[Info-vax] Using VMS for a web server

[terry+googleblog at tmk.com]

On Friday, June 5, 2015 at 8:54:34 PM UTC-4, David Froble wrote:
> If this is the case, then why are people still using VMS?
> 
> So, why then is anyone still using VMS?
> 
> Perhaps there are VMS problems that customers can use VMS to solve in 
> the most effective manner?

They may have a system that meets their needs where they don't feel the need for new programming / sysadmin support. If the VMS system just sits in the corner performing its task, then things may be perfectly fine until it breaks. At which point the customer may have to face the hard reality of pricing for spares / support on the existing system, and/or the cost to migrate to newer, supported hardware and software. At which point it is entirely possible that they'll do the least expensive thing to limp along until the system is replaced by something else. This may be complicated by no longer having the source code (or having it, but it no longer compiles) to their applications.

By definition, HP only has accurate information (and has presumably passed at least some of it along to VSI) about customers who are still paying HP for hardware and/or software support. The rest of their "customer database" is likely inaccurate (companies that have gone out of business/merged, who no longer run VMS, or who still run VMS but don't see any benefit to what HP is offering).

Paying customers are almost definitely a small subset of the number of non-hobbyist systems deployed, although I'm sure there are some MAJOR customers still paying for support.

The challenge for VMS, and thus for VSI, is to bring a new VMS platform to the attention of both of the following groups:

  1) Shops still running VMS, but either without support or who pay some 3rd party for support
  2) Shops that either never ran VMS or who already migrated away from VMS for some reason

In other words, attracting new business. It would seem that VSI feels that there is enough of a market there to justify doing the work they're doing. I mentioned in much older posts that the desires of those two groups are pretty much the opposite of each other - the first group wants to run their existing applications (possibly without recompiling, if they no longer have the sources or there is no compiler support for them). For this group, the less change the better. And "modern" VMS and the hardware it runs on need to be priced competitively enough that they aren't dismissed out of hand in favor of migration. The second group wants the exact opposite, feature-wise - a modern operating system with modern language and library support where they can re-implement their existing applications and grow. But they also want an environment where VMS + hardware is cost-competitive with what they're already doing. In fact, for these potential customers the advantage needs to be more than that - VMS needs to make the case for being a better solution for them, not just one "that doesn't cost much more"