Re: [GTALUG] JSON Standard for SQL Thoughts

[Christopher Browne via talk]

On Fri, 5 Jun 2020 at 17:23, Nicholas Krause via talk 
wrote:

> Greetings,
>
> I'm aware a few people here keep up more with certain database trends
> then me including Chris Browne. I'm aware of the newest standard of
> SQL supporting JSON.  Does anyone have any thoughts on this or how
> it may affect things going forward in the database world.
>
> Postresql supported operators before but seems their moving to the
> current SQL standard implementation, last I checked.


The PostgreSQL JSON types had been getting some pretty positive
attention, and with what got drawn into JSONB around the 9.5 days
https://www.postgresql.org/docs/9.5/functions-json.html
made it pretty attractive in ways that XML never got to.

JSONB was interesting in that it compresses material and provides
indexing inside the JSON document, so you can have functional
indexes on parts of the content, as well as GIN indexes

Sensible use of it shouldn't make all the usual relational bits go
away; tis best to define a table that has JSON with some mix
of JSON and other stuff.

ntdb=# create table our_users( user_id serial primary key);
CREATE TABLE
ntdb=#
ntdb=# create table message_encoded_json (
ntdb(#   id serial primary key,
ntdb(#   user_id integer not null references our_users(user_id) on delete
restrict,
ntdb(#   created_on timestamptz default now(),
ntdb(#   json_blob jsonb
ntdb(# );
CREATE TABLE
ntdb=# create index msg_gin on message_encoded_json using gin (json_blob
jsonb_ops);
CREATE INDEX
ntdb=# create index msg_gin2 on message_encoded_json using gin (json_blob
jsonb_path_ops);
CREATE INDEX

I'm not at all sure which of those indexes are preferable; that hasn't made
it into my
knowledge.

The above is the sort of pattern that makes sense; if there's a wild bundle
of stuff that
might go into a table of billing transactions, you'd still have ordinary
traditional attributes
like a transaction ID, transaction types, and dates in the lifecycle that
are pretty
universal.

I think I was in one of Oleg Bartunov's talks on their own "jsonpath" query
language,
possibly this one https://www.pgcon.org/2019/schedule/events/1274.en.html

The query system for JSON bits seemed pretty brain-melty; remembering what
little sets of characters going together searched for whatever seemed
painfully
similar to, oh, say, APL.

https://github.com/obartunov/sqljsondoc/blob/master/README.jsonpath.md
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] SpaceX Sent NASA Astronauts Into Orbit Using Linux

[John Moniz via talk]

Good to know. I've been wondering what the OS is in Crew Dragon and suspected it was Linux.  Elon Musk's Tesla cars also use Linux. On Jun 5, 2020 2:57 PM, James Knott via talk  wrote:https://futurism.com/the-byte/spacex-nasa-astronauts-linux
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] JSON Standard for SQL Thoughts

[Nicholas Krause via talk]

On 6/5/20 6:25 PM, Dave Cramer wrote:



On Fri, 5 Jun 2020 at 17:23, Nicholas Krause via talk > wrote:


Greetings,

I'm aware a few people here keep up more with certain database trends
then me including Chris Browne. I'm aware of the newest standard of
SQL supporting JSON.  Does anyone have any thoughts on this or how
it may affect things going forward in the database world.

Postresql supported operators before but seems their moving to the
current SQL standard implementation, last I checked.


I think the situation is the reverse. The standard is based on most of 
the early work from PostgreSQL



Dave Cramer


Dave,

Last I checked it wasn't, but if it that's cool. It would be nice for 
the  standard to use a common set of operators or data types. Most of 
the time the SQL standard seems to be pretty good about it, the only 
real black mark is date types.


Thanks,
Nick



--
Fundamentally an organism has conscious mental states if and only if 
there is something that it is like to be that organism--something it is 
like for the organism. - Thomas Nagel

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] JSON Standard for SQL Thoughts

[Dave Cramer via talk]

On Fri, 5 Jun 2020 at 17:23, Nicholas Krause via talk 
wrote:

> Greetings,
>
> I'm aware a few people here keep up more with certain database trends
> then me including Chris Browne. I'm aware of the newest standard of
> SQL supporting JSON.  Does anyone have any thoughts on this or how
> it may affect things going forward in the database world.
>
> Postresql supported operators before but seems their moving to the
> current SQL standard implementation, last I checked.
>

I think the situation is the reverse. The standard is based on most of the
early work from PostgreSQL


Dave Cramer
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] JSON Standard for SQL Thoughts

[Nicholas Krause via talk]

Greetings,

I'm aware a few people here keep up more with certain database trends
then me including Chris Browne. I'm aware of the newest standard of
SQL supporting JSON.  Does anyone have any thoughts on this or how
it may affect things going forward in the database world.

Postresql supported operators before but seems their moving to the
current SQL standard implementation, last I checked.

Thanks,
Nick
--
Fundamentally an organism has conscious mental states if and only if 
there is something that it is like to be that organism--something it is 
like for the organism. - Thomas Nagel

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] SpaceX Sent NASA Astronauts Into Orbit Using Linux

[Slackrat via talk]

James Knott via talk  writes:

> https://futurism.com/the-byte/spacex-nasa-astronauts-linux

Well Bully for Elon Musk. That's a truly momentous achievement.

But if he can repeat the feat on a reasonably frequent basis, RosCosmos
will really be crying because they made untold billions the last 9
years or so ferrying US flyers back and forth, despite the hullabaloo in
the Press about the latest NASA/Roscosmos agreement written to give the
impression that such hitch-hiking was a recent phenomenon.

-- 
William Henderson
aka Slackrat
http://billh.sdf.org/slackware.jpg
9HS5203 ON HamSphere Ham Radio
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

[GTALUG] SpaceX Sent NASA Astronauts Into Orbit Using Linux

[James Knott via talk]

https://futurism.com/the-byte/spacex-nasa-astronauts-linux
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Java

[Giles Orr via talk]

On Fri, 5 Jun 2020 at 13:04, Slackrat via talk  wrote:
>
> Can anyone give me a pointer to what precisely I might do to correct the
> situation. The first three suggestions do not apply:
> #   Reduce memory load on the system
> #   Increase physical memory or swap space
> #   Check if swap backing store is full
>
>
> The situation requires a cold boot to correct and that ABSOLUTELY
> nothing be done in the user account except launching an X server
> prior to launching the application.
>
>
> [QUOTE]
>
> bash-4.3# ./HamSphere_4.010a
> Fatal error: out of memory
> Uncaught exception: Out of memory
>
> An earlier version of the application at least privides a logfile when
> it crashes
>
> Server:www.hamsphere.com:8010
> OpenJDK 64-Bit Server VM warning: INFO:
> os::commit_memory(0x7f942b028000, 262144, 0) failed; error='Cannot
> allocate memory' (errno=12)
>
> #
> # There is insufficient memory for the Java Runtime Environment to continue.
> # Native memory allocation (mmap) failed to map 262144 bytes for committing 
> reserved memory.
> # An error report file with more information is saved as:
> # /inconnu/hs_err_pid23337.log
> [/QUOTE]
>
> bash-4.3$ free
>   totalusedfree  shared  buff/cache   
> available
> Mem:5999704  217532 50332322760  748940 
> 5392888
> Swap:   7219376   0 7219376
> bash-4.3$
>
> LogFile - too long to be reasonably opened by emacs, but "cat
> /hs_err_pid23337.log |less" works fine but, at least to me, is not
> helpful.
>
>
> System is Slackware 14.2 on an
> "Ideapad 330 (17, Intel) | Durable, Easy-to-Use 17.3" (sic) ... Lenovo" 
> disaster.

If you'll forgive me fumbling about in an area I barely know anything
about ...  I'll point out that Java sometimes means very different
things by "out of memory" than we think it does, and in this case
we're only finding out about it after it's filtered through the
application logging.  I think this probably means "Java has run out of
the memory you said it could have."  You didn't know you told it about
how much memory it could have, did you?  That's because there are
default values, and it's a pretty common problem to bump up against
them.  (Again, not totally sure that's what's happening here.)

I would try:

export JAVA_TOOL_OPTIONS="-Xmx1024M"

(You should perhaps check if this is already set FIRST, but doing this
at the command line is specific to the current shell.)  Then try
running the application again.

And read this:

https://stackoverflow.com/questions/5374455/what-does-java-option-xmx-stand-for

for a passable explanation of a horrible subject.  Read not just the
accepted answer, but the other stuff as well.

I hope this is what you need.

-- 
Giles
https://www.gilesorr.com/
giles...@gmail.com
---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Continuing Printer Woes

[D. Hugh Redelmeier via talk]

| From: Stewart C. Russell via talk 

| But … they're inside your router's firewall? Sure, they dial out for updates
| sometimes, but what doesn't?

Depending on a single firewall isn't considered good form.
Each machine should be hardened too.  Otherwise your network is
crunchy on the outside and tender on the inside.

This matters less in simple home networks.  But it is still a good idea if 
some of your other devices have security you cannot trust (eg. smart 
phones with apps that have some chance of being evil).  Windows 10 and 
most Linux distros have some hardening on each node.  I don't know or 
trust printers.

I don't remember the details, but we know that an HP Printer in
Saddam's Iraq was secretly working for the NSA.---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk

Re: [GTALUG] Continuing Printer Woes

[Stewart C. Russell via talk]

On 2020-06-04 10:51 a.m., D. Hugh Redelmeier via talk wrote:


I don't like putting printers on my network.  …   They don't have a
firewall that I can examine and configure.


But … they're inside your router's firewall? Sure, they dial out for 
updates sometimes, but what doesn't?


The one disadvantage of AirPrint - if you want to control access - is 
that anyone on your network can print at any time. So if you're lining 
up a special job on special paper, you better let other folks know first.


 Stewart

---
Post to this mailing list talk@gtalug.org
Unsubscribe from this mailing list https://gtalug.org/mailman/listinfo/talk