[Info-vax] OpenVMS printing to PDF

[johnwallace4 at yahoo.co.uk]

On Wednesday, 20 May 2015 21:41:18 UTC+1, JF Mezei  wrote:
> On 15-05-20 11:11, Jan-Erik Soderholm wrote:
> 
> > You are way out of reality with JS, callable HTML5 rendering...
> 
> 
> Man, am I ever happy to see that someone else is accused of living in a
> different Universe :-) Welcome to the club Hoff :-)
> 
> In fairness to Hoff, when presented with a problem, it is a good
> exercise to look at a variety of ways to solve it. And Hoff has
> presented many ways.
> 
> With regards to Hoff and CDA:   VSI could brush up the API documentation
> and decide that this will be a strategic part of VMS and encourage
> people to write converters for their file formats.
> 
> Remember that Digital abandonned CDA and sold much of it to other
> companies. There is no incentive to write converters to an architecture
> that has been abandonned.

There may be an incentive to write converters (or whatever is
appropriate, which might be e.g. to update JetForms?) if the new
or updated products offer a better solution to a particular requirement
than stuff that is readily available.

"Better" implies a value judgement; two sets of people faced with the
same set of requirements may not agree what the "best" answer is.

For example, some people *are* going to be understandably sceptical
about whether a pseudo-interactive browser is an appropriate component
of a "data to print" subsystem.

Having personally tried to play this "document formatting from
scratch, non-interactively" game on Windows XP a few (5?) years back,
and got basically nowhere despite being in a reasonably competent
Windows-centric environment, I'd start off as one of the sceptics,
especially if there was any expectation of using today's browser
toolkits (together with the wide variety of incompatibilities
between browser implementations, and the wide variety of
implementation-specific security holes associated therewith,
ranging from JPEG-processing holes, to JavaScript holes, to who
knows what).


Back then, I did end up looking at HTML as an output format for
subsequent interactive use, and it worked after a fashion, but
it wasn't right for generic automated printed report generation.

It'd be great to be convinced otherwise, if appropriate evidence
were available.

"Everybody else does xyz" isn't necessarily evidence. Everybody else
frequently isn't allowed to think that there may be other more
appropriate approaches for people willing to "think different",
rather than (e.g.) stick with their favourite IT/development
department monoculture.

One size does not fit all.