Web Prolog, Elixir and COBOL

[Transfinite Numbers]

Saving a post that will be deleted from SWI-Prolog
forum, because of some violation who knows what.
So preserving its content here:

Instead of Web Prolog, you could also call it Magic Potion
Programming Language. (BTW: I am not sure whether Japan or
so understands the ethymology of Elixir)

Elixir
An elixir is a clear, sweet-flavored liquid used for medical
purposes, to be taken orally and intended to cure one's illness.
When used as a pharmaceutical preparation, an elixir contains at
least one active ingredient designed to be taken orally. The
word was introduced in late Middle English, through Latin
from Arabic al-ʾiksīr (الإكسير), which in turn is the Arabization
of Greek xērion (ξήριον) "powder for drying wounds" (from ξηρός
xēros "dry"). They are used as solvents or vehicles for th...

(P.S.: The use of magic related terms in software product names
is not uncommen. We found for example General Magic, which was
Palo Alto Start-Up. And this Isreal based Case Tool Magic
Software, I had some exposure in the far past, still exists)

(P.P.S.: General Magic is known for promoting Telescript at
those times. This is yet another spin on Actors. It allows
function shipping seen as an agent going from one place to another.
Does not Web Prolog also explore a little bit this aspect?)

BTW: Cobol had already its 50th birthday, and it looks to me
they have a realistic view what an ISO standard means. The below
screenshot is from stack overflow tag description, and we also
find that the standards latest revision is form 2014.
Not only corrigenda:

Standards

The official COBOL standard is very slow to evolve due to a
strong desire to keep the language relevant without compromising
core strengths such as execution efficiency, interoperability
with other languages and backward compatibility with earlier
versions of the language.

The current COBOL ISO/IEC standard is ISO/IEC
1985:2014 Information technology -- Programming languages -- COBOL.

Vendors of course continue to issue new releases of their COBOL
compilers with unique non-standard features from time to time.
As a consequence, COBOL variants vary more than one might expect
given the availability of standards.

[Transfinite Numbers]

It was quite interesting to discover/respectively rediscover
COBOL. And that they have an ISO standard which considerably
evolved. Its seems also that COBOL added objects to their

language and it also runs on JVM and .NET.

[Bruce Axtens]

Don't forget GnuCOBOL which is open-source and runs on Windows, Linux
and Mac.

[Transfinite Numbers]

Its seems to be alive with a couple of
contributors, according to wikipedia.

3.1 release candidate coming!

[j4n bur53]

One dissatisfaction with the SWI Web Prolog proposal,
it is not clear whether it proposes something structured
or something unstructured.

Most of the examples show just "predicate calls",
enriched maybe by "predicate shipping". But it is not
clear whether the code is robust, has some nursery

and whether it is steadfast. But still it was rewarding
to discuss the matter, since it highlighted again that
probably the most important usecase of some "Web Prolog"

might be structured parallelism. And as a surprise I
started again thinking about debugging actors. Unstructured
actors might have their place, but somehow they might

need some features with hindsight of a later structured
use. What do others do? Interestingly I found Dart
mentioning better debugging:

talks about writing and debugging async Dart code
Dart Today & Tomorrow (Dart Developer Summit 2015)
https://www.youtube.com/watch?v=FiXiI2Atexc

But the above is an old video relatively short and
only giving some demo without a lot of background
information. Not sure what materialized.

Interestingly they call their local actors isolates.
In Java an isolate is something between process and
thread there is a special package for it.

Not sure what Dart isolates are and whether they
are the same as Java isolates or not. Well they
drumming about isolates again, through flutter:

Isolates and Event Loops - Flutter in Focus
24.166 Aufrufe - Am 14.06.2019 veröffentlicht
https://www.youtube.com/watch?v=vl_AaCgudcY

Transfinite Numbers schrieb:

[Transfinite Numbers]

Elixir guy does Naruto run near Area 51
https://www.youtube.com/watch?v=33DjpFy4lcU

LoL

On Monday, September 2, 2019 at 1:23:09 PM UTC+2, Transfinite Numbers wrote:

[j4n bur53]

Yet another contender for Prolog in the browser:

ProscriptLS is developed by Lindsey Spratt
(Lindsey dot Spratt at LogicForSystems dot com).
It is a fork of the original Proscript project
by Matt Lilley.

- builtins to work with the DOM (Document Object Model) of HTML,
- an interactive Prolog tracing debugger that can be used
in the client web page (using a JQuery Terminal),
- and support for modules.

The original Proscript project implemented the bulk
of the system, particularly the WAM, the WAM compiler,
and the basic architecture of the system.

https://lindseyspratt.github.io/proscriptls/index.html

Transfinite Numbers schrieb:

[j4n bur53]

Pitty, even the calculator example doesn't work:

/favicon.ico:1 Failed to load resource: the server responded with a
status of 404 ()
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:add_digit'/1,in('bootstrap_js:call_with_module/2:(execute(user:add_digit/1),8)'))
proscriptls.js:3322 Uncaught (in promise) ReferenceError: promise is not
defined
at promise_backtrack (proscriptls.js:3322)
at promise_callback (proscriptls.js:3305)
at proscriptls.js:3292
at async request_result (proscriptls.js:3291)
5
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:add_digit'/1,in('bootstrap_js:call_with_module/2:(execute(user:add_digit/1),8)'))
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:fix_current'/0,in('bootstrap_js:call_with_module/2:(execute(user:fix_current/0),5)'))
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:operate'/1,in('bootstrap_js:call_with_module/2:(execute(user:operate/1),8)'))
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:fix_current'/0,in('bootstrap_js:call_with_module/2:(execute(user:fix_current/0),5)'))
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:calculate'/0,in('bootstrap_js:call_with_module/2:(execute(user:calculate/0),5)'))
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:add_digit'/1,in('bootstrap_js:call_with_module/2:(execute(user:add_digit/1),8)'))
proscriptls.js:9589 Uncaught exception:
existence_error(procedure,'user:add_digit'/1,in('bootstrap_js:call_with_module/2:(execute(user:add_digit/1),8)'))

https://lindseyspratt.github.io/proscriptls/examples/calculator.html

And on GitHub, I dont see a way to raise issues.

https://github.com/lindseyspratt/proscriptls

Where are the issues?

j4n bur53 schrieb:

[Transfinite Numbers]

I wrote "thinking about debugging actors again".
Here I found something in the wild:

Intro: Jaeger - Yuri Shkuro, Uber & Pavol Loffay, Red Hat
https://www.youtube.com/watch?v=WRntQsUajow

[Transfinite Numbers]

Some reasoning about parallel streams, prototyped
in Prolog, and then transfered to the Scala type system:

Daniel Beskin—Compile-time Logic Programming in Scala: what for?
https://www.youtube.com/watch?v=9i06TyYM_lI

[Transfinite Numbers]

Putting debugging on the agenda again could be indeed
rewarding. There was an interesting remark by Jan W.
on SWI Discourse about determinism modes and debugging.

So I tried it myself, now from Jekejeke Prolog,
without any changes to the system:

Jekejeke Prolog 4, Development Environment 1.4.1 (20 August 2019)
(c) 1985-2019, XLOG Technologies GmbH, Switzerland

?- [user].
app([],X,X).
app([X|Y],Z,[X|T]) :- app(Y,Z,T).
^D
Yes

?- debug, leash([]), visible([fail]), spy(app/3).
Yes

?- app([1,2],[3],[1,2,3]).
Yes

?- app([1,2],[3],[1,4,3]).
* 1 Fail app([2], [3], [4, 3])
* 0 Fail app([1, 2], [3], [1, 4, 3])
No

I am not sure whether new trace/2 as in SWI is
necessary. Maybe its possible to make debug mode
more intelligent and let its implementation check

for the leash mode and do some optimizations. But
trace/2 could be nevertheless interesting, since
I guess it allows more modes on a per predicate basis.

[Transfinite Numbers]

It would be interesting to see how the different
determinism mode specifier would map to different
leash/visible combinations.

But I think its important that the debugger can
do choice point optimization, i.e. drop redo ports
when there are no choice points left.

My take on a mapping would then be, multiplicity
is how database people would write it:

Category Multiplicity Offending
det 1 fail, redo
semidet c redo
multi m fail
nondet mc

But how do you deal with predicates that
are multimoded, like:

append(+list(T), +list(T), -list(T)) is det.
append(+list(T), -list(T), +list(T)) is semidet.
append(-list(T), +list(T), +list(T)) is semidet.

Maybe what would be needed is a couple of
breakpoint test builtins and a couple of
breakpoint action builtins, and then

introduce the concept of advice point.
From SICStus Prolog:

" mentioned earlier, there are two kinds of
breakpoints: spypoints and advice-points.
The main purpose of spypoints is to support
interactive debugging. In contrast with this,
advice-points can help you to perform non-
interactive debugging activities."
https://sicstus.sics.se/sicstus/docs/4.0.2/html/sicstus/Advice_002dpoints.html#Advice_002dpoints

This is easiest done if the debugger itself is
written in Prolog, although keeping the overhead
low is probably an art.

[Transfinite Numbers]

What could be a more modest and open proposal than
Web Prolog?. Well we can inspect Web Prolog since
it is open source. For example rpc/2 uses:

setup_call_cleanup(
http_open(ExpandedURI, Stream, Options),
read(Stream, Answer),
close(Stream)),
wait_answer(Answer, Query, Offset, Limit, QueryAtom, Parts, Options).
https://github.com/Web-Prolog/swi-web-prolog-lite/blob/master/rpc.pl

So setup_call_cleanup/3 is used for the purpose to
secure that the stream resource is released. It
could be also used to detect a cut in the

continuation and to cleanup the remote server. But
it is used in the prototype and looks as indispensable,
I guess any new Web Prolog standard efforts would

need this control construct also make a standard first.
Currently its not part of the ISO core standard, but
nevertheless featured by the WG:

ISO/IEC JTC1 SC22 WG17
Draft proposal for setup_call_cleanup/3
https://www.complang.tuwien.ac.at/ulrich/iso-prolog/cleanup

The scope of setup_call_cleanup/3 are of course also network
resources. One could also make as a standards goal to
iterate the ISO core standard, and for example add

setup_call_cleanup/3 to it. Another standard that also
evolved was the COBOL ISO/IEC standard is ISO/IEC
1985:2014 standard. This would make the ISO Prolog

processor agnostic, whether the ISO Prolog processor
is really used for web prolog or not. But it would
kind of make every complying Prolog processor fit

also for web prolog.

[Transfinite Numbers]

Onions, Onions, Onions. How to sell a language?
Take Java, what does interests you more?

This onion (some made up onion shells):

+------------------------+
| Byte Code JIT |
| +------------------+ |
| | GC | |
| | +------------+ | |
| | | Threads | | |
| | | ... | | |
| | +------------+ | |
| | | |
| +------------------+ |
| |
+------------------------+

Or this onion (some made up onion shells):

+------------------------+
| java.sql.* |
| +------------------+ |
| | java.util.* | |
| | +------------+ | |
| | | Java | | |
| | | .. | | |
| | +------------+ | |
| | | |
| +------------------+ |
| |
+------------------------+

Same with Web Prolog. How can we get the full picture
whether Web Prolog is worth the effort. Through
a picture like this:

+------------------------+
| Actor |
| +------------------+ |
| | ISOTOPE | |
| | +------------+ | |
| | | ISOBASE | | |
| | | ... | | |
| | +------------+ | |
| | | |
| +------------------+ |
| |
+------------------------+

Or maybe a picture like this:

+------------------------+
| TRUSTED SMART ACTORS |
| +------------------+ |
| | SMART ACTORS | |
| | +------------+ | |
| | | ACTOR | | |
| | | ... | | |
| | +------------+ | |
| | | |
| +------------------+ |
| |
+------------------------+

[Transfinite Numbers]

The usual setup_call_cleanup/3 has a drawback
concerning the modelling of Erlang inside Prolog.
Lets assume the call inside setup_call_cleanup/3

is a perpetual process. This means this call will
never reach this point here:

setup_call_cleanup(
Setup,
Call, %%% exist never reached when Call is perpetual
Cleanup)

As a result Cleanup is also never called. Lets have
a look under the how, how setup call cleanup is
realize in my own system:

setup_call_cleanup(A, G, C) :-
sys_atomic((once(A), sys_cleanup(C))),
current_prolog_flag(sys_choices, X),
G, %%% exist never reached when G is perpetual
current_prolog_flag(sys_choices, Y),
(X == Y, !; true).

sys_atomic/1 and sys_cleanup/1 safely install a
cleanup handler, but this cleanup handler is never
triggered since G never exists.

We need some information about some commitment of
G, concerning the resources it is using, and concerning
whether he doesn't need any resources anymore.

Take the receive construct, and assume it needs some
lock resources. A good moment to release the lock resource
is also when we are finished with scanning the

actor inbox and do remove our message of choice:

%% some lock is acquired to listen for the mbox
clause(mbox(X),true,Ref),
(condition1 -> erase(Ref),
%% good moment to release the lock again
action1;
condition2 -> erase(Ref),
%% good moment to release the lock again
action2;
...
conditionn -> erase(Ref),
%% good moment to release the lock again
actionn;
fail).

So how could this be done?

[Transfinite Numbers]

Probably we need goal expansion, since creating
a cut transparent goal skeleton would help. We
can then use a cut to trigger the cleanup:

sys_atomic((once(...), %%% do lock acquire here
sys_cleanup(...))), %%% register lock release here as cleanup
clause(mbox(X),true,Ref),
(condition1 -> erase(Ref),
!, %%% will trigger cleanup
action1;
condition2 -> erase(Ref),
!, %%% will trigger cleanup

action2;
...
conditionn -> erase(Ref),

!, %%% will trigger cleanup
actionn;
fail).

This translation has the same benefit as setup_call_cleanup/3,
namely that the cleanup is called in case an error happens
inside the call.

But the translation has the additional benefit that it
can deal with perpetual calls, that do not exit and run
forever.

[Transfinite Numbers]

Maybe the below design is a little over the
top, basically in actors, the actor is the
only consumer of his own mail box.

The below design might be needed in some
other circumstances, but I guess a setup_call_cleanup/3
before the mail box is analyzed and with

the help of the logical cursor semantics,
everything works fine. So I was a little over-
pressimistic here. As a result we don't need

cuts, and can go with if-then-else solely!

[Mostowski Collapse]

Ok a simple design for Prolog RPC with setup_call_
cleanup/3 via the actor model is possible. Crucial
is the following little addition to our library(

experiment/broker) prototype:

exit(P, M):
The predicate succeeds in interrupting the actor
P by an error M. Exit is blocking until acknowledge
is received.

This primitive does not use the actor inbox
but directly affects an actor. Further exit/2 is
idempotent in that it does nothing, when the

actor already exited. This prevents orphan actors
when a Prolog RPC call is cutted away by a cut
or exception in the continuation. We also added

handling of events inside the Prolog RPC call
itself, so that the Prolog RPC call can do choice
point elimination and promote exceptions:

?- rpc('localhost:3010', father(tom, X)).
X = sally ;
X = erica

?- rpc('localhost:3010', X is 1/0).
Error: Division by zero.
is/2

See also:

Preview: Prolog RPC via the actor model. (Jekejeke)
https://gist.github.com/jburse/ee89628428b391199c86b866cd3c2dfc#gistcomment-3114824

[Mostowski Collapse]

We can do a comparison with some Prolog RPC
prototypes surround SWI-Prolog. This is one
take from Torbjorn Lager:

rpc(URI, Query, Options) :-
pengine_spawn(Pid, [
node(URI),
exit(true)
| Options
]),
options(pid(Pid), Options, Pid),
pengine_ask(Pid, Query, Options),
wait_answer(Query, Pid).

Also comparing with:

https://github.com/Web-Prolog/swi-web-prolog-lite/blob/master/rpc.pl

Main differences when comparing to our recent
Prolog RPC prototype. Some marks of our Prolog
RPC prototype are:

* Pure Erlang primitives realization, no pengine_xxx.
* Pure Erlang primitives realization, no http_xxx.
* No separate pengine_ask needed for the first solution.
* Repeat fail realization of the wait_answer loop.
* Automatic tear down via exit/2 from events in the continuation.

The tear down can be seen in this example where a cut
happens in the continuation. No stop message support is
needed by the remote part of the Prolog RPC:

?- rpc('localhost:3010', father(tom, X)), !.
X = sally

Wont leave any orphaned actor on the remote site.
Instead simply the remote part is immediately exited
with the help of a local setup_call_cleanup/3 usage.

Maybe this complete reduction to the actor model has
a few advantages:

* transport: We can delegate provisioning of some
transport to the actor platform, its not necessary
to write a different rpc for different actor platforms.
As long as the end point that we want to invoke is
representable as what actor platform spawn understands,
rpc/2 should work.

* interleaving: In as far as the actor platform
allows multiplexing, multiple rpc calls can
interleave their messages that are exchanged
between local part and remore part, increasing
resource utilization.

* state: Like actors, the RPC call will optionally
have state. At the remote part of the Prolog RPC
the execution can involve side effects and blocking,
in particular backtracking will always adress the
same actor in the remote site.

Open issues:

security: It might be desirable to identify the
messages that the remote part sends back to the local
part as replies, and have some rules in place. Best
would be if the actor platform would already provide
this security.

I saw that the Java DatagramSocket calls SecurityManager.
I guess its possible to implement policies that an
applet is only allowed to commmunicate with its
originating server, etc... But then maybe we also
want a SecurityManager on the server.

[Mostowski Collapse]

Ok we solved some issue about cropped stack
traces during Prolog RPC. We used JDK 1.7
suppressed exception lists. Unfortunately,

because of the Android platform, we cannot
use JDK 1.7 only JDK 1.6. So we used what
we had already rolled in the past in the

form of:

/**
* Make an engine exception from two engine exceptions.
* Will do the following transformation:
* Before: First = cause(X1,..cause(Xn-1,Xn)), Second = Y.
* After: Result = cause(X1,..cause(Xn-1,cause(Xn,Y))).
*/
public EngineException(EngineException e1, EngineException e2);

There is now a convenience predicate throw/2 which
will also automatically fill the stack trace of
the suppressed exception that is to be appended.

Here is an example run:

?- [user].
foo :- rpc('localhost:3010', _ is 1/0).

Yes
?- foo.

Error: Division by zero.
is/2

Error: Got error from remote.
foo/0


We noticed that the solution is not perfect.
The internationalization of an exception
during printStackTrace() is done locally,

we might not have all remote text properties.
Maybe printStackTrace() should also do rpc/2?

See also:

Preview: New predicate throw/2 in module "control". (Jekejeke)
https://gist.github.com/jburse/ee89628428b391199c86b866cd3c2dfc#gistcomment-3115919

[Mostowski Collapse]

A new module "bytes" is mainly motivated by the actor
model we are pursuing. We noticed that we might need
in the future some way to concatenate messages parts

that were originally one message. The API inside
the new module "bytes" consists of:

* memory_write(S):
The predicate succeeds in S with a new write
memory socket.
* memory_read(B, S):
The predicate succeeds in S with a new read memory
socket for the block B.
* memory_get(K, B):
The predicate succeeds in B with the block
of the Prolog output stream K.
* memory_get(K, O, B):
The predicate succeeds in B with the atom of
the Prolog output stream K in the encoding options O.

Since the module "bytes" is located in the package
"structure" from the runtime, memory sockets are
now part of the runtime. Before there was a module

in the development environment which is now obsolete.
We already used the memory sockets to bootstrap the
new predicates term_block/[2,3].

See also:

Preview: New Prolog module "bytes". (Jekejeke)
https://gist.github.com/jburse/be96f5b89eacc303cb9c8730f4313da6#gistcomment-3125559

[Mostowski Collapse]

My latest rpc/3 implementation uses setup_call_cleanup/3.
But it also uses catch/3 which unformtunately in my
system uses Java native stack.

But I noticed there is potential for another predicate
like setup_call_cleanup/3. By looking at SWI-Prologs
new intercept/3.

I guess intercept is an instance of a more general
new logical predicate. It might be not new to all Prolog
systems. And it is only logical in as far as it has

goal arguments.

* try_call_finally(Try, Goal, Finally):
The predicate succeeds whenever Goal succeeds. It calls
Try during call and redo port and it calls Finally
during exit and fail port. It also calls Finally
during exception port. Further in can optimize away
redo fail pairs, if the Goal succeeded without leaving
choice points.

This can be used to implement intercept:

intercept(Goal, Ball, Handler) :-
try_call_finally(
asserta(signal_handler(Ball, Handler)),
Goal,
once(retract(signal_handler(_,_))).

send_signal(Ball) :-
once(signal_handler(Ball, Handler)),
Handler.

The overhead of intercept is related to overhead
of a try_call_finally/3. In my Prolog system I have
not yet an optimal solution, since I need Java native
stack. But I noticed a relationship to

setup_call_cleanup/3 as follows.

Port try_call_finally setup_call_cleanup
Call Yes Yes
Exit Yes No
Redo Yes No
Fail Yes Yes
Exception Yes Yes
Cutter No Yes

Also both predicates detect when there are no
more choice point and optimize away. Call the
Finally respectively Cleanup more earlier. I only
added determination check recently to my try_call_finally/3
and it is currently used in with_input_to/2 and
with_output_from/2.

The later predicates now show choice point elimination.
In principle try_call_finally/3 could be standardized
like setup_call_cleanup/3. It has a lot of use cases.
intercept/3 being one use case. But also with_input_to/2
and with_output_from/2 are an example use case.

I also hope I can provide an implementation that does
not use Java stack in the future. But for this I need
to adopt the setup_call_cleanup/3 design and find a way to
not call Finally during the cutter port. This is the port
for a cut or exception in the continuation.

So its still work in progress.

[Mostowski Collapse]

Maybe yet another new built-in could be useful,
namely one that combines the setup_call_cleanup/3
and catch/3 in my Web Prolog rpc/3 implementation.

Not yet sure. They do different things, setup_call_cleanup/3
exits the remote actor. And catch/3 does some ball
wrapping. Maybe just find a new catch/3 which doesn't

use Java native stack. If I could controll the cutter
port, I could also implement a different catch/3.
Currently setup_call_cleanup/3 uses under the hood

sys_cleanup/1 which installs a choice point Goal.
But we could also install a choice point Closure and
call the Closure with a flag that indicates whether

we are cutting from the continuation or inside the goal.
I think there is a way to detect this situation since
its already detected during my environment trimming.

[Mostowski Collapse]

Corr.:
Maybe this here is better:

send_signal(Signal) :-
once(signal_handler(Ball, Handler)),
Ball = Signal,
Handler.

[Mostowski Collapse]

Here is a follow up to the cause/2 Prolog throwable
ball. I borrowed the idea from Java. Its found in
many places in Java in different incarnations.

For example a place I just discoverd recenty,
where something like this is used, is SQLException:

public class SQLException
extends Exception
implements Iterable<Throwable> /* this is special here */
void setNextException(SQLException ex) /* this is also special here */
SQLException getNextException() /* this is also again special here */
https://docs.oracle.com/javase/7/docs/api/java/sql/SQLException.html

Now there were some further questions on SWI
discourse about the new intercept/3. My own
opinion is, that it has some holes.

The proposal for intercept/3 is to call
send_signal/1. But maybe the library also
provides a pre-bundled useful signal handler?

How would one program a handler that does the
same as SQLException, so that after the intercept/3,
we can also somewhere grab the list of exceptions?

Does the library offer something like that? For my
cause/2 proposal I have recently added a snoc function
(LISP cons at end), which will be available with

upcoming release 1.4.2. This was described below.
What is a snoc function. Well since the cause/2
construct is like LISP consing device,

snoc is also a little LISP inspired lingo:
https://stackoverflow.com/questions/7674277/in-functionaljava-list-what-does-snoc-mean

[Mostowski Collapse]

Now this snoc causes me a new headache. It is not
that it has some errors. It wasn't implemented in
a destructive way by some point swizzling.

When we went from a Java method, which was already
treating the input arguments as values, and returning
a new value, we went to a new Predicate is

sys_append_cause/3, again not using anything
destructive. But the new Prolog snoc, was made the
in context of Web Prolog RPC to chain

exceptions vertically. But collecting syntax errors
is rather something to collect exceptions horizontally.
I do not yet have a means to distinguish the two.

[Mostowski Collapse]

But here is already an idea to distingusih the
two. The idea can be easiest explained at hand of
the exception framework we have in Jekejeke Prolog.

Jekejeke Prolog has some special built-ins that
do not have access to the stack trace. But nevertheless
they need to be able to throw an exception.

For this purpose we have InterpreterMessage, which
can be extended to InterpreterException by adding
a stack trace. So we made Java classes for this

common ISO Prolog convention:

error(Error_term, Implementation_defined_term)
http://www.univ-orleans.fr/lifo/software/stdprolog/exceptions.html

The Java classes are:

- Java class InterpreterMessage:
This class captures the error term, i.e. examples:
syntax_error(implementation_dependent_atom)
evaluation_error(Error)
Etc..

- Java class InterpreterException
The class captures the throwable ball itself, i.e. examples
error(_, _)
warning(_, _) /* this is special here */
cause(_, _) /* this is also special here */

The cause/2 construct allows "vertical" chaining. This
is a chaining on the level of Java class InterpreterException.
We could also introduce a chaining of the level

of Java class InterpreterMessage. This would especially
facilitate the chaining of syntax errors. The corresponding
new construct would allow "horizontal" chaining.

I am thinking about this here:

multi(_, _)

or this here (more conform to the usus of xxx_error):

multi_error(_, _)

If we have a consing device on the InterpreterMessage level,
we can easily provide vertical and horizontal chaining.
Woa! Cool!

[Mostowski Collapse]

There is one remaining problem. Where put line number,
column number, or a copy of the line, whatever, which
is usually put into the Implementation_defined_term part.

This is seen here:

Welcome to SWI-Prolog (threaded, 64 bits, version 8.1.19)

?- catch(read(X), E, true).
1 + .
E = error(syntax_error(operator_balance), stream(user_input, 8, 7, 337)).

And here:

Jekejeke Prolog 4, Runtime Library 1.4.2 (November 12, 2019)

?- catch(read(X), E, true).
1 + .
E = error(syntax_error(cannot_start_term),
[text_pos('1 + .\n ^'), pred(read/1), pred(catch/3)]).

For "horizontal" chaining, maybe this needs to be
abandonned. It would be more easier if a syntax_error
term would already carry such information such

as line number, column number, or a copy of the line,
whatever. Otherwise we would need even more grouping.
Like two different horizontal levels,

one level where all errors on the same line are
grouped, and one level for such pakets of errors
and the line differs. But we might also repeat

line information, and leave it to tooling to perform
grouping. A lot of new Java tools and output logs do
such grouping. I have never studied them.

Having such "horizontal" chaining available does not
mean that we totally need to abandon intercept/3 or
some such. It might be still a good design to have

some predicates emitting errors into some sink
to keep the snoc small. But I am currently facing
a parsing problem, where snocing of multiple errors

would be a fantastic solution.

[Mostowski Collapse]

BAT. Somebody recently told me he dropped
a BAT on Github. Where is my BAT?

But what is BAT?

It all happens on the Ethereum blockchain.
And we see Brandan Eich again and a new browser.
https://vimeo.com/209336437

[Mostowski Collapse]

Corr.:
Brendan Eich

[Mostowski Collapse]

Maybe should run a same sex marriag campaign
via BAT. Or do all brave browsers automatically
filter LGBT content? LoL

same sex marriag even legal in Switzerland
https://www.blick.ch/people-tv/schweiz/id15709162.html

[Mostowski Collapse]

coronavirus coming from a P4 lab in wuhan
itself? If this is true, you China are
officially declared stupid.

I saw a TV special about some german lab where
they studied some pig viruses. Question is
how livestock handles some viruses, that

usually don't exist in the borders of EU.
The lab was on an island far away from
civilisation, and you were not allowed to

visit it as a reporter.

https://en.wikipedia.org/wiki/Friedrich_Loeffler_Institute

[Mostowski Collapse]

About the P4 lab:
https://www.dailymail.co.uk/health/article-7922379

[R Kym Horsell]

Interesting. I hear they just discovered swine fever (ASF) in
wild pigs in Poland nr the German border.

--
Anyone who, when faced with reality denies it and then talks down to you
is just a damned putz, that's all there is to it.
-- Marvin the Martian aka Mike Varney [well-known climate denier], 27 May 2012

[Mostowski Collapse]

Maybe China doesn't need Smart Cities,
rather Safer Hospitals or dunno what.

Maybe Traditional Chinese medicine (TCM)
isn't apt to modern challenges.

If the virus has a long incubation time,
it probably spread already around the globe.

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> Maybe China doesn't need Smart Cities,
> rather Safer Hospitals or dunno what.
> Maybe Traditional Chinese medicine (TCM)
> isn't apt to modern challenges.
> If the virus has a long incubation time,
> it probably spread already around the globe.

Early modeling (Eric Taylor, Johnathon Read) seems to say at least 100s of
cases were expected to already be in major population centers and the lock-
down was "ineffective".

The latest models have brought down estimates of the "R_0" parameter
(ratio of 2nd infections per each primary) closer to seasonal flu
from Read's "upto 3.8".

But a lot remains unknown. We'll just have to wait and see.

On the down side they still haven't stopped MARS from a few years back.
It's still killing.

[R Kym Horsell]

R Kym Horsell <k...@kymhorsell.com> wrote:
> Mostowski Collapse <burs...@gmail.com> wrote:
>> Maybe China doesn't need Smart Cities,
>> rather Safer Hospitals or dunno what.
>> Maybe Traditional Chinese medicine (TCM)
>> isn't apt to modern challenges.
>> If the virus has a long incubation time,
>> it probably spread already around the globe.
> Early modeling (Eric Taylor, Johnathon Read) seems to say at least 100s of

^ Toner

> cases were expected to already be in major population centers and the lock-
> down was "ineffective".
> The latest models have brought down estimates of the "R_0" parameter
> (ratio of 2nd infections per each primary) closer to seasonal flu
> from Read's "upto 3.8".
> But a lot remains unknown. We'll just have to wait and see.
> On the down side they still haven't stopped MARS from a few years back.

^ MERS
> It's still killing.


From email. :)

[j4n bur53]

Hashtag #武汉肺炎 on twitter.

R Kym Horsell schrieb:

[Mostowski Collapse]

There is a new dimension in batshit crazy.

WHO has leveled the global risk to "high".

Cambodia has a first confirmed case.

[Mostowski Collapse]

So there is a deadly coronavirus in cats:
https://www.anicura.de/wissensbank/katzen/fip-bei-katzen/

And china is known for cat eating:
https://en.wikipedia.org/wiki/Cat_meat#China

Could be also a source, who knows?

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> So there is a deadly coronavirus in cats:
> https://www.anicura.de/wissensbank/katzen/fip-bei-katzen/
> And china is known for cat eating:
> https://en.wikipedia.org/wiki/Cat_meat#China
> Could be also a source, who knows?

...

Virologists say it's 96% similar to a known bat cov
but possibly came via some other mammal. All the sequencing
done up to about 1 wk back showed samples in human
patients evolved only recently from a common ancestor
around Oct 2019.

[Mostowski Collapse]

I experienced some problems using this
tag and Seamonkey browser. Somehow too
much multimedia and browser started choking.

So I switched to Chrome, which was better.
Today I installed this version of Seamonkey.
SeaMonkey 2.53.1 Beta 1 released

https://www.seamonkey-project.org/releases/2.53.1b1

The about:networking page tells me it does
use HTTP/2, for example on the google search
start page. Also twitter browsing now better.

But some Firefox November 2019 build should
also have HTTP/3. Keeping an eye on these things
because of unfinished Actor Model work.

Lets say in 2 weeks or so, China has overcome
the epidemic, because its too warm and the
virus anyway cannot survive. Who knows?

Maybe there is a backslash in 2020 autumn.
In either case, China might pick oneself up,
although with a high number of casualities.

So the show must go on!

Am Sonntag, 26. Januar 2020 23:35:48 UTC+1 schrieb Mostowski Collapse:

[Mostowski Collapse]

Nevertheless I think it is crazy, that these P4
labs are planned in big cities. Other cities

are Beijing and Kunming according to this
Nature article:

Inside the Chinese lab poised to
study world's most dangerous pathogens
https://www.nature.com/news/1.21487

Thats like internet access without firewall
etc.. Or unprotected sex.

[Mostowski Collapse]

The Nature article from 22 February 2017
has a strange soap box. Dont know what

the intention was. Research has no borders?
Spreading has no borders?

“Viruses don’t know borders.”

[j4n bur53]

Nature: Bat cave solves mystery of deadly SARS virus
01 December 2017 - David Cyranoski
https://www.nature.com/articles/d41586-017-07766-9

So how did the virus travel from Kunming, Yunnan
to Wuhan and mutate? Maybe animal meat was imported.
Or the virus took the direct shuttle

via some lab probes distributed by P4 labs. Such
a shuttle is cosy a nice, no temperature or UV
problem, if you are not a virus that can spread

via dust. These P4 labs are ideal to extend
the reach of any virus.

LoL

R Kym Horsell schrieb:

[Mostowski Collapse]

Think I am excaggerating? Why the assumption
the virus is man made? Isn't natur itself
much better in playing the random game?

Could be that my viewpoint is distored,
to much believe in artificial things, like
for example artificial intelligence, are

partially possible. Now we hear from research
race towards gene editing, especially for
reproduction science. And here is an example

of a former computer scientist, now
with microsoft, who turned into molecular
programming, with the explicit goal of:

The goal of molecular programming
is the systematic manipulation of
matter at the molecular scale, for
applications in engineering + technology
and biology + medicine.
http://lucacardelli.name/

“Viruses don’t know borders.” There could
be yet another meaning, no borders between
computer viruses and biological viruses.

[Mostowski Collapse]

In as far this brilliant film:

Contagion (2011) - Steven Soderbergh
https://www.youtube.com/watch?v=4sYSyuuLk5g

could be outdated and whitewashing.
Nevertheless hope I didn't spoil it,
see and enjoy (sic!) for yourself.

[j4n bur53]

Woa!

"Scientists have engineered the the world’s first pigs
with Huntington’s disease in a development that could
lead to a breakthrough for humans with the condition.

The joint Chinese-American team used a genetic editing
technique to introduce a segment of the human gene
which causes the neurodegenerative disease."
https://www.telegraph.co.uk/science/2018/03/30/huntingtons-disease-pigs-created-hope-gene-editing-breakthrough/

Mostowski Collapse schrieb:

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> Think I am excaggerating? Why the assumption
> the virus is man made? Isn't natur itself
> much better in playing the random game?

>...

Don't multiply assumptions. :)

But spooky parallel -- the Wuhan virus has been "re created"
in at least one lab in Melbourne AU.

The Chinese released the sequence (Wuhan is an RNA virus)
and a lab here in Melb rebuilt the sequence and by some magic
re-created the 3d topology to make something that seems
to be "the same" as the animal virus.

It theory is is possible to take a wild virus, tinker with in
using sequences and re-squencers, and build a disease that is
active in humans.

But "from scratch" maybe beyond the state of the art at present.

In Nature there are 1000s of coronaviruses (which already seem to
have an affinity to jump between mamals either infecting the gut
and/or lungs in some) and millions of viruses. It's only a matter of
time before one jumps from some species or even out of a melting
iceberg somewhere into someone and/or vice versa.

And around and around it goes, reversing the usual flow of entropy. :)

[Mostowski Collapse]

Yeah, USA and Australia send interesting
signals (about insider knowledge?).

For USA, recouperation of expats, what do
they do for quaranteen? Well they fly them
first to Alaska (sic!). Rumors or true?

Source, middle of this YT video:
Inside Wuhan: Q&A - Ben Kavanagh
https://www.youtube.com/watch?v=D4vFj5GsBnE

You can find similar information about
evacuation by Australia. Something with
Hazmat dunno what.

I like this approach, precautions, pre-
cautions, ...

[Mostowski Collapse]

The virus has an extrem high mutation rate,
in meme land (trends in Switzerland). Yesterday
the following was trending:

#coronarvirus

today is trending:

#coronaravirus

and tomorrow:

#Carbonaravirus

Go, go, you can do it!

[j4n bur53]

"research security" or small-minded trade war?
https://cen.acs.org/research-integrity/misconduct/Harvard-chemist-Charles-Lieber-charged/98/i5

Mostowski Collapse schrieb:

[Mostowski Collapse]

Machine Learning result:

Phylogenetic Tree of Early Coronavirus
http://virological.org/t/clock-and-tmrca-based-on-27-genomes/347

We are starting to see more structure in
the tree and overall the genetic data is
highly suggestive of a single-point introduction
into the human population followed by sustained
human-to-human transmission.

Median 95% HPD
01 Dec 2019 20 Oct 2019 - 20 Dec 2019

[Mostowski Collapse]

Not sure what the Location and Date
in the samples label says? Patient
Location or Lab Location?

So the virus traveled continents
already in a few weeks?

[j4n bur53]

Ok, now I see why CNBC is always talking
about day zero could be november 2019.

The pinned thread says it:

Phylodynamic Analysis | 44 genomes | 29 Jan 2020
http://virological.org/t/phylodynamic-analysis-44-genomes-29-jan-2020/356/1

There it is mentioned:

Estimated TMRCA 95% interval
22-Nov-2019 19-Oct-2019 – 17-Dec-2019

Mostowski Collapse schrieb:

[Mostowski Collapse]

Website biorxiv currently not reachable.
Possibly about this here?

"Indian scientists have just found HIV (AIDS)
virus-like insertions in the 2019-nCov virus
that are not found in any other coronavirus."
https://defence.pk/pdf/threads/indian-scientists-find-novel-coronavirus-possibly-designed.651871/

Not sure whether this is a conspiracy theory.
HIV and Corona Virus anyway share some similarity,
for example a compression algorithm (sic!):
https://en.wikipedia.org/wiki/Ribosomal_frameshift

Read about this compression mechanism a few days
ago. Not sure whether gene researches have too much
fantasy or not. LoL

Such slipper slopes could be also timers?
Just guessing. Like a coin flip?

[Mostowski Collapse]

Maybe a timer would be too unreliable.
More some "translational bifurcation".

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> Website biorxiv currently not reachable.
> Possibly about this here?
> "Indian scientists have just found HIV (AIDS)
> virus-like insertions in the 2019-nCov virus
> that are not found in any other coronavirus."
> https://defence.pk/pdf/threads/indian-scientists-find-novel-coronavirus-possibly-designed.651871/
> Not sure whether this is a conspiracy theory.
> HIV and Corona Virus anyway share some similarity,
> for example a compression algorithm (sic!):
> https://en.wikipedia.org/wiki/Ribosomal_frameshift
> Read about this compression mechanism a few days
> ago. Not sure whether gene researches have too much
> fantasy or not. LoL
> Such slipper slopes could be also timers?
> Just guessing. Like a coin flip?

There is a lot of junk going around. :)

The fine print is Feigl-Ding et al found sections "similar to" HPV1.

So we have to ask what is the chance that a virus -- one
30,000 bases long (ncov) and the other 10,000 bases long (HIV) --
will have sequences that "look like" some other virus and, related, how
long and similar a section would need to be to be a NON-CHANCE resemblance. :)

You might suspect "pretty long" and "pretty similar".

The folks at YCombinator have looked at this story and come down
on the side of it being down to pure chance.
They point out there are other parts in the 2019ncov that look like
parts of other viruses as well.

I think there is also some doubt anyone building a virus weapon would
gravitate to splicing in HIV in particular.

If they had some special reason it doesn't make much sense to insert
anything less than a complete gene. HIV has 9 genes roughly
1000 bases long on average.

If you found something less than 1000 bases in common then it
is not likely the result of deliberate editing or -- if so -- it's a
botched job and they need to hand back their evil scientist membership card.

[Mostowski Collapse]

Now every country is showing their ABC competency:

C-130 Military
https://twitter.com/Bundesheerbauer/status/1223588489813536768

The Brits were chicken, and used a Spanish crew.
Why were they exiting EU exactly?

LoL

[Mostowski Collapse]

"Brexit gives Britain a chance to fix its immigration policy"
https://www.republicworld.com/world-news/uk-news/brexit-uks-chance-to-fix-immigration-system-restore-control.html

How about an immigration policy for viruses? LoL
Yeah 'make the most of opportunities' and 'unleash the potential'.

[Mostowski Collapse]

1918: Ships, Troops
2019: Airplanes, Civilians

Which one is faster?

1918 Spanish Flu historical documentary
https://www.youtube.com/watch?v=UDY5COg2P2c

[j4n bur53]

The results of the study were published in October
last in the PLOS Neglected Tropical Diseases
journal, originally established by the Bill
and Melinda Gates Foundation.

Coronavirus : Wuhan institute's study on bats
and bat hunters in Nagaland to be probed
NEW DELHI , February 03, 2020 01:01 IST
https://www.thehindu.com/news/national/study-on-bats-and-bat-hunters-of-nagaland-come-under-the-scanner/article30722099.ece

Mostowski Collapse schrieb:

[j4n bur53]

Thats hard to digest, just for the record:

"one mountain does not allow two tigers to live together"
https://greatgameindia.com/coronavirus-chinas-secret-plan-to-weaponize-viruses/

j4n bur53 schrieb:

[j4n bur53]

But it is not yet perfect enough

"Eyes of Darkness" Autor Dean Koontz - Wuhan-400
https://books.google.de/books?id=EdKNU1DzdbcC&pg=PT261&dq=Wuhan-400&hl=de&sa=X&ved=0ahUKEwjMw46sy7bnAhVOxIsKHS4dD8oQ6AEILDAA#v=onepage&q=Wuhan-400&f=false

LoL

j4n bur53 schrieb:

[R Kym Horsell]

R Kym Horsell <k...@kymhorsell.com> wrote:

> Mostowski Collapse <burs...@gmail.com> wrote:
>> Website biorxiv currently not reachable.
>> Possibly about this here?
>> "Indian scientists have just found HIV (AIDS)
>> virus-like insertions in the 2019-nCov virus
>> that are not found in any other coronavirus."
>> https://defence.pk/pdf/threads/indian-scientists-find-novel-coronavirus-possibly-designed.651871/
>> Not sure whether this is a conspiracy theory.
>> HIV and Corona Virus anyway share some similarity,
>> for example a compression algorithm (sic!):
>> https://en.wikipedia.org/wiki/Ribosomal_frameshift
>> Read about this compression mechanism a few days
>> ago. Not sure whether gene researches have too much
>> fantasy or not. LoL
>> Such slipper slopes could be also timers?
>> Just guessing. Like a coin flip?
> There is a lot of junk going around. :)
> The fine print is Feigl-Ding et al found sections "similar to" HPV1.
> So we have to ask what is the chance that a virus -- one
> 30,000 bases long (ncov) and the other 10,000 bases long (HIV) --
> will have sequences that "look like" some other virus and, related, how
> long and similar a section would need to be to be a NON-CHANCE resemblance. :)

...

Bingo.

Quick retraction of a faulty coronavirus paper was a good moment for science
Statnews, 03 Feb 2020
Titled "Uncanny similarity of unique inserts in the 2019-nCoV spike protein
to HIV-1 gp120 and Gag," the paper claimed to find similarities between the
new coronavirus and HIV, the virus that causes AIDS. The use of the word
"uncanny" in the title, together with "unlikely to be fortuitous" in the
abstract, led some to think that the authors were suggesting the virus had
somehow been engineered by humans.
The paper, from academic institutions in New Delhi, India, was critical and
alarming, if true. Except that it wasn't.
The paper was almost immediately withdrawn, but not before plenty of
handwringing from researchers who complained that the appearance of such
shoddy work on a preprint server without vetting by peer reviewers is
precisely why the hoary old model of science publishing is better at keeping
junk science out of the literature.

[j4n bur53]

The paper is still there:

Uncanny similarity of unique inserts in the
2019-nCoV spike protein to HIV-1 gp120 and Gag

https://www.biorxiv.org/content/10.1101/2020.01.30.927871v1

Another paper is also interesting:

The novel coronavirus 2019 (2019-nCoV) uses the SARS-
coronavirus receptor ACE2 and the cellular protease TMPRSS2
for entry into target cells
https://www.biorxiv.org/content/10.1101/2020.01.31.929042v1

R Kym Horsell schrieb:

[Mostowski Collapse]

From the second paper, blame it on food or
animal closeness then:

"Thus, SARS-CoV was identified in Civet cats and
raccoon dogs, which are sold as food sources
in Chinese wet markets."

So how solve the problem? Maybe like this:

"The Jewish laws of kashrut and Islamic dietary
laws both forbid the consumption of cat meat.
Kashrut and Halal Laws both disallow the consumption
of any terrestrial predators. To be considered
kosher in the case of mammals, it must not be a
predator and it must both chew cud and have
cloven hooves."
https://en.wikipedia.org/wiki/Cat_meat#Jewish_and_Islamic_tradition

LoL

[Mostowski Collapse]

"This rule thus excludes the camel from
the list of kosher animals because although
the camel does ruminate, it does not possess
true "hooves" – it walks on soft toes which
have little more than a nail merely giving
an appearance of a "hoof". Similarly, the pig,
although it has cloven true hooves, does not
ruminate. Other animals such as dogs or cats
exhibit neither characteristic, and are thus
also forbidden for Jewish consumption."
https://en.wikipedia.org/wiki/Cloven_hoof

MERS was camel related? Right?

So this wouldn't be kosher:

The Food Ranger - Dubai Food
https://www.youtube.com/watch?v=1fVwoquZuU4

LoL

[Mostowski Collapse]

So did he (The Food Ranger) use the left
or right hand to grab the food? When I watch
these videos to pass time, I also always

check whether people (from restaurant,
kitchen, etc..) use gloves or not. One sick
person is quickly a super spreader.

On the other hand, when visiting a cheese
factory as a child, we were allowed to
put our bare hands into fresh cut curd,

to get an idea of the texture. Maybe they
thought some more bacteria will not harm?
Such risky behaviour maybe not anymore seen.

Well not really, again no gloves:

Cheesemaking - visiting a Swiss dairyman
https://www.youtube.com/watch?v=l9lwiOnMVVU

LoL

[Mostowski Collapse]

Well I guess this cheese is hand-made.

[Mostowski Collapse]

Now I got it. Numbers are updated around 00:15 GMT.
Just a few minutes ago it was 20700, now its 23900.

(Total Confirmed) Holy, Moly.

[Mostowski Collapse]

So whats the Elixir, to make Safe Cities, not
only Smart Cities with Tik Tok on the Phone?

240–280 nm: Disinfection, decontamination of
surfaces and water (DNA absorption has a peak at 260 nm)
https://en.wikipedia.org/wiki/Ultraviolet#Applications

Source:

Influenza as a bioweapon - J R Soc Med. 2003 Jul; 96(7): 345–346.
inactivation systems (e.g. ultraviolet) for ventilation systems
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC539539/

[Mostowski Collapse]

More details:

Inactivation of the coronavirus - J Virol Methods. 2004 Oct;121(1):85-91.
https://www.ncbi.nlm.nih.gov/pubmed/15350737

[j4n bur53]

Holy Cow

280 m from food market another lab?
Wuhan Center of Disease Control
https://twitter.com/OSINTHK/status/1228664201452765185

Not sure whether this is only a harmless
bureaocratic institution or not. Also
from the same twitter thread:

China Science Communication - Youth in the Wild
Some animal sample harvester footage
https://www.youtube.com/watch?v=ovnUyTRMERI

R Kym Horsell schrieb:
> Mostowski Collapse <burs...@gmail.com> wrote:
>> coronavirus coming from a P4 lab in wuhan
>> itself? If this is true, you China are
>> officially declared stupid.
>> I saw a TV special about some german lab where
>> they studied some pig viruses. Question is
>> how livestock handles some viruses, that
>> usually don't exist in the borders of EU.
>> The lab was on an island far away from
>> civilisation, and you were not allowed to
>> visit it as a reporter.
>> https://en.wikipedia.org/wiki/Friedrich_Loeffler_Institute
>
> Interesting. I hear they just discovered swine fever (ASF) in
> wild pigs in Poland nr the German border.
>

[Mostowski Collapse]

Gain-of-Function Research
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4996883/

GOFR is a subset of “dual-use research”
i.e., research that can be used for both
beneficial and malevolent purposes.

‘Dual-use research of concern’ (DURC)
refers to dual-use research for which the
consequences of malevolent use would be

exceptionally severe.

But how do you GOFR viruses? Being better
as nature. Currently a virus is nearly
as small as microchip circuit.

Virus diameter 20–1000 nm
MOSFET Masks UV Light 200 nm

So maybe nanon chips can be used to select
viruses? The UV Light would only destroy
the viruses, but it might be

nevertheless be used to produce something
that could maybe outperform nature in
mutation and selection?

[Mostowski Collapse]

Samsung has started EUV production it seems.
This needs vacum and special mirrors. And
you can go down to 10nm.

https://news.samsung.com/global/samsung-electronics-begins-mass-production-at-new-euv-manufacturing-line

https://en.wikipedia.org/wiki/Extreme_ultraviolet_lithography

[Mostowski Collapse]

Semantic Network!
https://www.wuhanvirus.sg/cases
You can grab nodes with the mouse.

Whats the meta-model? There is something
like airplances and hospitals involved.

Anybody up for a Prolog take?

[Mostowski Collapse]

Deep Learning to the rescue:

Zhang, H.; Saravanan, K. M.; Yang, Y.;
Hossain, T. Deep learning based drug
screening for novel coronavirus
2019-nCov. 2020, 19, 1–17.

GPU Accelerated matching:

Fischer,A.; Sellner,M.; Neranjan, S., Lill, M.A
and Smieško, M. Inhibitors for Novel Coronavirus
Protease Identified by Virtual Screening of
687 Million Compounds, 2020

https://chemrxiv.org/articles/Inhibitors_for_Novel_Coronavirus_Protease_Identified_by_Virtual_Screening_of_687_Million_Compounds/11923239

[Mostowski Collapse]

What nobody from the demented kitchen cabinets
around the world could do,

Greta Thunberg could do it. Now this here is
trending on Twitter:

#flattenthecurve

Cudos

[Mostowski Collapse]

Corona now killing a lot of stock bubbles.

[Mostowski Collapse]

Some damage ranking seen as of now:

1. Germany (-12.5%)
2. Switzerland (-9.5%)
3. USA (-8.5%)

Woa!

[Mostowski Collapse]

Not to mention the monday damage.

[Mostowski Collapse]

Irony, school skipping on friday isn't
anymore a thing. Schools anyway closed.

#ClimateStrikeOnline

Am Donnerstag, 12. März 2020 00:32:35 UTC+1 schrieb Mostowski Collapse:

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> Some damage ranking seen as of now:
>
> 1. Germany (-12.5%)
> 2. Switzerland (-9.5%)
> 3. USA (-8.5%)
>
> Woa!

OK. Here's something to think about.

Write a simple lil program that predicts something interesting about
the covid19 pandemic.

Here is an example (just numeric, not prolog yet :) of the kind of
thing that would be interesting, at least in a school classroom context:

When will e.g. Russia visus caces catch up to Italy?

Rationale: italy is in a pretty bad position. Large number of cases.
Very fast doubling time. Very poor testing regime so it doesnt really
know how many cases there are -- it's just reporting the cases it found.
(You can figure this out from the serial correlation of the relevant
time series).
Ergo if country X is increasing exponentially then it will eventually
catch up with where Italy is now. How many days will that take?

To belabour the point:


From ARMA(1,1) regression:
mod1=y = 1.11667e-06*exp(0.236644*x) Russia data (from Johns Hopkins database)
mod2=y = 0.00190806*exp(0.225086*x) Italy data
jday russia italy
76 72.2249 51270.3
77 91.5082 64212.5
78 115.94 80421.7
79 146.895 100723
80 186.114 126148
81 235.805 157992
82 298.762 197874
83 378.529 247823
84 479.592 310381
85 607.639 388731
86 769.872 486858
87 975.42 609756
88 1235.85 763677
89 1565.81 956452
90 1983.86 1.19789m
91 2513.53 1.50027m
92 3184.62 1.87899m
93 4034.88 2.3533m
94 5112.16 2.94735m
95 6477.05 3.69135m
96 8206.36 4.62315m
97 10397.4 5.79018m
98 13173.4 7.2518m
99 16690.5 9.08237m
100 21146.7 11.375m
101 26792.7 14.2464m
102 33946.1 17.8427m
103 43009.3 22.3467m
104 54492.4 27.9877m
jday 104 russia catches up with jday 76 italy in 28 days time

1 more days work:
Add English-like front end and allow questions like "when X", "will ever X",
or whatever other Q's you think are cute.

Think of it as the Chat80 but with virus pandemics.

[j4n bur53]

How do you fit a sigmoid curve? Spreading
is not purely exponential. There are not
indefinitely many people to infect.

R Kym Horsell schrieb:

[R Kym Horsell]

j4n bur53 <janb...@fastmail.fm> wrote:
>
> How do you fit a sigmoid curve? Spreading
> is not purely exponential. There are not
> indefinitely many people to infect.

In the first day of development dont worry about
more than 30% of the pop'n (i.e. the usual flu season)
having the disease.

The exact numerical modeling your logic program can use to
get the numbers can be tuned up later. As a first step I tend
to use the actual data from the field, maybe fixing observing
biases if that's at all possible. At least you can show people
and say it's based on numbers that have been actually seen,
not hypothetical models that work from first principles and
guessed parameters. :)

My leanings these days is trying to build very simple models
that boards and random policy makers can understand and have a
solid mental model of how they work and what their limitations are.

At the moment e.g. politicians have no real idea how "the experts" are
getting their projections. They ask a question and a panel
of experts appears to vote on various numbers and some kind of
consensus is reached and if you they extra special lucky that day
there may be error bars on the estimate.

It would be nice if AI's could be used to run "what if" scenarios
in realtime via an office k/b and answer questions from non-experts
that would allow them to creep up on the underlying subject matter
without blowing any major neural circuitry and hitting a
Kruger-Dunning black hole.

> R Kym Horsell schrieb:
>> Mostowski Collapse <burs...@gmail.com> wrote:
>>> Some damage ranking seen as of now:
>>> 1. Germany (-12.5%)
>>> 2. Switzerland (-9.5%)
>>> 3. USA (-8.5%)
>>> Woa!
>> OK. Here's something to think about.
>> Write a simple lil program that predicts something interesting about
>> the covid19 pandemic.
>> Here is an example (just numeric, not prolog yet :) of the kind of
>> thing that would be interesting, at least in a school classroom context:
>> When will e.g. Russia visus caces catch up to Italy?

...

--
[Never been sober for 2 days straight:]

[Kym Horsell 03 Aug 2017 0:1:2 PDT:]
> Put down da shine an sober up for a day or 2.

Neber!
-- AlleyCat Computers aka Hillbilly Davis, 03 Aug 2017

[R Kym Horsell]

R Kym Horsell <k...@kymhorsell.com> wrote:

...

> My leanings these days is trying to build very simple models
> that boards and random policy makers can understand and have a
> solid mental model of how they work and what their limitations are.
>
> At the moment e.g. politicians have no real idea how "the experts" are
> getting their projections. They ask a question and a panel
> of experts appears to vote on various numbers and some kind of
> consensus is reached and if you they extra special lucky that day
> there may be error bars on the estimate.
>
> It would be nice if AI's could be used to run "what if" scenarios
> in realtime via an office k/b and answer questions from non-experts
> that would allow them to creep up on the underlying subject matter
> without blowing any major neural circuitry and hitting a
> Kruger-Dunning black hole.

>...

No sooner asked than answered! (Sort of).

[Herd Immunity:]
The UK Only Realised "In The Last Few Days" That Its Coronavirus Strategy
Would "Likely Result In Hundreds of Thousands of Deaths"
BuzzFeed, 16 Mar 2020
Scientists advising the government say an aggressive new approach adopted to
attempt to "suppress" the virus may have to be in place for 18 months.

[The Johnson Plan:]
Robert Rohde @RARohde 15 Mar 2020 0:50Z
Apparently the UK believes the best way to protect vulnerable people
from coronavirus is to achieve herd immunity. My video, from last year,
explains herd immunity in the context of vaccines. Without a vaccine,
~60% of the UK would have to get sick to achieve the effect.
<https://t.co/flT8roMEAl>twitter.com/RARohde/status...

Adam Kucharski @AdamJKucharski 14 Mar 2020 13:41Z
I am deeply uncomfortable with the message that UK is actively pursuing
`herd immunity' as the main COVID-19 strategy. Our group's scenario
modelling has focused on reducing two main things: peak healthcare
demand and deaths... 1/

[Mostowski Collapse]

Brexit Boris wants to restore UK pension
funds etc.. by eliminating half of its
population through coronavirus. Cheers.

LoL

[Mostowski Collapse]

Looks like a pretty nice whole
body inflamation already:

https://www.theguardian.com/world/2020/mar/17/coronavirus-uk-cases-how-many-are-in-your-area

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> Looks like a pretty nice whole
> body inflamation already:
>
> https://www.theguardian.com/world/2020/mar/17/coronavirus-uk-cases-how-many-are-in-your-area
>

Health officials are softening up the public already, saying
20,000 deaths would be a "good outcome".
Down from the previous estimate of 100,000.

Exponential functions and tsunami waves are soo deceptive.
They look slow-moving and relatively harmless until it's on
top of you.

[Mostowski Collapse]

EMA was based in London until January last
year, when Brexit saw it relocate to Amsterdam.
https://www.theguardian.com/world/2020/mar/14/coronavirus-vaccine-delays-brexit-ema-expensive

[Mostowski Collapse]

Thats a nice Sigmoid:
https://www.worldometers.info/coronavirus/country/south-korea/

Some countries are far from it, like for
example Germany. Also a sigmoid is not
necessarily the end result, there can

be relapses, so the result could be
a sum of different sigmoids.

[Mostowski Collapse]

Currently in the accelerating branch of sigmoid,
number of cases doubling every 3 days:

Comparison of different countries
https://twitter.com/RobertRethfeld/status/1241088603834458115

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> Currently in the accelerating branch of sigmoid,
> number of cases doubling every 3 days:
>
> Comparison of different countries
> https://twitter.com/RobertRethfeld/status/1241088603834458115
>

...

The tsunami is about to hit NY.

Every day Johns Hopkins updates its time series and I scrape the data
and compare each country and region in there (methodology changes over time
just to keep things "interesting") and compares simple TS model
against situation in Italy.

NYS is presently 3-4 days away from caseloads equalling Italy.

UK is less than 2 wks from that point.

Even Australia with 3rd-largest testing regime is only 4 wks from Italy
levels. They're having trouble with making kids and country bumpkins
take it seriously. Despite empty supermarkets, closed borders and
all sport events cancelled or playing for cameras only people under
50 have trouble accepting something it wrong. Yesterday councils
around Sydney had to close down beaches because it was sunny and
1000 of people came down to the beaches and were refusing to
leave because they claimed it was all overblown and nothing was really wrong.

Interesting case is Russia. They have 100s of ack'ed cases now with
(I think) a 3-4 day doubling time. Many other hospital cases are
listed as "pneumonia of unknown origin". But even the officially-declared
numbers now show the same kinds of curves for other problem centers.

It's gunna get much worse before it gets better.

There will be some stings in the tail with early papers indicating
the virus can hide in the body and other details that point to
second and other waves of disease coming in several months time.
Most likely this will become an annual thing and pnemonia deaths rates
will essentially be double prev levels from now on.

Economics is the dismal science, they say. Deductive logic unerringly
leads to an apparently "sub optimal" solution most of the time.
That old "2 icecream sellers on the beach problem" that each
incrementally move to the center of the beach trying to maximize their
market. Their moves are "inevitable" and "pointless" in a sense.

Treating covid19 with incremental solutions ("tightening the knob"
on the radiation like the NY governor was saying on TV) when
you know a tsunami is coming is a bit the same. It's something
authorities and random beancounters have to go through before they
inevitably find the strategy is not good enough.

--
[HOCKEY SCHTICK!]
John Pratt @Jackthelad1947 19 Mar 2020 07:11Z
Hockey sticks #coronavirusaustralia and Global temperature are
teaching Australians about exponential growth. #ClimateEmergency
Demand #ClimateAction and socially isolate now. Your future
depends on what you do today. #auspol #qldpol #StopAdani
#TheDrum @TheCairnsPost pic.twitter.com/j0EDrp1P55

Aly-Khan Satchu @alykhansatchu 25 Feb 2020 02:25Z
"The greatest shortcoming of the human race is our inability to
understand the exponential function." - Professor Allen Bartlett
@macronomics1 <https://t.co/tR8YdkcL1D>twitter.com/macronomics1/s...
24-FEB-2020 :: The Viral Moment has Arrived #COVID19 #coronavirus
#2019nCoV <https://t.co/ZzNABgH9m4>j.mp/38UwVym

[Mostowski Collapse]

Well, the 50 year old are subject to misinformation.
RKI now publishes gender and age. Look for yourself:

https://experience.arcgis.com/experience/478220a4c454480e823b17327b2bf1d4

I made a screenshot:
https://gist.github.com/jburse/13b6f57c6e043a12befda681868f55ea#gistcomment-3223271

The 50 year old are to most affect, at least the
most infected. Lets say 2 weeks from now, we

will see whats the death rate.

[Mostowski Collapse]

Coronatest for $10 based on, under discussion:
https://en.wikipedia.org/wiki/Real-time_polymerase_chain_reaction

[Mostowski Collapse]

"First attempt at 50 min (from sample to result)
CoV test with a 3 min extraction and rapid
cycle RT-qPCR on a portable qPCR machine (16
samples per run) with readout on a mobile
phone. Needs optimisation but looking promising.
Collaboration with Jonathan Edgeworth at St Thomas'"
https://twitter.com/Justin_OGrady/status/1240577369439698944

[R Kym Horsell]

Mostowski Collapse <burs...@gmail.com> wrote:
> "First attempt at 50 min (from sample to result)
> CoV test with a 3 min extraction and rapid
> cycle RT-qPCR on a portable qPCR machine (16
> samples per run) with readout on a mobile
> phone. Needs optimisation but looking promising.
> Collaboration with Jonathan Edgeworth at St Thomas'"
> https://twitter.com/Justin_OGrady/status/1240577369439698944

The blood drop tests were always the quickest.
I have no idea on accuracy.
I did note that Spain had contracted to get tests from China
and their validation showed they were only "30% accurate".
Trouble never comes in 1s.

> Am Donnerstag, 26. M"arz 2020 19:41:19 UTC+1 schrieb Mostowski Collapse:
>> Coronatest for $10 based on, under discussion:
>> https://en.wikipedia.org/wiki/Real-time_polymerase_chain_reaction
>>

--
>[It can be] some day in summer being the same max temp as some day
>in winter and summer still being warmer than winter.
But that ONE day wasnt warmer was it. :)
-- Matt Sykes, 8 Oct 2013

[Mostowski Collapse]

"Spider" query by geospatial
data visualization engine:

Spring Break vs. COVID19
https://www.youtube.com/watch?v=cq2zuE3ISYU

[Mostowski Collapse]

Some news about Web Prolog?

Elixir has Nerves
Prolog has _____ ?

See also:
Interactive Robotics with Elixir and Nerves
https://www.youtube.com/watch?v=zBzKoV2iFdQ