[Info-vax] Any stronger versions of the LMF planned ?, was: Re: LMF Licence Generator Code

[Jan-Erik Söderholm]

Den 2021-08-12 kl. 08:58, skrev Lawrence D’Oliveiro:
> On Thursday, August 12, 2021 at 2:58:11 AM UTC+12, Bill Gunshannon wrote:
> 
>> Here we are 30 years later and there is still
>> new COBOL being written every day.
> 
> By whom, and for what?
> 
> What I found ironic was the premise that COBOL was written specifically for “business-oriented” uses, eschewing any of that “scientific” stuff like mathematical notation and floating point, or even any decent dynamic string handling.
> 
> Then, a decade or two later, came along these things called “relational databases”, which were enthusiastically adopted by businesses--the very market that COBOL was supposedly optimized for.
> 
> But it turns out the best way to interface to a relational DBMS is to generate SQL query strings.

No, it is not. It is inefficient (since it needs to "compile" the
SQL at each execution) and dynamical SQL statements are the source
for "SQL injection" issues.

Proper application code used fixed precompiled SQL statements
with paramater markers.


> And for that, you need decent string handling, with facilities for format substitution, argument quoting and the like. None of which were envisaged in the original design of COBOL.

Just totaly wrong. Cobol works perfectly together with RDBMS's.


> 
> So today, even a language like Python, Perl or (spit) PHP would be a better fit for “business needs” than COBOL ...
>