Request for a new option: -savesuccess *or* simply -save

[ThSITC]

May I please request a NEW OPTION -savesuccess  in NetRexx 1.04 RC2
which does actually SAVE
the successfully reported as *compilation successful* members of a list
of compilations ?

Currently, the NetRexx Compikler may report a lot of compilations as
successful, but does *not*
store this result int the proper .class files.

The whole result is both misleading and frustrating:

You get the report abcdef.nrx compilation successful, but when You are
compiling a bunch,
if any other program fails to compile, no compilation is done at all!

Thus, an option -savesuccess *or simply* -save might help developers :-)
Thomas Schneider





_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[Kermit Kiser]

Hi Thomas!

I commend you on requesting an option rather than trying to change the
translator behavior for everyone as in your previous requests concerning
this issue. You can open a ticket on Kenai if you wish but I doubt that
the option will get implemented because the majority of programmers
believe that this is a very bad idea.

What will you do if you have an application with 100 modules in a
directory structure and you implement a new feature that changes 50
modules but only 25 compile OK while another 25 fail to compile and your
system is now broken and there is no easy way to put things back together?

Of course you can easily write your own compile script using the
documented APIs for NetRexxC to compile your independent modules
separately and save the successful results the way that you wish.

If I have misunderstood your use case, please clarify what you are doing
and why you need this feature.

-- Kermit


On 6/3/2015 6:07 AM, Thomas Schneider wrote:

> May I please request a NEW OPTION -savesuccess  in NetRexx 1.04 RC2
> which does actually SAVE
> the successfully reported as *compilation successful* members of a
> list of compilations ?
>
> Currently, the NetRexx Compikler may report a lot of compilations as
> successful, but does *not*
> store this result int the proper .class files.
>
> The whole result is both misleading and frustrating:
>
> You get the report abcdef.nrx compilation successful, but when You are
> compiling a bunch,
> if any other program fails to compile, no compilation is done at all!
>
> Thus, an option -savesuccess *or simply* -save might help developers :-)
> Thomas Schneider
>
>
>
>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>
>
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[ThSITC]

First of all, Kermit & Rene: thanks for your excellent work on NetRexx
3.04 (sorry my mis-spelling 1.04 below :-()

Second I do observe a couple of things with NetRexx 3.04 RC1:

1.) The average *delay* after starting a compile is drastically *longer*
than 3.03. In my case, the delay is longer than 1/2 minute!
2.) When an import statement references a *non existing/not accesiible
module*, former NetRexx versions did report that immediately.
in NetRexx 3.04 those imports are reported as a javac error -- after
generating the program.

As I am novice to NetRexx 3.04 I'm just *reporting* that, *not complaining*!

I do have to rethink about Your answer/question in Your second paragraph.
Keep up the good work, I'm only 1 user ;-)

Kind regards,
Thomas.
==============================================================================================================


Am 04/06/2015 um 08:55 schrieb Kermit Kiser:

> Hi Thomas!
>
> I commend you on requesting an option rather than trying to change the
> translator behavior for everyone as in your previous requests
> concerning this issue. You can open a ticket on Kenai if you wish but
> I doubt that the option will get implemented because the majority of
> programmers believe that this is a very bad idea.
>
> What will you do if you have an application with 100 modules in a
> directory structure and you implement a new feature that changes 50
> modules but only 25 compile OK while another 25 fail to compile and
> your system is now broken and there is no easy way to put things back
> together?
>
> Of course you can easily write your own compile script using the
> documented APIs for NetRexxC to compile your independent modules
> separately and save the successful results the way that you wish.
>
> If I have misunderstood your use case, please clarify what you are
> doing and why you need this feature.
>
> -- Kermit
>
>
> On 6/3/2015 6:07 AM, Thomas Schneider wrote:
>> May I please request a NEW OPTION -savesuccess  in NetRexx 1.04 RC2
>> which does actually SAVE
>> the successfully reported as *compilation successful* members of a
>> list of compilations ?
>>
>> Currently, the NetRexx Compikler may report a lot of compilations as
>> successful, but does *not*
>> store this result int the proper .class files.
>>
>> The whole result is both misleading and frustrating:
>>
>> You get the report abcdef.nrx compilation successful, but when You
>> are compiling a bunch,
>> if any other program fails to compile, no compilation is done at all!
>>
>> Thus, an option -savesuccess *or simply* -save might help developers :-)
>> Thomas Schneider
>>
>>
>>
>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>>
>>
>>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[rvjansen]

Thomas,

I have an idea about the delay and it stems from the environment that I know that you are using. I will send you some testcases.

Can you send us an example of the second issue?

best regards,

René.

> On 4 jun. 2015, at 14:17, Thomas Schneider <[hidden email]> wrote:
>
> First of all, Kermit & Rene: thanks for your excellent work on NetRexx 3.04 (sorry my mis-spelling 1.04 below :-()
>
> Second I do observe a couple of things with NetRexx 3.04 RC1:
>
> 1.) The average *delay* after starting a compile is drastically *longer* than 3.03. In my case, the delay is longer than 1/2 minute!
> 2.) When an import statement references a *non existing/not accesiible module*, former NetRexx versions did report that immediately.
> in NetRexx 3.04 those imports are reported as a javac error -- after generating the program.
>
> As I am novice to NetRexx 3.04 I'm just *reporting* that, *not complaining*!
>
> I do have to rethink about Your answer/question in Your second paragraph.
> Keep up the good work, I'm only 1 user ;-)
>
> Kind regards,
> Thomas.
> ==============================================================================================================
>
>
> Am 04/06/2015 um 08:55 schrieb Kermit Kiser:
>> Hi Thomas!
>>
>> I commend you on requesting an option rather than trying to change the translator behavior for everyone as in your previous requests concerning this issue. You can open a ticket on Kenai if you wish but I doubt that the option will get implemented because the majority of programmers believe that this is a very bad idea.
>>
>> What will you do if you have an application with 100 modules in a directory structure and you implement a new feature that changes 50 modules but only 25 compile OK while another 25 fail to compile and your system is now broken and there is no easy way to put things back together?
>>
>> Of course you can easily write your own compile script using the documented APIs for NetRexxC to compile your independent modules separately and save the successful results the way that you wish.
>>
>> If I have misunderstood your use case, please clarify what you are doing and why you need this feature.
>>
>> -- Kermit
>>
>>
>> On 6/3/2015 6:07 AM, Thomas Schneider wrote:
>>> May I please request a NEW OPTION -savesuccess  in NetRexx 1.04 RC2 which does actually SAVE
>>> the successfully reported as *compilation successful* members of a list of compilations ?
>>>
>>> Currently, the NetRexx Compikler may report a lot of compilations as successful, but does *not*
>>> store this result int the proper .class files.
>>>
>>> The whole result is both misleading and frustrating:
>>>
>>> You get the report abcdef.nrx compilation successful, but when You are compiling a bunch,
>>> if any other program fails to compile, no compilation is done at all!
>>>
>>> Thus, an option -savesuccess *or simply* -save might help developers :-)
>>> Thomas Schneider
>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Ibm-netrexx mailing list
>>> [hidden email]
>>> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>>>
>>>
>>>
>>
>> _______________________________________________
>> Ibm-netrexx mailing list
>> [hidden email]
>> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>>
>
> _______________________________________________
> Ibm-netrexx mailing list
> [hidden email]
> Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
>

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/