More prolog vs LAMP vs MEAN stack vs JAVA CRAP vs rails

[endlessboo...@gmail.com]

So lets assume a few 24 cpu 250g ram servers with 12 disk sas raid 10

running freebsd 10.3 and the raid 10 is ZFS mirrors

swi prolog vs the other websters

can you win?

can you do it with less code??

can this be real?


can you also run reports for the business?

[Andrew Davison]

I run a small web site with Prolog. Not that Prolog adds anything really, I just use as a powerful scripting language.

[endlessboo...@gmail.com]

did you use annie's web tutorial?

dynamic content?

do you use prolog as data store? and flush to disk periodically?

hows performance

[j4n bur53]

Since our recent release 1.1.4 had some changes in
the programming API and we updated all our deployment
examples, here it comes:

<shameless-selfpromotion>

Greatly simplified web and mobile Java coding for Prolog.

We updated our deployment guide to conform to our new programming API
for Jekejeke Prolog. In this new programming API the application
programmer can directly use strings, there is no need to wrap them in a
class TermAtom. But the new programming API for Jekejeke Prolog has also
further goodies. To sum up the following changes are available:

Interactor Interface:
The new programming API goes with Java iterator inspired methods.
The class Knowledge base has a method iterable() which will return an
Interpreter. The class Interpreter has a method iterator() which will
return a CallIn. The new programming API now also allows multiple
iterators at the same time without spawning a thread.

Class Path Loading:
The preferred way to load a Prolog text from class path has
changed. We abandoned the path/1 constructor and now go exclusively with
the library/1 constructor when a Prolog text should be loaded from the
class path. When using the library/1 constructor it is not necessary to
mention the extension (.p, .pl or .pro).

Lightweight Atomics:
We wastely simplified the code when using atomics in the
programming API. Prolog atoms/functors, Prolog numbers and Prolog
references can now be directly passed and returned by Java strings, Java
numbers and Java classes. So creating a compound =< becomes a one liner
TermCompound("=<", Integer.valueOf(agefrom), vars[COLUMN_AGE]).

When developing mobile applications for Android with Jekejeke Prolog we
recommend using the HAXM emulator. On the windows platform this emulator
has the fastest start-up time and also the most fluid execution. We even
have the impression that the HAXM emulator is faster that many mobile
devices and more or less on par with the Java JVM. For the HAXM emulator
VT has to be enabled in Bios.

Last but not least we have updated the example android manifest XML. The
changes in the android manifest and the use of the lasted Android SDK to
compile the mobile example results in a more modern look & feel when we
run the application. The mobile application is a little bit unorthodox
since we create all widgets on the fly. Nevertheless the basic styling
of the widgets depends on android manifest and SDK.

Intel® Hardware Accelerated Execution Manager
https://software.intel.com/en-us/android/articles/intel-hardware-accelerated-execution-manager

Jekejeke Prolog Mobile Deployment Example:
http://www.jekejeke.ch/idatab/doclet/prod/en/docs/05_run/15_stdy/08_deploy/10_mobile/package.html

</shameless-selfpromotion>

endlessboo...@gmail.com schrieb:

[endlessboo...@gmail.com]

oracle is cancer
I would never ever use java.
ever ever!!!
www.happstack.com
swi prolog webserver http://www.pathwayslms.com/swipltuts/html/
lisp coreserver or restas or ningle or lisp on lines
werc.cat-v.org

[j4n bur53]

Android doesn't use the oracle Java, after project
Harmony some years ago there was a branching.

You need to call Google cancer, and not Oracle.

But I admit its inverwoven, except if you directly
develop on an android device, for example with AIDE:

http://play.google.com/store/apps/details?id=com.aide.ui

endlessboo...@gmail.com schrieb:

[j4n bur53]

> You need to call Google cancer, and not Oracle.

Oh you cant do that, since Rob Picke together with
Robert Griesemer develop Go at Google.

Good move Google!

j4n bur53 schrieb:

[j4n bur53]

This is it:
http://research.google.com/pubs/r.html

So I hope you don't get shizo about these facts:
- Android = Google and Java = bad
- Go = Google and Unix = good
---------------------------------------
= Division by zero

j4n bur53 schrieb:

[endlessboo...@gmail.com]

NO ORACLE IS CANCER, google are bad as well

[endlessboo...@gmail.com]

www.gnoga.com

www.happstack.com

www.cliki.net

picolisp

yaron brook capitalism is moral

molyneux fossil fuels are moral alex epstein

[endlessboo...@gmail.com]

https://www.youtube.com/watch?v=44E6ujRLchs

racket scheme
swi prolog
gforth

[polymorph self]

google is cancer, they supported obama, bing.com all the way or duckduckgo.com

[polymorph self]

android is better than apple but still shitty
I dont really need a smartphone.....and I am investigating the ubuntu phone
:)
I use a desktop
laptops are for the weak minded

[polymorph self]

rob pike is awesome, and he just works at that shithole for the cash, he is on record saying he misses plan 9 and its dev environment

[polymorph self]

apple google twitter facebook oracle cancer

openbsd on a desktop with icewm good! :)

ubuntu phone or better yet don't use a smartphone.....!! its just a government control tool and surveillance

:)

[polymorph self]

remember capitalism is best, and great depression was caused by FDR and government

ayn rand has always been right
https://www.youtube.com/watch?v=lBY2yESW-Xs&index=9&list=PL34C495F517639744

[burs...@gmail.com]

> he is on record saying he misses plan 9 and its dev environment

Plus the idea all we need is a kernel, a shell and utilities
https://www.youtube.com/watch?v=JoVQTPbD6UY

Where is Prolog here?
The top level as a shell?
A lot of little modules as utilities?
Is SWISH the same thing only remote?
Is Prolog stagnant since 50 years?

[burs...@gmail.com]

The invention of the data scientist back then, before python:
https://www.youtube.com/watch?v=tc4ROCJYbm0

P.S.: The female programmer is not Margaret Hamilton(1) but
Lorinda Cherry(2). Maybe the template for Sheldon Cooper's
girlfriend in the TV series TBBT.

(1) https://de.wikipedia.org/wiki/Margaret_Hamilton_%28Wissenschaftlerin%29
(2) https://en.wikipedia.org/wiki/Lorinda_Cherry

P.P.S.: I would say Albert Hopkins(3) could have been
the template for Sheldon Cooper himself:
https://www.youtube.com/watch?v=YIBhPsyYCiM

(3) https://en.wikipedia.org/wiki/Albert_L._Hopkins

[burs...@gmail.com]

Disclaimer: The above post is not safe for work (NSFW),
you would need to watch these videos and look for these
people, and your boss might not be happy that you spend
so much time watching videos.

[polymorph self]

are you proposing a prolog os?

will it be smaller than linux?

less code?

[polymorph self]

in theory it seems as you add more categories
searches get faster
as the search disqualifies at least half for each new true flase category
like say I was looking for a date
under 20 miles away
height
race
political party
likes horror movies
are all categories which would make a search way faster, at least 50% for each category as it narrows down the pool

it would seem then you jsut need few 100 category to make an search blinding fast

[kint...@gmail.com]

> at least 50% for each category as it narrows down the pool
>
> it would seem then you jsut need few 100 category to make an search blinding fast

the last search in your list might be "blindingly fast" but that does not obliviate consideration of the previous searches , your first category search was blindingly slow , and that time still contributes to the overall time .

~~~ kintalken

[polymorph self]

no no
1 search
very fast
but since prolog can break down the data set into 100 shards, it will be fast

no? instead of panning the whole blob of data