If OpenVMS is profitable, where's the money going?

[John Eerenberg]

Antoniov,

> John,
> I don't agree with you: HP is handling it
> better then Compaq;
I hope you are right in the general case.

My comments were made in the context of the financial industry.

Tell me one CIO/CFO/CEO in the financial markets in the US especially the southest that knows about VMS and is smart enough to make it "polically correct" in their own orgraniation?

To address the general case outside the financial industry, I am willing to bet that within the Fortune 100 that almost all CIO/CFO/CEO politicaly support IBM, HP, etc. products but it will be a rare CIO/CFO/CEO to explicitly understand and desire VMS.

Can you name one CIO/CFO/CEO within the Fortune 100 that understands the massive benefits of VMS in terms of how it helps their business, desire's VMS for their organization due to these benefits, and goes on record with their peers backing them down when someone says, "Oh. VMS, that dead, antiquated OS." ?

The reality is that all us techies can be (or for newbies -- become) enthusiastic along with many low level managers, but the culture has to change at HP sales -- that is a given. And, perhaps more importantly, the culture has to change at at the CIO/CFO/CEO level of the large companies too (maybe the non-large companies too?). These executives are proud about their technology decisions; they privately brag about how "this or that technology" will enable them to stomp the competition, make them more efficient, reduce costs, etc. The "this or that technology" does not include VMS. I would honestly like to read about a few dozen CIO/CFO/CEO within the Fortune 100 put VMS in a positive light. :-)

If I can read such commentary, I will then take a few examples and present them to a mid-level manager I know. ;-)

My point is A) this could be the reason why products subordinate to VMS are gaining market share and VMS is in limbo at best. B) To steal and change a line from James Carvel, "IT'S THE CULTURE, stupid."

[labadie_1]

I agree with others that I feel HP thinks Vms is for big customers: but there is a paradox: no Erp (Sap, Oracle applications...) is available on Vms. And I know a good number of sites have left Vms because they were installing an Erp, not because they were unhappy with Vms. And with an Erp, we can be sure this customer will be on Vms for long.

:-(

[Antoniov.]

John, Labadie,
focus is "there is no enough software for Vms" so I repeat HP should help software house that's developing using this platform.
My post abou C compiler and A/S recall the real difficult for developments and customer too.
Big customer can't find erp, little customer pay for free software on concurrent platforms.
Vms is old? When Linus Toward thought to linux, many people said *nix is old but now Linux is used worlwide; Vms is more strong and more secure than Linux: if OS is released with Apache, PHP, Perl inside may be offered as secure web server.
It's only an idea but focus is on software avaiable (free and payment software).

@Antoniov

[Willem Grooters]

Gerard is right.
That's just one of the major issues, as I already pointed out (as stated, we had this on our back). I don't think this is an issue for the sales department directly (although sales could play a big role in it), probably more a management issue.
AKAIK, companies that deliver ERP packages (JDEdwards, SAP, BAAN are the ones I know of) run on HP-UX - I would think these are HP-business partners. I won't say they need to be forced to port their applications to VMS, but some pressure from HP could help. Both HP-UX and VMS are now HP products so HP won't lose anything!
It has been shown before - at least, it has been stated - that porting a Unix application to VMS wasn't that difficult with the GNV initiative. So what is the reason that at least the basic ERP stuff hasn't been ported yet? If this is because of the uncertainty of VMS's future and market, then there is quite a lot of explanation to be done - mostly by the partner-channel department, but by ESG management as well!

[Willem Grooters]

Antonio,

Nice thought on "contacting a softwarehouse" for developing thes products.
I guess a lot of VMS-professionals would welcome the idea but I'm afraid it won't happen - unless (see my previous port) bigger ISV's would be involved to have their application ported (this time: properly: to VMS, not to GNV).
Nevertheless, I think that even in a less formal circuit there is a lot of potential. There is a good deal of OS-related and system-management-related software available - as you will undoubtably know. But the main issue is all about business applications: non-technical stuff. You don't do this on your own - and not without some secured payment (unless you have other ways of income; but how many of us can be called a multi-millionaire...).
But I agree that lower prices, or inclusion of licenses of certian products would help. TCPIP should be part of the OS anyway (no extra licenses), and compilers need to be affordable. It need to be paid, though, by your customer!
Otherwise, you'll be stuck with (Unix-based...) open source software, as far as it runs on VMS...

[John Eerenberg]

Willem,
> "Gerard is right."
> "That's just one of the major issues, as I already pointed out (as stated, we had this on our back). I don't think this is an issue for the sales department directly (although sales could play a big role in it), probably more a management issue."

Agreed. And it makes sense that if the sales offices have some level of autonomy, then management, perhaps at the Carly level, has to set the direction that would help clear out any biases or, forget bias, just clear away the apathy toward VMS.

I believe there is a rather big disconnect. OpenVMS enigneering continues to achieve and push the technical excellence of VMS, yet, sales and marketing do not understand and appreciate it. As I have said before and others agree with, VMS is seen more as a sales vehicle to migrate to another platform.

On the subject of adding more free compilers, etc. to VMS I think this will help developers and attract them, make things easier. Outside the developer community, a danger exists that if too much free stuff is added, the value of VMS is undermined. In my experience, price is not really a critical factor. Those that want cheap will achieve it. Those that want the best, must be convinced that the extra bucks gives them the best. Too low a price will make them suspecious. I look at IBM and see technical mediorce products being sold as "the best" because IBM will do back flips to support it when mud hits the fan. While HP-UX may have this kind of reputation, I really don't believe HP has the culture to pull this off for VMS.

Much more to discuss, but that's all for now. ;-)

john

[Antoniov.]

John,

"In my experience, price is not really a critical factor. Those that want cheap will achieve it"

I think price is an important factor for mass diffusion; it's no unique factor but when a potential customer look for two products, if he feel them as similar then price become critical.
I agree low price and cost can't help so I specified a few software: A/S pathwork and C compiler, because HP-UX has already theese two product included (TCP/IP is included since a few years).
About Apache, PHP, Perl and other open source, they may included in distribution kit (again HP-UX make this).
It's not only a price question: today any operating system need to exchange data with other worlds, so keep payable A/S Pathwork (for example) limit connectivity of VMS.
I think Vms may be more expensive than *NIX or Windoze platform because more stable and it's better but can't have less usefull characteristics.

@Antoniov

[Ian Miller.]

the way the licensing is to be done on itanium VMS (bundles - small, medium, large) and the lower price on itanium hardware may help depending on what is included at level.
mainly if the right things are included in the cheapest level.

[Antoniov.]

Good idea,
so HP can promote either Vms either Itanium.

@Antoniov

[Brad McCusker]

think some software could be avaiable with OS without licensing; it's already happened for TC/IP and DCPS: could be same for Advanced Server/Pathworks


Getting a bit off topic here (what is the topic anyway?)

The reason Adavanced Server for OpenVMS and PATHWORKS for OpenVMS are not bundled with the OS is becuase of the original licensing agreement with AT&T (and, further upstream, Microsoft, I assume) which requires a royalty payment for each and every connection to the LanManager server.

I know there have been attempts to negotiate or buyout the contract, and, each effort proved to be un-economical.

Beleive me, I'm the guy who used to have to stand in front of audiences and explain how the licensing worked, if there was any way to get rid of licensing for PW/AS, I think we would have.

Brad

[Ian Miller.]

re pathworks - we will have to use samba then :-)
re subject of this thread - I don't know but its interesting nevertheless

[Antoniov.]

Brad,
I didn't tought the royalty side of problem; thank for your explanation.
However, as Ian posted, may be usefull offer samba: HP already include it with HP-UX (sorry if I repeatly recall HP-UX).

@Antoniov

[John Eerenberg]

Yeah, we may be a little off topic concnering the original post and the Inquierer.net article.

About that article, if the $ figures are in the ball park, it would seem to validate that VMS is healthy financially.

In general, I think this discussion is very good since we are all wanting to make VMS the best it can be and successful!

john

[Willem Grooters]

Ian,
ARE you getting an old man - presumably a lot of us seem to be - or just missing your original?

Ian Miller Apr 23, 2004 16:37:44 GMT N/A: Question Author <------------- ????

--------------------------------------------------------------------------------
re pathworks - we will have to use samba then :-)
re subject of this thread - I don't know but its interesting nevertheless <-----------????

Never mind ;-D

Brad: Original tpic was: What is HP doing with the money earned licensing VMS? In other words: what do licensees pay licenses for? IMHO, I don't mind the shift of atention.

Ian: SAMBA could be a nice alternative if it was as stable and secure as AS. anyone ever tested Samba for buffer overflow? It's coded in C and therefore _could_ be vulnarable (like SSL proved to be).

A new pricing scheme - bundling - on Itanium is a nice idea, but will this transfered to AXP and VAX as well, and what would teh amount be. If just 1-5%, who cares. Should be noticable!
Furthermore, I don't think that price of the license of the OS itself matters very much. I suggest that if support was cheaper and more leveled to what a company needs, it would help even more. So a 1 hr respons would be far more expensive than 1 day. Understood and accepted. THAT makes sense. Also, that if you find a problem with the OS, you can have at elast some sort of support - if only you could drop it somewhere. But correct me if I'm wrong - I'm not involved in such matters.
As a developer, I would pay a fair price on compilers if I could be sure they would fit the hardware and adhere to VMS standards. Documentation inluded, please! I wouldn't mind cheaper compilers though.

Willem

[Jan van den Ende]

Brad:


of the original licensing agreement with AT&T (and, further upstream, Microsoft, I assume) which requires a royalty payment for each and every connection to the LanManager server.


would that not equally aply to Samba?
I don't have any personal knowledge, but a seem to have heard that Samba also uses LanMan, and if I read Antoniov correctly, THAT is included into base HP-UX?
Someone should be talking to somebody high enough in "the price deciding team" about equal monks and equal cassocks...

Jan

[Brad McCusker]

>>would that not equally aply to Samba?

No. Samba was developed without use of Microsoft technolgy (I assume).

AT&T licensed NT file and print server code (sometimes known as LanManager, SMB, CIFs) from Microsoft for the purpose of developing an NT compatible file and print server for UNIX. AT&T developed what they call the "Advanced Server for UNIX", and licensed it to various UNIX vendors, including HP and DEC. DEC in turn ported it to its UNIX flavors and also ported it to VMS. Somewhere along the way DEC also renamed it to PATHWORKS, and then later on changed the name to Advanced Server.

Did I lose you yet?

The point is, Advanced Server for OpenVMS makes use of original Microsoft sources. Yes, some of the source code actually has a Microsoft copyright, an AT&T copyright, and then the DIGITAL/Compaq/HP copyright.

Samba does not have any of that baggage.

As for Samba on VMS, there does exist ports, and, I fully support that work. The work I am currently involved with (UNIX Portability) should make it much easier to port current versions of Samba.

Hope this helps.

(By the way, I too like the discussions in this thread, regardless of what the original topic is or was)

Brad

[Ian Miller.]

I posted originally June 5th 2003. I wonder if this thread will still be added to in June 2004 :-) As long as it stays interesting.

[Antoniov.]

Ian,
you posted some very very interesting thread.
Congratulation!

@Antoniov

[Steve Reece]

I tried running Samba on OpenVMS Alpha v7.3 and never got it to work successfully...

It's useful that Brad has posted to this thread though and with the context of Samba knocking around as one thing that Ian and I discussed off-line a few weeks ago was the (lack of) availability of a configure (?)function on VMS. This is typically used by unix/linux code through their installation and compilation process to select which bits of which product exist and to establish what the environment is that the software has to be configured to. To the best of my knowledge though, nothing like this exists on VMS plug-compatible.

Is your team looking at this kind of thing please Brad?

More on the initial line of the thread, while VMS sales are profitable, they also cost more to create. A Pre-Sales person, an account manager, a VMS Ambassador and probably a reseller being used for the average SME (small/medium-sized enterprise) buying in VMS for the first time. Compare that to a Proliant server with Windows on it where the box has little margin but virtually sells itself.

I see no way that HP would agree to sell off the entire product line of OpenVMS in both software and services. The two would have to go together and would put a serious dent in the headcount at HP as well as in the revenues generated. Anyone who looks at the VMS jobs market though will recognise that it's pretty stagnant - maybe that's because of all the headcount reductions from HP and Compaq/Digital before them, flooding the market with people that are good at their jobs though?

[labadie_1]

The problem with Samba was with 2.2.8 ?

http://www.pi-net.dyndns.org/anonymous/jyc/

I think Vms has today a big weakness: no Websphere (Ibm), Jonas or Jboss. Bea Weblogic 8.1 is available on Linux Solaris Windows, but not on Vms.

I know that a group of people are porting Jonas to Vms, but I think this should be Hp work, as Sap should be available on Vms, like it used to be (long ago).

[Martin P.J. Zinser]

Hello Steve,

to prolong this thread a bit more ;-)

I do believe that being able to just do
a ./configure is one of the main drivers for doing GNV.

Going this route has some implications:

* You do need GNV on each system you want to
build the software on
* GNV is not available for VAX
* Current versions of GNV requires VMS 7.3-1 or higher

For me personally these restrictions make this approach not attractive for the freeware projects I do support for general distribution. So I did some work on this on my own. I developed a Perl script that reads a config.h.in and produces a DCL that will check for the properties found in there.

Note: Perl is needed by the person porting the software in the first place. The result is a DCL script that can be run on any VMS system without other requirements on installed software.

Clearly this is work in progress, I do not have yet tests for all possible conditions in a config.h.in, but the script is easily extenable and does help me nicely with my own ports. This and a number of other DCL snippets that form my construction kit for building VMS build procedures for freeware can be found at

http://zinser.no-ip.info/wwwar/mgr/btools/

Greetings, Martin

[Keith Parris]

> Bea Weblogic 8.1 is available on Linux Solaris Windows, but not on Vms. <

The BEA website indicates Weblogic 8.1 SP1 is supported on OpenVMS today. See http://e-docs.bea.com/platform/suppconfigs/configs81/hpovms73_alpha/81sp1.html

The OpenVMS Rolling Roadmap at http://h71000.www7.hp.com/openvms/roadmap/openvms_roadmaps_files/OPENVMS_ROADMAPS.PDF puts BEA Weblogic 8.1 SP3 support in 2005, and notes a BEA Weblogic campaign as going on now.

[Ian Miller.]

As I said to Steve see also
http://www.oooovms.dyndns.org/

[John Eerenberg]

I like discussing all the features and things that VMS can do, but it is, in the big picture, noise.

As a techie, I've watched VMS enhanced many times and become many times more valuable. At the same time, VMS' value in the market place has diminished. The techies and managers that know and understand VMS are shrinking. Most IT managers don't understand VMS and its value, its distinction, it's capabilities now and in the future.

The trade rags are not neutral; they make comments either positive or negative. The recent 64bit computing article in Computerworld is very telling of this commentary -- the Alpha "is [now] abandoned." I think the Alpha is tightly associated with VMS that as Alpha fades away, VMS is thought of as fading away (it doesn't help that Itanium is not broadly accepted yet).

I love VMS and I'll be there at the bootcamp digging into all the internals and features being presented -- as well as attending the crash dump class the week before. Most of us make a good living because of VMS. Today's perception is that *nix is a much better career path. I have trouble arguing that one.

Normally, I would not put so much time into posting more then one or two responses. I believe VMS is worth the time to discuss and to keep focused on the issues raised in this thread. I am hoping something will result from all this.

Is this a BoF topic at bootcamp?

[Willem Grooters]

John,

Is this a BoF topic at bootcamp?

For what I know: Not really - at least, officilally. Though I think it SHOULD be a topic, but as you already stated, it's more a SALES matter.
In that perspective:
Martin/Sue (as tyhe ones that handle the BoF (AFAIK):
Can something be arranged that this matter really IS addressed as well, or is it out-of-scope? As you both very well know, this is something that lives in this community and HP (slaes) _should_ be woken up!

Willem