[Info-vax] Questions and observations about OpenVMS

[Arne Vajhøj]

On 3/6/2021 4:44 PM, Forrest Aldrich wrote:
> 1. Current/modern applications for OpenVMS.  What companies are using it 
> and for what reason do they continue to remain on OpenVMS vs another OS 
> (apart from older proprietary code that simply works, as-is).

The apart piece is probably the main reason.

A lot of VMS installations will be running some old application,
porting to another platform has been considered and rejected due to
some mix of cost and risk.

> 2. Direction of OpenVMS.  Apart from older industries, I wonder about 
> the future direction of OpenVMS.  They will be completing the port to 
> x86_64, which I believe opens many opportunities for the platform. Could 
> we see OpenVMS become a desktop server, or in smaller roles where 
> clustering may also be applied.  Might we see companies like Autodesk, 
> Adobe, et al embracing the platform for their products.   What would 
> that mean to smaller shops that run their own servers -- or, might we 
> even see OpenVMS operating in the cloud as well.    In these scenarios, 
> I wonder if OpenVMS's native clustering capabilities would benefit 
> resource-intensive applications -- imagine being able to cluster a few 
> boxes you happen to have around as a "standard feature" (Linux lacks 
> this, mostly).

Very few see any future for VMS as desktop computer.

It seems likely that VMS will to some extent move to public cloud.
Most other platforms are.

Everybody is clustering today, but the industry trend is application
clusters not OS clusters.

> 3.  Licensing, relative costs.  OpenVMS is an extremely expensive OS to 
> license, which may present challenges in its migrating into other 
> industries.  Depending on #2, it might be difficult for a systems admin 
> to justify the cost, versus other common solutions available.   I live 
> close to OpenVMS's home office; my communications with them have been 
> met with cold responses -- I've recommended they consider different 
> licensing models.  For example, their community-based licensing is nice, 
> but it's a one-off and you won't get upgrades.   As a systems admin, and 
> depending on #2, I'd prefer to have a full-on, non-commercial license 
> that receives all the updates -- there could be many benefits of getting 
> OpenVMS into the hands of smaller shops and sysadmins.

VSI need to make money. With the VMS volume then VMS cannot be
low cost. VMS will have to go for the market where companies are
willing to pay.

I believe the indication is that VSI will be moving to a
Redhat like licensing model.

> 4.  Modern/general perception of VMS.  When I mention VMS to anyone, 
> I've seen eyes roll and mentions of "dinosaur" etc., but I see a 
> potential "gold mine" if OpenVMS could work in the above --  with the 
> right partnerships to bring more commercial (and open) code to the 
> platform.   My sense is that OpenVMS probably has "enough" high-end 
> commercial industry that they can afford not to care, but for how long? 
>   They have been mum on future plans, but I can't imagine it doesn't 
> include at least some of this.

ISV support and open source support is important for VMS.

> 5.  Bringing OpenVMS into the modern age.  From scanning different 
> articles, it seems there remains some work in bringing OpenVMS into the 
> modern world, for example updates to DCL and other relative quirks.

DCL is a bit old/fashioned.

But investing in DCL enhancements is not what is going to
boost VMS sales.

Arne