442fa7f91611.shtml

Acquisti consistenti prima della riforma che ha abolito il voto capitario. La famiglia Boschi ha sicurissimamente passato insider trading a Londra, tramite noto ladro, truffatore, nazifascista, immensamente ricicla soldi mafiosi, che affatto va' in Tanzania a fare del bene, in quanto vi va' a riciclare cash di (sua) LL Lega Ladrona, come per suoi gusti sessuali di tipo depravatissimo: avanzo di galera Davide Serra di Algebris e Twitter. Dove prenderanno, le mazzette, ora, i vermi nazifascisti Pier Luigi Boschi di Banca Etruria e sua zoccolona ( di fatto) Berlusconicchia Maria Elena Boschi ( bastarda puttanazista che vuole sgozzare la giustizia via estremissimamente ingiusta salvaberlusconi http://www.blitzquotidiano.it/rassegna-stampa/libero-renzi-per-fare-la-pace-offre-la-salva-berlusconi-che-fara-mattarella-2090491/ .... che qui, non per niente, slingua un topo di fogna corrotto, ndranghetista, fascista, estortore di soldi alla Banca Popolare di Lodihttp://www.repubblica.it/2005/l/sezioni/economia/banche21/ipolitici/ipolitici.html .. pezzo di merda criminalissimo Paolo Romani
http://www.corriere.it/methode_image/2014/08/08/Politica/Foto%20Politica%20-%20Trattate/6ebdfe07bdd8cb1fe88af8343f8a5b1c-012-kXsC-U43030145012273wcB-593x443@Corriere-Web-Sezioni.jpg?v=20140808175213 )?
A) Alle Bahamas
B) Alle Bermuda
C) A Panama
D) Ad Hong Kong
E) A Singapore
F) Alle Mauritius ( "roba" tipo Svizzera e' da anni 70, 80: stile nazimafioso pedofilo Silvio Berlusconi e suo B-o-ttino Craxi.. dai.. please)
Lauti premi a chi azzecca per primo.
---
BRAVO, BRAVO, DAVVERO BRAVISSIMO ELIO LANNUTTI A QUERELARE STO VERME BERLUS-CORROTTTISSIMO DI MATTEO RENZI:
http://www.ilfattoquotidiano.it/2015/01/21/denuncia-per-renzi/1357263/
CHE SBEFFEGGIA PM PER BENE, EROICI, SALVA NAZIONE ( SPESSO FATTI ESPLODERE, COME IL NAZIMAFIOSO PEDOFILO STRAGISTA SILVIO BERLUSCONI FECE FARE CON GLI ETERNI GIOVANNI FALCONE E PAOLO BORSELLINO). TIPO QUELLI DI PALERMO, BARI, MILANO, NAPOLI, DICENDO, ANZI, RAGLIANDO LORO: "'OOOO OO CHE PAURA, MI FANNO, OO OO"
http://tv.ilfattoquotidiano.it/2014/09/10/renzi-anm-protesta-brrrrr-che-paura-sciopero-sindacati-polizia-illegale/295911/
CHE RABBIA MOSTRUOSISSIMA, QUESTO VERMINOSO, CRIMINALISSIMO TRAFFICARE FRA POR-CO-RRUTTORE MAXIMO SILVIO BERLUSCONI E POR-CO-RROTTO MAXIMO MATTEO RENZI!
https://ilgrandetsunami.wordpress.com/2015/01/17/berlusconi-che-ne-sara-di-me-il-2-febbraio-carmelo-lopapa/
"IO TI VOTO LE RIFORME (ODIOSISSIMAMENTE MAFIOSE E FASCISTE, OSSIA BERLUSCONIANISSIME) CHE STAI APPRONTANDO ( VEDI SENATORI NON ELETTI E CAPOLISTA BLOCCATI, COSA CHE ANCHE I VERMINOSI MATTEO RENZI E SILVIO BERLUSCONI DEGLI ULTIMI 8 DECENNI, OSSIA ADOLF HITLER, BENITO MUSSOLINI, ALFREDO STROESSNER, FRANCISCO FRANCO, EMILIO EDUARDO MASSERA, AUGUSTO PINOCHET E POL POT AVREBBERO SENTITO TANTISSIMO PUDORE AL SOL PROVARE A PENSARNE), TU METTI AL QUIRINALE UN FANTOCCIO DI MIA PROPRIETA' CHE COMPRO QUANDO VOGLIO QUALE GIULIANO AMATO, VALTER VELTRONI O ANNA FINOCCHIARO ... O MEGLIO ANCORA, SE PARLIAMO DI MIEI FASCIOBAMBOCCI ALLA PIERFERDINANDO CASINI O GIANNI LETTA... TUTTI MIEI PUPAZZI CHE MI HAN GIA' GARANTITO CHE CON SEI EURO E MEZZO CASH, MI FIRMEREBBERO TUTTE LE GRAZIE CHE VOGLIO IN NOME DELLA MIA.... PACIFICAZIONE ALLA VASELLINA... E ... SPECIALMENTE ...GIUSTO PER ANDARE SUL SICURO.... MI FAI ANCHE E SUBITO UNA NORMINA DECAPITANTE NOIOSISSIMI CONCETTI COME DEMOCRAZIA E GIUSTIZIA CHE IMPONGA IL MIO TORNARE IN POLITICA, COSI' CHE POSSA FOTTERE IL POPOLO CIUCCIO, LE LEGGI, DOZZINE DI (GRANDISSIMI) MAGISTRATI COME ILDA BOCASSINI, EDMONDO BRUTI LIBERATI, NINO DI MATTEO, ROBERTO SCARPINATO, FABIO DE PASQUALE, HENRY WOODCOCK, PASQUALE DRAGO, ATTRAVERSO LA ( BASTARDAMENTE VIGLIACCHISSIMA) IMMUNITA' EVITA GALERA, CHE MI RI RITROVEREI"!
http://www.ilfattoquotidiano.it/2015/01/18/salva-berlusconi-alessandro-pace-manina-renzi-reato-falso/1349562/
http://www.ilfattoquotidiano.it/2015/01/08/salva-berlusconi-mucchetti-renzi-venga-senato-spiegare-successo/1322595/
http://www.ilfattoquotidiano.it/2015/01/06/salva-berlusconi-coppi-ammette-quella-norma-segnale-per-quirinale/1318110/
ECCO DOVE CI PORTANO BASTARDI LAVA CASH MAFIOSO A GO GO COME I MALAVITOSINCRAVATTATI DAVIDE SERRA DI ALGEBRIS E TWITTER INSIEME AL RENATO VALLANZASCA UNITO AD UGO FANTOZZI DELLA FINANZA, NOTO AVANZO DI GALERA PAOLO BARRAI NATO A MILANO IL 28.6.1965, DI CRIMINALISSIMO WMO, CRIMINALISSIMA BSI ITALIA SRL DI VIA SOCRATE 26 MILANO E CRIMINALISSIMO BLOG MERCATO "MERDATO" LIBERO ( DUE VERMI REPELLENTI CHE RICICLANO ALL'ESTERO VAGONI DI SOLDI DI COSA NOSTRA, CAMORRA, NDRANGHETA O LADRATI SE NON PURE FRUTTO DI MEGA MAZZETTE IN DIREZIONE LL LEGA LADRONA ED EX PDL POPOLO DI LADRONI; IN CONGIUNZIONE CON BANCHIERI DELINQUENTISSIMI, SPESSO PURE MANDANTI DI OMICIDI O "SUICIDATE", COME FATTO CON DAVID ROSSI DI MONTE PASCHI, QUALI GLI ASSASSINI ENNIO DORIS E MASSIMO DORIS DI BANCA MEDIOLANUM; O QUALE "O MASSONE CAMORRISTA" GIUSEPPE SABATO DI BANCA ESPERIA
http://www.gruppoesperia.it/chi-siamo/giuseppe-sabato.html
https://books.google.it/books?id=B1mEj0GtktIC&pg=PT304&lpg=PT304&dq=GIUSEPPE+SABATO+LICIO+GELLI&source=bl&ots=Gqtu0KYRmD&sig=d2TOz9sZDY6563zIPxwnNYcbxb4&hl=it&sa=X&ei=I-i_VOOsBMLlUonCgZgI&ved=0CFMQ6AEwCA#v=onepage&q=GIUSEPPE%20SABATO%20LICIO%20GELLI&f=false
TUTTI DEL GRUPPO MA-F-INIVEST DI " STEFANO BONTATE, MARCELLO DELL'UTRI, TOTO RIINA, LICIO GELLI, BERNARDO PROVENZANO E SILVIO BERLUSCONI: " OO CHE CASO, OO")! E PROPRIO MENTRE VIENO ACCLARATO CHE STO VERME COLERICO E STECCATISSIMO DI MATTEO RENZI, COME INTUITO DA GENIO BORSISTICO ED EROE CIVILE MICHELE NISTA DA ANNI E NON "SOLO" 11 MESI, E' IN POLITICA, IN PRIMIS, PER PROTEGGERE IL TOPO DI FOGNA DI SUO PADRE, TIZIANO RENZI. ACCERTATO BANCAROTTIERE FRAUDOLENTISSIMO, ACCERTATO NEOPIDUISTA LADRONE E TRUFFATORE! CHE HA SODOMIZZATO UN MILIONE DI EURO A FIDI TOSCANA E LI HA FATTI PAGARE AL POPOLO CIUCCIO, VIA SUO BASTARDO NAZIMAFIOSO POR-CO-RROTTO DITTATORE MATTEO RENZI!
http://www.beppegrillo.it/2015/01/i_conflitti_dinteressi_della_famiglia_renzie.html
https://www.youtube.com/watch?v=A7Ngp6JrK9A
http://robertoiacobone.altervista.org/debiti-azienda-di-famiglia-renzi-pagati-dal-governo-renzi/?doing_wp_cron=1421500410.9769570827484130859375
VOGLIAMO A PALAZZO CHIGI STEFANO FASSINA SUBITO! INSIEME AL PD PER BENE, QUELLO ANTI MAFIA FASCISTA DI MATTEO RENZI! INSIEME, OVVIAMENTE, A M5S E SEL! A FARE IL SUO VERO LAVORO, OSSIA LA ZOCCOLA DI STRADA, STA BATTONA HITLERIANA, CHE SI CREDE MODELLA MA E' CESSO STRA COLMO DI CELLULITE, DI MARIA ELENA BOSCHI http://www.dagospia.com/img/foto/08-2014/maria-elena-boschi-bikini-rosa-in-spiaggia-a-marina-di-pietrasanta-581617_tn.jpg FIGLIA DI ALTRO VERME CRIMINALISSIMO: MEGA LAVA SOLDI MAFIOSI PIER LUIGI BOSCHI DI BANCA ETRURIA (DOPO AVER PASSATO TUTTA UNA VITA A TRAFFICARE CON COOP VICINISSIME A MAFIA, CAMORRA E NDRANGHETA, NON PER NIENTE, STILE "RENZUSCONIANISSIMI" SALVATORE BUZZI E MASSIMO CARMINATI)! AL QUIRINALE UN UOMO O DONNA VERA, ALLA NINO DI MATTEO O ILDA BOCASSINI, CHE FACCIA TRASLOCARE IL CANCROMICIDA DEL MONDO INTERO, SILVIO BERLUSCONI, DA PALAZZO GRAZIOLI A PALAZZO UCCIARDONE E SUBITO. O VERA RIVOLUZIONE SARA'! RIVOLUZIONE RIPRISTINANTE VERA DEMOCRACIA Y LIBERTAD! PS SEMPRE VINCENTISSIMI I GENI BORSISTICI GEORGE SOROS E MICHELE NISTA A PUNTARE SULLA SPAGNA. PARREBBE CHE DOPO AVER SAPUTO CHE IL MIGLIORE FIUTO PER QUALSIASI COSA AL MONDO, MICHELE NISTA, VEDA NON MALE LA SPAGNA, GEORGE SOROS ABBIA DECISO DI METTERCI SUBITO, MANCO FOSSERO NOCCIOLINE, 500 MILIONI DI EURO, NELL'AUMENTO DI CAPITALE DI BANCO SANTANDER. NON SARA' UN PAESE IMMUNE DI DIFETTI, LA SPAGNA, COME NON LO E' ALCUN PAESE DEL PIANETA TERRA. ANCHE LI, GLI SCANDALI PER CORRUZIONE NON MANCANO ( MA SONO AL MASSIMO UN DECIMO, RISPETTO A QUELLI DELLA CLOACA DI RENZUSCONIA)! PERO', NONOSTANTE MEZZO SECOLO DI NAZIFASCISMO, OGGI LI VI E' DEMOCRAZIA VERA. SIA AL POTERE MARIANO RAJOY DELL'OPUS DEI O IL PROMETTENTISSIMO PABLO IGLESIAS DI PODEMOS. NON VI SONO, SULLA GOVERNATIVA POLTRONA DI MADRID, VERMI STRAGISTI, FASCIOCAMORRISTI E PEDOFILI ALLA SILVIO BERLUSCONI, CHE SI FAN LE LEGGI PER GONFIARSI LE TASCHE DI SOLDI LERCISSIMI OLTRE CHE PER SGOZZARE A MORTE DEMOCRAZIA E GIUSTIZIA, OGNI GIORNO. E QUESTO, O IN PROPRIO, O COMPRANDOSI RAGAZZINI CORROTTISSIMI CHIAMANTISI MATTEO, COME MATTEO RENZI (OGGI). O IL NUOVO ADOLF HITLER: MATTEO SALVINI (DOMANI). MENTRE L'UNICO PADRONE, L'UNICO VERO BOSS DEL CANCROMICIDA DEL MONDO INTERO, SILVIO BERLUSCONI, UN ALTRO MATTEO, MATTEO MESSINA DENARO, SORRIDE E DICE " BRAVO MIO PRESTANOME BEDDU SILVIO BERLUSCONI, HAI TRASFORMATO L'ITALIA IN RENZUSCONIA, CHE IN REALTA' SEMPRE BERLUSCONIA E', AAAAAAA... COME PIACE A MMMIA, AAAA.... E' TUTTO UNA COSA NOSTRA, SILVIUZZEDDU BEDDU .. CONTINUA COSI' CHE TI TROVIAMO QUALCHE ALTRA BEDDA PROSTITUTA DI 12-14 ANNI PELLU TEMPU LIBERO, AAAAA.... QUESTA VOLTA CAMBIAMO, AAAA... TE LA TROVIAMO FILIPPINA E LA FACCIAMO PASSARE PER LA NIPOTE DEL RE DELLA THAILANDIA, BHUMIBOL ADULYADEJ, IL RE PIU' RICCO DEL MONDO... CHE SPESSO E VOLENTIERI "ABOLISCE UFFICIALISSIMAMENTE LA DEMOCRAZIA"... SI ... SILVIUZZEDDU BEDDU DA COSA NOSTRA, TI TROVIAMO UNA BAMBINA FILIPPINA DI 12 ANNI DA SBAVARE E TOCCARE QUANTO VUOI... E LA FACCIAMO PASSARE PER LA NIPOTE THAILANDESE DI BHUMIBOL ADULYADEJ, AAAA.... COSI' VEDRAI CHE QUANDO TELEFONI, PREOCCUPATISSIMO, DA PARIGI ( TANTO, FRA POCO, NELLA TUA DITTATURA DELLE BANANAS DI RENZUSCONIA, SUBITO, IL PASSAPORTO, TI RIDARANNO), I POLIZIOTTI O QUESTORI, SOLO E SEMPRE LA TUA VOLONTA', FARANNO, AAA"!!!
You received this digest because you're subscribed to updates for this group. You can change your settings on the group membership page.
To unsubscribe from this group and stop receiving emails from it send an email to comp.lang.c+++unsubscribe@googlegroups.com.
Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)