Oh ok ... guess I didn't realize that WET is being provided primarily for 
'backward compatibility'.
Thanks for the clarification on the other bits.

> Using cURL or libcurl is not inherently dangerous. Any code that goes
> into production should be peer reviewed. You can write bad code in any
> language using any tool.

Again ... you've over-generalized a very specific scenario I said folks should 
be wary of.
I didn't say curl or any other tool is dangerous, piping source-unknown scripts 
to bash is!

- KB

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On Friday, July 24, 2020 10:15 AM, David Crayford <[email protected]> wrote:

> On 2020-07-24 12:02 PM, kekronbekron wrote:
>
> > > I wouldn't. I would recommend using a sophisticated networking library
> > > like Java or whatever your favorite language is on the JVM.
> > > Can't figure out if you're kidding...
>
> No, I'm not kidding! IMO, unless you have a critical requirement to web
> enable legacy languages then I would avoid WET at all costs. A quick
> browse of the samples is enough to conclude that
> while it may work it is hideously complicated compared to similar
> function in modern languages. So, why not just use a better language? I
> almost died laughing when I saw how complicated it
> is to parse JSON in REXX using the WET.
>
> > > Who told you that? My employer offers a cURL port for z/OS and it's well
> > > maintained with support for production environment.
> > > Ok, Rocket's curl?
> > > What's the percentage of clients that want a separate product for 
> > > something that also comes with (or at least used to?) the OS (Ported 
> > > Tools).
>
> FYI, IBM sold ported tools to Rocket years ago. There is no cURL that
> comes with z/OS.
>
> > Yes, everything I'm saying is subjective...
> > Adoption would be much higher if Ported Tools' curl were actively developed.
>
> It is! And you can download it for free if you just want to write
> in-house tooling. If you want to use cURL in production then you will
> have to buy support.
>
> > Eh ... I didn't say curl-ing a script is dangerous and CWET isn't.
> > I meant piping any source-unknown script direct for execution is not a 
> > great idea.
>
> Using cURL or libcurl is not inherently dangerous. Any code that goes
> into production should be peer reviewed. You can write bad code in any
> language using any tool.
>
> > -   KB
> >
> > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
> > On Friday, July 24, 2020 8:53 AM, David Crayford [email protected] wrote:
> >
> > > On 2020-07-24 11:12 AM, kekronbekron wrote:
> > >
> > > > Just mentioned ASM / COB CWET for options really.
> > > > They're a a lot more involved than the Python client (when that's 
> > > > available).
> > > > curl is ok as a user, but when you want to productionize something, I 
> > > > would think the recommendation would be to use CWET.
> > > > I wouldn't. I would recommend using a sophisticated networking library
> > > > like Java or whatever your favorite language is on the JVM.
> > >
> > > > Not saying curl is a bad tool, it is handy & does what it does.
> > > > Ease of use does not mean it's the solution of choice in many 
> > > > controlled environments.
> > > > By loved I mean does it get upgrades/improvements?
> > > > Who told you that? My employer offers a cURL port for z/OS and it's well
> > > > maintained with support for production environment.
> > >
> > > > I don't know I'm just asking..
> > > > curl-ing a shell script directly is bit ... dangerous.
> > > > That's purely subjective. I don't see why cURL would be any more
> > > > dangerous than writing a Python script or using CWET.
> > >
> > > Lots of people are using Git for DevOps on z/OS and that uses cURL for
> > > ssh and https transport.
> > >
> > > > Not in this case as the script is available to inspect.
> > > > --
> > >
> > > For IBM-MAIN subscribe / signoff / archive access instructions,
> > > send email to [email protected] with the message: INFO IBM-MAIN
> >
> > For IBM-MAIN subscribe / signoff / archive access instructions,
> > send email to [email protected] with the message: INFO IBM-MAIN
>
> --
>
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to [email protected] with the message: INFO IBM-MAIN

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to