On 19/11/2014 9:58 PM, Kirk Wolf wrote:
This is not a comment about the LUA4Z port, but at first glance I find the
module system a little confusing.
It's quite simple once you grok it
http://www.lua.org/manual/5.1/manual.html#5.3. It's very flexible. You
can add your own module system quite
I'm a bit late to the party, but I finally got around to installing Lua4z.
The installer and ivp worked fine. For a first release, I think that it
is extremely well done.
I also wanted to mention a presentation that David did in SHARE this summer
which I believe demonstrates some of the
In 5449b49c.3040...@gmail.com, on 10/24/2014
at 10:08 AM, David Crayford dcrayf...@gmail.com said:
I don't want to be too disparaging about REXX but I've profiled it
extensively and it is not an effecient implementation.
That depends on what it is. What did you profile? OS/2 CREXX? OS/2
On 28/10/2014 6:39 AM, John McKown wrote:
xlc only supports HFP and BFP natively via the float compiler option, so
I left it out.
The C/C++ compiler has builtins for both packed decimal and decimal
floating point. If you were feeling adventurous you could write bindings
for both. Lua
On Mon, Oct 27, 2014 at 8:12 PM, David Crayford dcrayf...@gmail.com wrote:
On 28/10/2014 4:20 AM, John McKown wrote:
This is likely a silly question. Does lua4z use HFP (zArch) or BFP (IEEE)
floating point? If I had to guess, I'd _guess_ HFP.
You guessed wrong :). Lua numbers are double
On Tue, Oct 28, 2014 at 6:53 AM, John McKown john.archie.mck...@gmail.com
wrote:
On Mon, Oct 27, 2014 at 8:12 PM, David Crayford dcrayf...@gmail.com
wrote:
On 28/10/2014 4:20 AM, John McKown wrote:
This is likely a silly question. Does lua4z use HFP (zArch) or BFP (IEEE)
floating point? If
On 28/10/2014 7:53 PM, John McKown wrote:
On Mon, Oct 27, 2014 at 8:12 PM, David Crayford dcrayf...@gmail.com wrote:
On 28/10/2014 4:20 AM, John McKown wrote:
This is likely a silly question. Does lua4z use HFP (zArch) or BFP (IEEE)
floating point? If I had to guess, I'd _guess_ HFP.
You
On Tue, Oct 28, 2014 at 7:32 AM, David Crayford dcrayf...@gmail.com wrote:
On 28/10/2014 7:53 PM, John McKown wrote:
On Mon, Oct 27, 2014 at 8:12 PM, David Crayford dcrayf...@gmail.com
wrote:
On 28/10/2014 4:20 AM, John McKown wrote:
This is likely a silly question. Does lua4z use HFP
On 28/10/2014 9:31 PM, John McKown wrote:
I've got a new build for Lua4z which includes sqlite3 as part of LuaSQL
https://github.com/keplerproject/luasql. I'm going to release it soon
with the regex extensions for POSIX and PCRE.
Wonderful. Thanks. I will now just wait. I'm__good__ at just
Installation error - likely due to your fix of upper casing input. Relevant
portion of install output:
Enter PATH directory in which to create command symlinks or 'none'
(/usr/local/bin)
/HM/bin
link_dir (/HM/BIN) is not a valid path.
realpath() failed: realpath: EDC5129I No such file or
Try again John. Should be good to go. And thanks for testing this for me!
On 28/10/2014 10:06 PM, John McKown wrote:
Installation error - likely due to your fix of upper casing input. Relevant
portion of install output:
Enter PATH directory in which to create command symlinks or 'none'
Nope. Transcript:
sh ./lua4z-1.0.0.bin
./lua4z-1.0.0.bin
The Lua4z software license agreement is available on the web at:
http://lua4z.com/license
Please read the license carefully.
The installed software includes a copy of the license
in the lua4z/doc directory.
Have you read, understood,
Oh, tested -- install works if I pre-create the LOADLIB and SAMPLIB.
--
The temperature of the aqueous content of an unremittingly ogled
culinary vessel will not achieve 100 degrees on the Celsius scale.
Maranatha!
John McKown
Works if I reply Y, in upper case. Problem?
if promptYN(The PDSE load module library '..loadlib..' does not exist,
create it?, y) then
should be?
if promptYN(The PDSE load module library '..loadlib..' does not exist,
create it?, y):upper() then
On Tue, Oct 28, 2014 at 9:47 AM, John McKown
Works GREAT! And it interoperates with the sqlite3 UNIX command from the
SQLite port for z/OS on the CBT. I used your example to create the SQLite
data base, then accessed it with the sqlite3 command from a shell prompt.
Curious: did you use the CBT version of SQLite or did you do your own?
On
On 2014-10-28 17:22, Ed Gould wrote:
Since we no longer have a monitor how about creating a list for this
product ?
http://lua4z.com/forum/
-- gil
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send
This is likely a silly question. Does lua4z use HFP (zArch) or BFP (IEEE)
floating point? If I had to guess, I'd _guess_ HFP.
On Thu, Oct 23, 2014 at 6:38 AM, David Crayford dcrayf...@gmail.com wrote:
Lua is a powerful, fast scripting language with vibrant user communities
on many platforms.
On Mon, 27 Oct 2014 15:20:39 -0500, John McKown wrote:
This is likely a silly question. Does lua4z use HFP (zArch) or BFP (IEEE)
floating point? If I had to guess, I'd _guess_ HFP.
I'd surely hope so. HFP is *so* 20th Century!
(You left one out.)
On Thu, Oct 23, 2014 at 6:38 AM, David
xlc only supports HFP and BFP natively via the float compiler option, so
I left it out.
On Oct 27, 2014 5:05 PM, Paul Gilmartin
000433f07816-dmarc-requ...@listserv.ua.edu wrote:
On Mon, 27 Oct 2014 15:20:39 -0500, John McKown wrote:
This is likely a silly question. Does lua4z use HFP
On 28/10/2014 4:20 AM, John McKown wrote:
This is likely a silly question. Does lua4z use HFP (zArch) or BFP (IEEE)
floating point? If I had to guess, I'd _guess_ HFP.
You guessed wrong :). Lua numbers are double precision IEEE BFP. This
did cause some problems. For example, the DB2 ODBC
On 24/10/2014 11:03 PM, Barkow, Eileen wrote:
LUA could use some good tutorials or at least some more samples of code.
The reference manual is no help for someone trying to learn the language.
There's a wiki that covers just about everything you could possibly want
to know about Lua which
On 24/10/2014 10:40 PM, John Gilmore wrote:
Like landing at Thanet, having Lua usable under z/OS UNIX is a good
thing. It may make its way against REXX, or again it may not.
It goes without saying that us mainframers tend to stick to what we
know. But I encourage as many people as possible
Am 25.10.2014 10:02, schrieb David Crayford:
On 24/10/2014 10:40 PM, John Gilmore wrote:
Like landing at Thanet, having Lua usable under z/OS UNIX is a good
thing. It may make its way against REXX, or again it may not.
It goes without saying that us mainframers tend to stick to what we
Thank you very much.
Of course, it is all about the quality of the implementation.
The implementation on a certain environment may be bad
or not very motivated ... but the language REXX should not
be blamed for this.
Some time ago (maybe 10 years) I tested REXX on OS/2 vs. Regina on Linux
on
FWIW, JavaScript runs on the mainframe in a JVM as well, which has the
advantage of being zAAP/zIIP eligible. I've done a fair bit of that. Some of it
to pretty good productive use.
Lua does look interesting and has the benefit of having an actual I/O library
included.
On Thu, 23 Oct 2014
-Original Message-
From: IBM Mainframe Discussion List On Behalf Of David Crayford
On 24/10/2014 6:50 AM, Bernd Oppolzer wrote:
Doesn't the example in the benchmark show performance problems in
EXECIO instead of REXX? EXECIO, IMO, is not part of the REXX
interpreter, but instead
I am trying to run a benchmark test as David suggested using the clists in
http://users.tpg.com.au/crayford/rexx-lua-c-io-benchmark.htm
the only big files I have to read (containing alot of data) are SMF archives.
LUA4Z reads them without a problem but
REXX/EXECIO gives this error:
Like landing at Thanet, having Lua usable under z/OS UNIX is a good
thing. It may make its way against REXX, or again it may not.
The matrix-algebra test case is not, however, an appropriate one for
the comparative evaluation of these two implementations of Lua and
REXX. It is not indeed a good
LUA could use some good tutorials or at least some more samples of code.
The reference manual is no help for someone trying to learn the language.
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf
Of John Gilmore
Sent: Friday, October 24,
On Fri, Oct 24, 2014 at 9:22 AM, Barkow, Eileen ebar...@doitt.nyc.gov
wrote:
snip
the only big files I have to read (containing alot of data) are SMF
archives.
LUA4Z reads them without a problem but
snip
So LUA4Z is probably better for this reason alone – it can read any file.
PMFJI here, but I just wanted to add one small issue I have with the lua
language - The fact that it does not use POSIX extended regular expression
syntax for string patterns. As a long-time [g]awk user, I am very experienced
and comfortable with POSIX RE syntax, but I am totally unfamiliar
On 24/10/2014 11:03 PM, Barkow, Eileen wrote:
LUA could use some good tutorials or at least some more samples of code.
The reference manual is no help for someone trying to learn the language.
The canonical programming guide is PIL (Programming In Lua)
http://www.lua.org/pil/contents.html.
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On
Behalf Of David Crayford
Sent: Thursday, October 23, 2014 10:08 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: [ANN] Lua4z: the Lua programming language on z/OS, with batteries
snip
OK, let's try
I keep getting this error both in MVS and UNIX.
LUA: //'xcics.lua4z.samplib(cpu)':4: attempt to perform arithmetic on field '?
' (a nil value)
stack traceback:
I would guess that, as previously suggested,
for i = 1, count do a[i] = a[i] * i end
would really rather be
for i = 1, count do a[i] = i * i end
Bob
Barkow, Eileen wrote:
I keep getting this error both in MVS and UNIX.
LUA: //'xcics.lua4z.samplib(cpu)':4: attempt to perform arithmetic on
This suggestion is probably unrelated to the actual issue, but I see the
odd characters ÝC¨ in your note, which I think should display as square
brackets. If your terminal emulator is currently set to use code page
37 (or similar), you might try code page 1047.
Barkow, Eileen wrote:
I keep
Thank you -
for i = 1, count do a[i] = i * i end fixed the problem.
I get an elapsed time of about 0.14 - 0.15 running LUA on both UNIX and MVSs,
But the REXX clist yields about 3.3
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf
Of
I had to change the brackets to x'AD' X'BD', the same as I use for java.
The for i = 1, count do a[i] = i * i end statement fixed the problem.
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf
Of Tom Brennan
Sent: Friday, October 24, 2014
On 25/10/2014 12:01 AM, Farley, Peter x23353 wrote:
PMFJI here, but I just wanted to add one small issue I have with the lua language -
The fact that it does not use POSIX extended regular expression syntax for string
patterns. As a long-time [g]awk user, I am very experienced and
Lua4z was built using the IBM-1047 code page but the lexer and parser
have been patched to tolerate IBM-037
http://lua4z.com/doc/manual/using.md.html#Square_brackets.
Unfortunately, output is printed in IBM-1047 which is why the square
brackets are garbled. I could fix this but you need to
On 25/10/2014 3:35 AM, Bob Rutledge wrote:
I would guess that, as previously suggested,
for i = 1, count do a[i] = a[i] * i end
would really rather be
for i = 1, count do a[i] = i * i end
Good eye Bob! When I copy and pasted the snippet into the e-mail it
dropped the square brackets.
Might I suggest changing the variable name from A to SQ?
Then you have the calculation of SQ[i] = i * i.
On Fri, Oct 24, 2014 at 9:32 PM, David Crayford dcrayf...@gmail.com wrote:
On 25/10/2014 3:35 AM, Bob Rutledge wrote:
I would guess that, as previously suggested,
for i = 1, count do a[i]
Lua is a powerful, fast scripting language with vibrant user communities
on many platforms. Lua celebrates its 20th birthday this year.
From the Lua website (http://www.lua.org/about.html):
Lua has a deserved reputation for performance. To claim to be as
fast as Lua is an aspiration of
Thank you. I guess this means that I need to learn lua now grin/.
On Thu, Oct 23, 2014 at 6:38 AM, David Crayford dcrayf...@gmail.com wrote:
Lua is a powerful, fast scripting language with vibrant user communities
on many platforms. Lua celebrates its 20th birthday this year.
From the Lua
On 23/10/2014 7:46 PM, John McKown wrote:
Thank you. I guess this means that I need to learn lua now grin/.
That shouldn't be too difficult for a man of your abilities 8^)
Scripting languages are generally very easy to learn. Lua is has an
almost identical feature set to Javascript. They
On Thu, Oct 23, 2014 at 7:00 AM, David Crayford dcrayf...@gmail.com wrote:
On 23/10/2014 7:46 PM, John McKown wrote:
Thank you. I guess this means that I need to learn lua now grin/.
That shouldn't be too difficult for a man of your abilities 8^)
Scripting languages are generally very
On 23/10/2014 8:42 PM, John McKown wrote:
Well, I got it installed. I ran into a problem with the
./lua4z-1.0.0.bin when I ran it a second time. The first time, it said it
couldn't find the PDSEs (LOADLIB and SAMPLIB) and did I want to define
them? All is good. I reply Y and they are
You need to enter the y in lower case - otherwise those dataset errors occur.
I am not up to snuff on UNIX lately and do not understand what the ? in the
environment variables mean.
They did work when I added them to my profile, but I never saw this syntax
before:
export LUA_PATH=./?.lua;\
On 23/10/2014 9:36 PM, Barkow, Eileen wrote:
You need to enter the y in lower case - otherwise those dataset errors occur.
I have now changed the installation script to fold all input to upper case.
I am not up to snuff on UNIX lately and do not understand what the ? in the
environment
On Thu, Oct 23, 2014 at 8:24 AM, David Crayford dcrayf...@gmail.com wrote:
Can I convince you to try it again. I've uploaded an updated IVP.
Well the install script does not give any error messages when I reuse the
same PDSE library names. But the ivp still fails. I can't shake the feeling
Thanks for the explanation about the ?s but I still do not see what they are
being used to substitute for.
I passed the UNIX IVP test, but my profile contains this:
$ echo $LUA_PATH
I'm at a loss! Can you do this quick test for me.
Invoke the Lua shell by typing lua in the shell
type
require cjson
post the stack trace
On 23/10/2014 9:51 PM, John McKown wrote:
On Thu, Oct 23, 2014 at 8:24 AM, David Crayford dcrayf...@gmail.com wrote:
Can I convince you to try it
I was certain that I did something wrong, but I don't know what. Somewhere
I have missed a step.
require cjson
error loading module 'cjson' from file
'/HealthMarkets/lua4z/lib/lua/5.1/cjson.so':
dynamic libraries not enabled; check your Lua installation
stack traceback:
[C]: in
On 23/10/2014 10:02 PM, Barkow, Eileen wrote:
Thanks for the explanation about the ?s but I still do not see what they are
being used to substitute for.
I passed the UNIX IVP test, but my profile contains this:
Lua has a module system that searches LUA_PATH substituting ? for the
module
Here is another thing I never saw before.
I do not have a PDSE in linklist (that I can use) so I am trying to copy the
LUA4Z.LOADLIB modules into a regular pds and get this error.
I do not want to have another linklist lib added for this product and was
always able to copy PDSE load modules
Thanks again for the explanation David - I am learning a lot installing this
product.
See my other posting about the PDSE to PDS issue.
-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf
Of David Crayford
Sent: Thursday, October 23, 2014
Unfortunately that's not possible. Lua4z is a C/C++ XPLINK application
and it requires a PO library so a PDSE is mandatory.
I understand your position but XPLINK is the build and it results in
much faster code. I know you probably already know this but you can
always dynamically add a PDSE to
There seems to be a problem loading the SO from a symlink. Can you try
it without a symlink? It works on my system but if I can narrow it down
to that and recreate it at our lab I can cut a fix.
On 23/10/2014 10:18 PM, John McKown wrote:
I was certain that I did something wrong, but I don't
No joy:
LIH1:TSH009:/HealthMarkets/lib$
rm liblua.so
LIH1:TSH009:/HealthMarkets/lib$
lua
Lua 5.2.0 Copyright (C) 1994-2011 Lua.org, PUC-Rio
require cjson
error loading module 'cjson' from file
'/HealthMarkets/lua4z/lib/lua/5.1/cjson.so':
dynamic libraries not enabled; check your
You are not running the correct version of Lua. You have another version
of Lua in your PATH which is why Lua 5.2.0 is being printed in the prompt.
That's a relief!
On 23/10/2014 10:53 PM, John McKown wrote:
No joy:
LIH1:TSH009:/HealthMarkets/lib$
rm liblua.so
Ah! yes. I have an alpha, which I got from you?, in /HealthMarkets/bin. So
the symlink commands (ln -s) for the new version into
/HealthMarkets/lua4z/bin silently failed. I removed that old version,
established the correct symlinks, and all is sunny and bright. This is not
likely to happen to any
Thanks again for the info. I will try to do without linklist for now.
I tried to set up the LUACONF member as documented and get this error:
And the parm //'SYS1.SAMPLIB(IVP)' is not accepted by the system.
FREE FI(LUA)
Ignore this - the DD should be LUACONF, not LUA
-Original Message-
From: Barkow, Eileen
Sent: Thursday, October 23, 2014 11:33 AM
To: 'IBM Mainframe Discussion List'
Subject: RE: [ANN] Lua4z: the Lua programming language on z/OS, with batteries
Thanks again for the info. I will try to
LOL, good catch! Better try it with UNUM. I would suggest defining LUA
source PDS data sets VB LRECL(256) BLKSIZE(32760). Lua reads config
files as if they were UNIX files as a stream, not record oriented card
image data sets. If that is not
in the doc I will make sure it is.
Thanks for the
On 23/10/2014 11:35 PM, Barkow, Eileen wrote:
Ignore this - the DD should be LUACONF, not LUA
Nope! The DD should be Lua and the member name should be LUACONF.
-Original Message-
From: Barkow, Eileen
Sent: Thursday, October 23, 2014 11:33 AM
To: 'IBM Mainframe Discussion List'
On 23/10/2014 11:35 PM, Barkow, Eileen wrote:
Ignore this - the DD should be LUACONF, not LUA
Nope! The DD should be LUA and the member name should be LUACONF.
-Original Message-
From: Barkow, Eileen
Sent: Thursday, October 23, 2014 11:33 AM
To: 'IBM Mainframe Discussion List'
On 23/10/2014 11:32 PM, John McKown wrote:
Ah! yes. I have an alpha, which I got from you?, in /HealthMarkets/bin. So
the symlink commands (ln -s) for the new version into
/HealthMarkets/lua4z/bin silently failed.
What I should do is test if an existing version of Lua exists and promt
to take
I got passed the LUA DD but cannot read the SAMPLIB
FREE FI(LUA)
IKJ56247I FILE LUA NOT FREED, IS NOT ALLOCATED
ALLOC FI(LUA) DA('XCICS.LUA4Z.CONF') SHR
CALL
On 23/10/2014 11:59 PM, Barkow, Eileen wrote:
I got passed the LUA DD but cannot read the SAMPLIB
FREE FI(LUA)
IKJ56247I FILE LUA NOT FREED, IS NOT ALLOCATED
ALLOC FI(LUA) DA('XCICS.LUA4Z.CONF') SHR
CALL 'XCICS.LUA4Z.LOADLIB(LUA)' 'XCICS.LUA4Z.SAMPLIB(IVP)'
lua: cannot open
On 23 October 2014 07:38, David Crayford dcrayf...@gmail.com wrote:
Lua4z is available at no charge. Download it now from:
http://lua4z.com
Sigh - just what the world needs - another free as in free beer free
software licence. As proprietary licences go, this one is perhaps less
bad than many,
On 24/10/2014 12:36 AM, Tony Harminc wrote:
On 23 October 2014 07:38, David Crayford dcrayf...@gmail.com wrote:
Lua4z is available at no charge. Download it now from:
http://lua4z.com
Sigh - just what the world needs - another free as in free beer free
software licence. As proprietary
Why not just release it under the MIT license like the main lua code from
www.lua.org is released under? Unless, of course, the company does not want
to release the source. That is certainly allowable under the original's MIT
license. And is one reason that the MIT license exists and is used. I.e.
I am waiting for the LUA4Z.LOADLIB to be added to linklist.
I do not have access to the ISPF command tables.
HOWEVER,
This job worked in batch.
I put the LUACONF member in LUA4Z.SAMPLIB and ran this:
//LUA EXEC PGM=LUA,PARM=IVP
//STEPLIB DD DSN=XCICS.LUA4Z.LOADLIB,DISP=SHR
All good points yet again. The reality is that I work for a vendor and
they sponsored the project with my wages. That's not a lot of money :)
but they are used to a world in where IP = source code and
anything different is alien to them. I'm very open to suggestions on
this topic because you
Also ran HELLO
So it appears that the LUACONF member just has to be in the same dataset as the
script to be run, not a separate CONF dataset.
//LUA EXEC PGM=LUA,PARM=HELLO
//STEPLIB DD DSN=XCICS.LUA4Z.LOADLIB,DISP=SHR
//*UA DD DSN=XCICS.LUA4Z.CONF,DISP=SHR
//LUA
On 24/10/2014 1:30 AM, Barkow, Eileen wrote:
Also ran HELLO
So it appears that the LUACONF member just has to be in the same dataset as the
script to be run, not a separate CONF dataset.
That's correct. Lua was born as a configuration language so we chose to
use a LUACONF member as part of
On 24/10/2014 1:25 AM, Barkow, Eileen wrote:
I am waiting for the LUA4Z.LOADLIB to be added to linklist.
I do not have access to the ISPF command tables.
If you site is configured to do so you can copy the sample command table
into a USRxCMDS member in your ISPTLIB. See the ISPF docs for
On Thu, Oct 23, 2014 at 12:29 PM, David Crayford dcrayf...@gmail.com
wrote:
All good points yet again. The reality is that I work for a vendor and
they sponsored the project with my wages. That's not a lot of money :) but
they are used to a world in where IP = source code and
anything
I have not checked the rest of the doc but the installation doc is not clear at
all about running on MVS.
For one thing, it does show any actual batch jcl example.
The doc indicates that the LUACONF member should be in a separate dataset but
does not state how to specify it in
relation to the
Point taken. I have forwarded your mail to the ID person and I'm sure he
will update it pretty quickly.
Now you have Lua installed why don't you bench-test it against REXX and
report if you get similar results to what I get on my machine
Good one Dave, glad you were able to convince your powers that be to get this
out the door.
Let's hope it gets some acceptance.
Now you have Lua installed why don't you bench-test it against REXX and
report if you get similar results to what I get on my machine
Now don't you go training
Doesn't the example in the benchmark show performance problems in EXECIO
instead of REXX? EXECIO, IMO, is not part of the REXX interpreter, but
instead
it is a vehicle (an external function) to do I/O from REXX on z/OS.
Other REXX implementations
(for example, on Linux or Windows) don't use
On 24/10/2014 6:50 AM, Bernd Oppolzer wrote:
Doesn't the example in the benchmark show performance problems in EXECIO
instead of REXX? EXECIO, IMO, is not part of the REXX interpreter, but
instead
it is a vehicle (an external function) to do I/O from REXX on z/OS.
Other REXX implementations
In 54490f13.4040...@gmail.com, on 10/23/2014
at 10:22 PM, David Crayford dcrayf...@gmail.com said:
It's one of the big pluses vs REXX
More specifically, the classic REXX included in z/OS and z/VM. Is
there an open requirement to incorporate OOREXX?
--
Shmuel (Seymour J.) Metz, SysProg
I am with Bernd on the Rexx I/O has been a bottleneck, but logically speaking
I/O can be a bottleneck being slow than memory functions. My Rexx experience is
with VM, Netview, TSO, z/OS. Also Orexx on Linux and Windows. So I speak from
experience as I am sure Bernd does.
Scott ford
85 matches
Mail list logo