[Info-vax] Ghostscript and HTML Browser on X86

[Andreas Gruhl]

Arne Vajhøj schrieb am Montag, 12. Dezember 2022 um 15:57:05 UTC+1:
> On 12/12/2022 6:53 AM, Andreas Gruhl wrote: 
> > We have successfully ported our first application (Planning for 
> > public transport companies, 1.6 Mio Lines of Pascal) to X86. 
> > 
> > But: the application generates HTML and Postscript files. On Alpha 
> > and Itanium we can directly show these to the user via an old Mozilla 
> > browser and via Ghostview/Ghostscript. The browser is automatically 
> > launched by our application and directed to the correct page. 
> > Everything works just fine. 
> > 
> > Where can we get an equivalent of these display products for X86? 
> > 
> > Well, we have Sourcecode for Ghostscript and will try to build it - 
> > it doesn't look easy though. 
> > 
> > But we see no chance of raising an HTML browser on X86. We do not 
> > need all the latest bells and whistles. The software might be 10 
> > years of age or even older - if it can be built on Alpha or Itanium, 
> > we could try building it on X86, provided we had access to the source 
> > code. 
> > 
> > And before you ask, why we not simply use display software on the 
> > user's desktop, here is my answer: the only connection between the 
> > desktop and our VMS server is via the X Protocol. This is commonly 
> > routed through products like Citrix or Teamviewer and the Exceed 
> > display server. So interactions between the VMS server and the 
> > desktop are quite limited.
> I believe GS and GV are written in C. And VMS C is VMS C. So it should 
> be possible to get those running on VMS x86-64. Maybe someone 
> even already got them running. 
> 
> Web browser is an entirely different story. I think the short 
> version is that it will not happen. VSI does not have a commercial 
> justification for porting a browser to VMS and the VMS community are 
> not able to mobilize enough volunteers to do so. Technically it is 
> of course possible - and when the CLang compiler become available for 
> VMS x86-64 it will become even easier, but I don't think it will happen. 
> 
> Bottom line is that you will need a different workflow. Time to sit down 
> with the network people and the security people to discuss what can 
> be done. 
Unfortunately these are different persons for each of our customers.
So we simpy can't have too much discussion with all of them.
We have to come up with a solution which minimizes discussions.
> 
> There are plenty of possibilities to discuss (hopefully there is at 
> least one that will be considered acceptable): 
> * VMS FTP/SFTP/FTPS files out to a Linux server inside the 
> secure zone and browser on that Linux server is accessed 
> via tunneled X (requires a new Linux server inside the secure 
> zone) 
> * VMS FTP/SFTP/FTPS files out to a Linux server outside the 
> secure zone where the files can be accessed from the desktop 
> (requires some outgoing connectivity, but outgoing is sometimes 
> easier to get approved than inbound) 
> * install a HTTP server on VMS and serve the files from that 
> (requires opening up port 80) 
> * install a HTTP server on VMS and serve the files from that 
> and setup -firewall-proxy server-firewall-VMS (requires 
> some configuration and possible even some hardware) 
> * find/write a tool that are able to copy files over X protocol 
> and use that tool to move the files from VMS to a Linux server 
> outside the secure zone where the files can be accessed from 
> the desktop (requires software, and even though it does not 
> require new protocols to be allowed then it may actually be 
> the same from a real security perspective) 
> 
> [there may be other options, but that was those I could think of] 
> 
Thanks for your suggestions. We will consider them.
But whatever we do - it won't make things easier for us.
But 
> Arne