OpenVMS Community License Program
-
Topic author - Valued Contributor
- Posts: 62
- Joined: Sun Jun 16, 2019 7:55 am
- Reputation: 6
- Location: Zurich
- Status: Offline
- Contact:
OpenVMS Community License Program
As it says on the News page.
That's good news!
OpenVMS cannot just exist, it must also be visible. And visibility requires a) support of normal hardware and b) availability.
In my opinion this ranks with choosing Visual Studio Code as a developer tool as one of the decisions that makes OpenVMS usable for the non-VMS community.
That's good news!
OpenVMS cannot just exist, it must also be visible. And visibility requires a) support of normal hardware and b) availability.
In my opinion this ranks with choosing Visual Studio Code as a developer tool as one of the decisions that makes OpenVMS usable for the non-VMS community.
-
Topic author - Valued Contributor
- Posts: 62
- Joined: Sun Jun 16, 2019 7:55 am
- Reputation: 6
- Location: Zurich
- Status: Offline
- Contact:
Re: OpenVMS Community License Program
To me it's obvious that if OpenVMS shall attract new users (and new developers and administrators) it must be easily available, without complicated hoops to jump through.
A community license that allows one instance of OpenVMS will not attract many new users.
Most people will try OpenVMS in virtual machines, create templates and deploy new VMs whenever needed. If this is not allowed, people simply will not try OpenVMS and will use Linux or FreeBSD instead.
In my opinion OpenVMS community licenses must not be much more restrictve than Windows licences.
We'll see what happens. But I have high hopes for OpenVMS on the technical side and I see a possibility of failure only on the marketing side.
A community license that allows one instance of OpenVMS will not attract many new users.
Most people will try OpenVMS in virtual machines, create templates and deploy new VMs whenever needed. If this is not allowed, people simply will not try OpenVMS and will use Linux or FreeBSD instead.
In my opinion OpenVMS community licenses must not be much more restrictve than Windows licences.
We'll see what happens. But I have high hopes for OpenVMS on the technical side and I see a possibility of failure only on the marketing side.
-
- Valued Contributor
- Posts: 68
- Joined: Sun Feb 23, 2020 7:49 am
- Reputation: 0
- Location: Fort Worth, TX USA
- Status: Offline
Re: OpenVMS Community License Program
Hi Maxxim,
Thank you for your replies and the effort you have put into them.
Thank you for your replies and the effort you have put into them.
Thanks. I will spread the word to those who wish to comment that they should join here and hopefully they will.marty.stu wrote: ↑Fri Aug 14, 2020 8:28 amWell, the best way of discussing this publicly is indeed this forum. A slightly less convenient communication channel is our email info@vmssoftware.com. If one wants to convey their thoughts in private, using it is fine. Still, in most cases we will only be able to forward them to higher management or answer the questions that aren't covered by NDA, so you shouldn't expect to get much more info than we are allowed to communicate on this forum.reinhardtjh wrote: ↑Thu Aug 13, 2020 6:38 pmI do hope that the people in charge can be persuaded to change their mind, however as, again, VMS Clustering is a major feature and there will be a certain number of Hobbyists that will want to do it. Especially with the upcoming x86 release. Is there an email or way of communicating with someone our request or is this forum the best place to be asking?
Thank you! That's good news.marty.stu wrote: ↑Fri Aug 14, 2020 8:28 amreinhardtjh wrote: ↑Thu Aug 13, 2020 6:38 pmAnother question that has come up is about the language compilers on Integrity. They aren't listed specifically in the license PAKs that were given yet they are in the download area for installation. Are they included in VSI's BOE? I know they weren't in the HP version. Was their omission an error or oversight?
This was indeed an oversight. We have replaced the archive on the server, and by following the same link in the email, you should be able to download an updated script that contains the full set of PAKs for the compilers on Integrity. ALPHA-LP already covers all of them, so the Alpha set remained untouched.
That's good to know, but the problem with the training kit is that it requires a Windows host. I don't even have a Windows system other than one in an Oracle VirtualBox emulator on my Mac. Others are in the same situation although I grant you probably not many. It would be nice if there was a usable free Alpha and to try a cluster would then require a much larger system to hold several such kits for a cluster. Plus it's an emulated Alpha and you are already providing clustering and HBVS for Alpha.marty.stu wrote: ↑Fri Aug 14, 2020 8:28 amYes, I follow more or less regularly what is being posted on comp.os.vms and have seen Simon's comment on HBVS and clustering, and he indeed has some valid points, but keep in mind that we are still providing the student kit (and there are no plans to change this in the foreseeable future) along with the emulator which can be used if one really wants to try how clustering and shadowing work on OpenVMS.reinhardtjh wrote: ↑Thu Aug 13, 2020 6:39 pmBased on your comments I think you have access to the discussions on comp.os.vms but here is a quote from someone there that also makes a good point:
That's probably true, but it just so happens that I am one that proves an exception. I am somewhat of a hardware junkie. Would it surprise you that I have a 27U rack in my office with the following:marty.stu wrote: ↑Fri Aug 14, 2020 8:28 amThe chances that anyone would buy a couple of Integrity servers with an MSA only to try clustering and volume shadowing seem to me pretty low, especially compared to having an x86 host with the amount of RAM sufficient to run a couple of emulated instances of Alphas.
- HPE P2000 G3 MSA dual FC controller
- D2700 disk shelf with 25 146GB and 300GB SAS drives
- HPE MSL2024 LTO4 tape library
- HPE C3000 Blade Enclosure
- HP Integrity BL860c I2 blade with 2 2-core CPU
- HP Integrity BL860c I2 blade with 1 2-core CPU
- HP Proliant BL460c G8 with 8 Core E5-2680
- HP Storageworks 8/24 FC Switch
- DEC MicroVAX 3100 Model 95
- 2 Compaq AlphaServer DS10 617Mhz
- Compaq XP1000 617MHz workstation
- Compaq 4200 series storage shelf with 14 73GB drives
- DEC InfoTower 100
marty.stu wrote: ↑Fri Aug 14, 2020 8:28 amThank you for your questions, John. This program was released only recently and we value any feedback about its implementation, terms and conditions, and technical aspects.reinhardtjh wrote: ↑Thu Aug 13, 2020 6:39 pmThank you for your time, Maxim and if there is a better way to let the people in charge know of our requests, please let me know so I can spread the word.
Kind regards,
Maxim Megalinskiy
VSI Training Team
John H. Reinhardt
VMS user since 1981
VMS user since 1981
-
- Valued Contributor
- Posts: 91
- Joined: Fri Jul 12, 2019 1:59 pm
- Reputation: 0
- Location: Netherlands
- Status: Offline
- Contact:
Re: OpenVMS Community License Program
I am VERY VERY disspointed with VSI's Community Licence Program.
Most importantly: the absence of clustering on Itanium.
No matter what reason, it will render current environments useless. I guess that most hobbyist licenees with Itanium run them in a clustered environment, for whatever reason; accepted by HPE since the first publication of the Itanium licenses. Removal renders these environments useless.
I wonder how many members of the community will accept this; how many will accept the (much extra) work to be done to come over it - if possible. Chances are that VSI (and therefore, OpenVMS) will loose very dedicated users/developers/advocates/supporters/free forum- and wiki contributers/free marketeers because they find themselves forced off OpenVMS.
Most importantly: the absence of clustering on Itanium.
No matter what reason, it will render current environments useless. I guess that most hobbyist licenees with Itanium run them in a clustered environment, for whatever reason; accepted by HPE since the first publication of the Itanium licenses. Removal renders these environments useless.
I wonder how many members of the community will accept this; how many will accept the (much extra) work to be done to come over it - if possible. Chances are that VSI (and therefore, OpenVMS) will loose very dedicated users/developers/advocates/supporters/free forum- and wiki contributers/free marketeers because they find themselves forced off OpenVMS.
-
- Site Admin
- Valued Contributor
- Posts: 96
- Joined: Tue May 21, 2019 6:56 am
- Reputation: 0
- Status: Offline
Re: OpenVMS Community License Program
John,reinhardtjh wrote: ↑Fri Aug 14, 2020 1:17 pmThat's good to know, but the problem with the training kit is that it requires a Windows host. I don't even have a Windows system other than one in an Oracle VirtualBox emulator on my Mac. Others are in the same situation although I grant you probably not many. It would be nice if there was a usable free Alpha and to try a cluster would then require a much larger system to hold several such kits for a cluster. Plus it's an emulated Alpha and you are already providing clustering and HBVS for Alpha.
I clearly understand that there are many fans in the OpenVMS world having lots of legacy hardware, and that's indeed a problem as they will have to change their environments quite heavily, but if we talk about attracting new people, the emulator should be enough as those who want to try something new (I mean the operating system they have only heard talking about) rarely buy legacy hardware. Although even if they are not numerous, such people do exist, I have to admit that. The issue that the emulator only runs on Windows is indeed valid, as all other existing opensource or free emulators are in quite poor shape or were removed from public access. I myself once tried to build es40 and although I was actually able to run it, it was painfully slow and did crash a lot, at least for me.
Anyway, thank you very much for your reports and feedback, Andrew and Willem, we have read your feedback and appreciate it as well.
Thank you for staying with us,
Maxim
Run to the bedroom, In the suitcase on the left You'll find my favorite axe.
-
- VSI Expert
- Active Contributor
- Posts: 41
- Joined: Mon Jun 10, 2019 9:23 am
- Reputation: 1
- Status: Offline
Re: OpenVMS Community License Program
From the Community License Q & A Forum:
Hello OpenVMS Community,
VMS Software, Inc.’s launch of the OpenVMS Community License program is a great success. The applications for participation exceeded our expectations, which is great evidence of the strength of OpenVMS in the market. We thank the company employees who worked tirelessly to make the program a reality.
The initial launch is version 1.0 of the community program. The VMS Software Community program was not intended to be identical to the HPE Hobbyist program. We anticipated differences between the HPE hobbyist license and VSI community license would generate discussion. Your ideas and feedback are important and valuable. Please provide your feedback here or in the VSI OpenVMS Forum, OpenVMS Community License Program - Q&A forum. We will consider changes to the program based on your feedback.
With the participation of community members, the program will evolve to meet the needs of the worldwide OpenVMS community.
Thank you,
VMS Software Management Team
Dave Sweeney, Chief Customer Advocate
Added in 1 hour 1 minute 38 seconds:
Hello OpenVMS Community,
VMS Software, Inc.’s launch of the OpenVMS Community License program is a great success. The applications for participation exceeded our expectations, which is great evidence of the strength of OpenVMS in the market. We thank the company employees who worked tirelessly to make the program a reality.
The initial launch is version 1.0 of the community program. The VMS Software Community program was not intended to be identical to the HPE Hobbyist program. We anticipated differences between the HPE hobbyist license and VSI community license would generate discussion. Your ideas and feedback are important and valuable. Please provide your feedback here or in the VSI OpenVMS Forum, OpenVMS Community License Program - Q&A forum. We will consider changes to the program based on your feedback.
With the participation of community members, the program will evolve to meet the needs of the worldwide OpenVMS community.
Thank you,
VMS Software Management Team
Dave Sweeney, Chief Customer Advocate
Added in 1 hour 1 minute 38 seconds:
Willem please describe exactly what you are using clusters for in your hobbyist environment so VSI can evaluate your requirements. If you are uncomfortable sharing the information in the forum, please send email to training@vmssoftware.com. As you may have seen in my recent post VSI welcomes constructive feedback on how to improve the program.willemgrooters wrote: ↑Sun Aug 16, 2020 4:34 pmI am VERY VERY disspointed with VSI's Community Licence Program.
Most importantly: the absence of clustering on Itanium.
No matter what reason, it will render current environments useless. I guess that most hobbyist licenees with Itanium run them in a clustered environment, for whatever reason; accepted by HPE since the first publication of the Itanium licenses. Removal renders these environments useless.
I wonder how many members of the community will accept this; how many will accept the (much extra) work to be done to come over it - if possible. Chances are that VSI (and therefore, OpenVMS) will loose very dedicated users/developers/advocates/supporters/free forum- and wiki contributers/free marketeers because they find themselves forced off OpenVMS.
Last edited by madsweeney on Tue Aug 18, 2020 11:00 am, edited 1 time in total.
Dave "MAD" Sweeney
Re: OpenVMS Community License Program
Code: Select all
Thanks for establishing the new Community License Program
(occasionally referred to, still, as "Hobbyist"). I can imagine how
much work was involved, and it's appreciated. Complaints follow.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
VSI_OPENVMS_V842_INSTALLATION_MANUAL.PDF
Obtaining the Distribution Image and Burning a DVD
Page 5:
11.DVD Method: ["11.DVD"? White space?]
Typical command: $ UNZIP -V zipfilename -d ddcu:[directory]
Page 6:
12.Virtual Disk Method: [Again, white space?]
Typical command: $ UNZIP -V zipfilename -d ddcu:[directory]
That "-V" option might be harmless, but, in the words of Inigo
Montoya, "You keep using that word. I do not think it means what you
think it means."
First, unquoted upper-case stuff in DCL (under someone else's
control) is always dangerous. Second, UnZip "-V" affects version
numbers, not file record attributes; and there's no obvious need to
preserve VMS file record attributes on a disk image (ISO-9660, or
other).
Zip "-V" causes Zip to _store_ VMS file attributes:
-V save VMS file attributes
UnZip "-V" is different:
-V retain VMS version numbers
If the VMS file attributes are stored in the archive, then, on a VMS
system, UnZip will restore those attributes unconditionally; no option
is needed for that. (And there's no option to stop it.)
For this stuff, plain-old "UNZIP" should be just fine. Please let me
know if you find evidence to the contrary.
Page 11:
independend
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Installation quirk on an rx2660:
What is the node's SCS node name? ITX
This is an IA64 node. LAN/IP is required for cluster communications.
PEDRIVER will be loaded. Answer YES to the following question to
allow this node to be a cluster member.
Enter this cluster's group number: 666
Enter this cluster's password:
Re-enter this cluster's password for verification:
"YES" seemed inappropriate for "Enter this cluster's group number".
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Post-installation:
%LICENSE-I-LOADED, VSI DFG was successfully loaded with 4 units
but (all?) others:
%LICENSE-W-NCLOAD, VSI OPENVMS-I64-BOE was loaded in noncompliant mode with 2 un
it(s). 4 units required
It looks as if I (as SYSTEM) will be seeing these messages regularly:
You have 1 new Mail message.
Subj: !!! Important: LICENSE COMPLIANCE REPORT !!! Action required.
ITX $ show lice /unit
OpenVMS I64/LMF Charge Information for node ITX
This is an HP rx2660 (1.59GHz/9.0MB), with 4 cores active
This platform supports up to 2 processor socket(s)
Type: PPL, Units Required: 4 (I64 Per Processor)
Type: PCL, Units Required: 4 (I64 Per Core)
It may be difficult to adjust psychologically to my junk being too
fancy for this stuff.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
DFG installation:
$ @SYS$SYSTEM:SETFILENOMOVE.COM
To set the NOMOVE attribute on these files you must boot from"
another system, such as the distribution CD, and then execute"
It seems as if it had been only recently, when I was running the
installation itself, that I _was_ using the distribution _DVD_ as my
system disk. To a casual observer, that seems as if it would have been
an ideal time to deal with this task. Perhaps as an option, but what's
the reason not to do it by default, if not always?
Note, too, all the trailing quotation marks which apparently were
left when some instances of:
$ say "<text>"
were (almost correctly) replaced by:
$ type sys$input
$DECK
<text>"
$EOD
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
%LICENSE-E-NOAUTH, VSI VMSCLUSTER use is not authorized on this node
-LICENSE-F-NOLICENSE, no license is active for this software product
-LICENSE-I-SYSMGR, please see your system manager
My system manager can't help me.
The lack of a VMSCLUSTER PAK will be seriously annoying. I've always
configured my (reasonably recent) part-time systems as no-VOTES cluster
members (with their own system disks), allowing easy access to disks on
my primary system. Much of that still works without the PAK, but some
basic operations fail with complaints like:
VMSCLUSTER license is not active
> [...] The VMS Software Community program was not intended to be
> identical to the HPE Hobbyist program. [...]
I don't need "identical", but less worse would be better.
As a quick example of the added convenience, consider the ease of
accessing a file like VSI_PAK_IA64.COM on another cluster member's
served disk immediately after installing the OS. No need for any
still-un-PAKed/un-configured network software, no need for any USB or
optical storage gizmo/medium, and so on. In fact, that still works,
even without the PAK, but leaving VAXCLUSTER=2 causes the expected
pain. For example:
ITX $ set host 0
I'm ITX. Who are you?
Username: system
Password:
VMSCLUSTER license is not active
%REM-F-NETERR, DECnet channel error on remote terminal link
%REM-S-END, control returned to node LOCAL:.ITX::
%SYSTEM-F-LINKDISCON, network partner disconnected logical link
Another use case is building stuff for multiple architectures in a
single source tree. Shared storage is convenient for many things.
Having to pack-ship-unpack architecture-specific files sucks much of the
joy from the job.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
The installation script says:
The following PATCH kits are present on the OpenVMS I64
distribution media.
Something like an actual location would have been nice. Especially
when VSI-I64VMS-TCPIP_SSH_PAT-V0507-ECO5D-4.PCSI$COMPRESSED was in:
ITX$DNA0:[KITS.TCPIP]
but VSI-I64VMS-VMS842L1I_UPDATE-V0100--4.PCSI$COMPRESSED was in both:
ITX$DNA0:[KITS.VMS842L1I_UPDATE_KIT]
and
ITX$DNA0:[VMS842L1I_UPDATE.KIT]
Better organization would be better.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Interestingly, the new (to me) MMS V4.0 emitted some unexpected
"%MMS-W-GWKACTNOUPD, Actions didn't update" complaints resulting from
case differences in some object-library module dependencies in one of my
projects. My fault, it seems. So, the educational value of access to
less obsolete software has been demonstrated to my satisfaction. (Of
course, I believed it before, but more evidence is good.)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
That may be all for now, but the future's still out there.
Forum software which preserved white space by default would be nice,
too.
-
- Valued Contributor
- Posts: 91
- Joined: Fri Jul 12, 2019 1:59 pm
- Reputation: 0
- Location: Netherlands
- Status: Offline
- Contact:
Re: OpenVMS Community License Program
I doubt I am the only one that will run into "problems" because there is no clustering on Itanium included in the community License.madsweeney wrote: ↑Tue Aug 18, 2020 12:01 pm
Willem please describe exactly what you are using clusters for in your hobbyist environment so VSI can evaluate your requirements. If you are uncomfortable sharing the information in the forum, please send email to training@vmssoftware.com. As you may have seen in my recent post VSI welcomes constructive feedback on how to improve the program.
Developing software to be run on both AXP and IA64, clustering allows me to have one codebase on disk(s) that are accesable by both platforms without copying anything; and I guess most members that develop software on both platforms, work that way. Because the ability is gone, changes on one platform need to be redone on the other; with all possible extra work in compiling, linking and testing (and debugging): it can no longer be garanteed that either platform has the same codebase than the other.
If VSI wants the community to contribute in developing new software on, or porting (Linux) software to openVMS, they should not just supply the software to do so - both required or handy - but also supply the facilities to do it as clean and simple as possible. Not just now, for both AXP and IA64, but - especially - for X86-64.
It is an educated guess, but my feeling is that this is shared by many - if not all - community members that do their development on OpenVMS.
Re: OpenVMS Community License Program
Hello Willem,
May be time to move to modern devop tools like a DCVS.
I have done this, 10 (15?) years ago and I have never had any synchronization problem.
https://about.gitlab.com/why/
The good new, is using mercurial/heptapod VMS developments can be integrated in such environment.
See for example https://foss.vmsgenerations.org/
We expect to have the CI part working with VMS soon.
I know customers who have started to use it for their developments.
Jean-François
May be time to move to modern devop tools like a DCVS.
I have done this, 10 (15?) years ago and I have never had any synchronization problem.
https://about.gitlab.com/why/
The good new, is using mercurial/heptapod VMS developments can be integrated in such environment.
See for example https://foss.vmsgenerations.org/
We expect to have the CI part working with VMS soon.
I know customers who have started to use it for their developments.
Jean-François
Last edited by jfp on Tue Aug 25, 2020 9:03 am, edited 1 time in total.
Re: OpenVMS Community License Program
Metoo.willemgrooters wrote: ↑Sun Aug 23, 2020 6:12 am
I doubt I am the only one that will run into "problems" because there is no clustering on Itanium included in the community License.
I run 2 clusters (AXP & IA64).
So upgrading the AXP will not be a real problem (apart that I will loose my permanent educational licenses).
The Itanium cluster is one with one boot-member and several satellites. One Of them is connected via fiberoptics to a MSA1000 & MSA1500. For me this is a way to use cheap second-hand hardware with a maximum of cpu-power/RAM while still be able to access all available disk-space. This cluster runs on temporarily hobyist licenses. When the cluster was first set-up it used free "school" licenses, which could be run on the university-campus and at homes of employees. I was even allowed to share the license code with anyone employed by my university (Only one license per "school" was issued). Several years ago this educational "school" program was merged into the hobyist program. Since John Egolf said in his E-mails that installation was allowed on more than one machine I never cared.
I really need this system for my "porting opensource to OpenVMS" projects.
Jouk
-
- VSI Expert
- Active Contributor
- Posts: 41
- Joined: Mon Jun 10, 2019 9:23 am
- Reputation: 1
- Status: Offline
Re: OpenVMS Community License Program
We are listening to feedback and will make changes to the program to meet the community needs. Your "problems" are not yours alone as evidenced by other contributors to the forum.willemgrooters wrote: ↑Sun Aug 23, 2020 6:12 amI doubt I am the only one that will run into "problems" because there is no clustering on Itanium included in the community License.madsweeney wrote: ↑Tue Aug 18, 2020 12:01 pm
Willem please describe exactly what you are using clusters for in your hobbyist environment so VSI can evaluate your requirements. If you are uncomfortable sharing the information in the forum, please send email to training@vmssoftware.com. As you may have seen in my recent post VSI welcomes constructive feedback on how to improve the program.
Developing software to be run on both AXP and IA64, clustering allows me to have one codebase on disk(s) that are accesable by both platforms without copying anything; and I guess most members that develop software on both platforms, work that way. Because the ability is gone, changes on one platform need to be redone on the other; with all possible extra work in compiling, linking and testing (and debugging): it can no longer be garanteed that either platform has the same codebase than the other.
If VSI wants the community to contribute in developing new software on, or porting (Linux) software to openVMS, they should not just supply the software to do so - both required or handy - but also supply the facilities to do it as clean and simple as possible. Not just now, for both AXP and IA64, but - especially - for X86-64.
It is an educated guess, but my feeling is that this is shared by many - if not all - community members that do their development on OpenVMS.
Dave "MAD" Sweeney