OS/2 - Possibilities (Use Cases)

[JTA]

Many forum postings have subjects like "need this/that utility or program to ..." and the topic might be something like "backup". This leads to discussions of "what works under OS/2", "has this app been ported", etc.

I propose that the use cases & associated problems get defined somewhat, as in the following "backup" example. When that's done for each use case / scenario, and everyone's suggestions are posted, it might become easier to see what works and is available. Here's an example:

Use Case: Backup
  - scenario: OS/2 is primary OS on physical box
    - data lives on a (JFS, other) volume
    - backup: native tools (only, in most cases); options might be:
      - any currently functioning OS/2 backup program
      - rsync utilities (send the data somewhere)
      - rexx/scripting (script the backup w/ associated native utilities/apps)
      - imaging: at machine level (DFSee, others)
  - scenario: OS/2 is in a VM
    - data could still live inside the os/2 volume. or ...
    - could also live on an external volume of Host-OS, and shared to VM
    - backup:
      - all of the native OS/2 solutions  (which can access shared folder)
      - all solutions available to the Host-OS world (depending on choice of Host-OS)
        - any Windows solutions
        - any Linux solutions
     - multitude of free/paid solutions in both camps
     - options to dedupe, as data lives on volume accessible to these methods
     - imaging: backup the container/VM (or perhapsjust the .vhd's or .vmdk's)

The idea is that, once in this use case format, the sets of solutions are more visible in either scenario, and anyone (new or otherwise) who stumbles upon the posting can instantly see available options in both scenario camps. If you've already solved it and moved on, your knowledge & solution will help everyone else (use cases AND solutions).

As others work out even more possibilities, the posting can be updated with the info (or new posting started, hopefully referencing the old posting).

Other possibilities I've recently seen: Multi-User, Compiler(s), Browser(s), Networking (to something or to services), Filesystems ... but by no means a complete list of nifty topics.

Please kick off other use cases that seem interesting or had vexed you previously ...

[JTA]

Looking at the above "backup" use case, there seem to be more possibilities if OS/2 is virtualized, vs running physical-only.

If OS/2 is a vm, and if the dataset can live on the "shared folders" feature of virtualization, then for example, I can run 64-bit software against the dataset.

I could "export" the dataset from OS/2, and work with it then, but I'm leaning towards solutions where no one OS really owns the dataset, they only get to access it; the Host-OS owns the dataset and manages it (not unlike networking, but with less overhead).

In the Windows world, MS has an unfortunate concept of "my folders", and it was set up on the C drive (OS + apps). This gets messy over time, and ultimately "pollutes" both image backups (with user data) and data backups (with system data); when you tried to change things, profiles sometimes went south. It turns out we can separate the two by implementing d:\home\user, some registry changes, and then c:\ is used only with OS+apps.

Unix/linux always had separate OS/apps/userdata, so it was clean(er) from the start.

With OS/2, you can do things simarly with multiple drives/partitions, but then nothing external gets at that dataset. This possibly becomes a use case where OS/2 virtual is better than OS/2 physical (given decent virtualization layers), wrt backup mngmnt.

As an aside, I've not yet had any grief from any of the virtualized device drivers (guest tools implementation) with either ArcaOS or Warp 4.52) ... they are "good enough" for most everything I'm doing.

[Martin Iturbide]

Hi JTA.

I don't have any more "use cases", I understand that this idea is with the goal to have a better way to communicate, so we can understand faster the environment of the user and try to help him faster.

I had experienced some issues, for example, when someone post that something is not working on the browser and I don't even get which browser is he/she using.

I can not force community members to have more structure on their request (of course), but I had found very useful to ask them to publish the "testlog" log file. Have you seen that kind of log? But this works when the having an issue with OS/2, instead of "how to" information requests.

Regards

[JTA]

If an end-user is having difficulties *using* something, then a log might give *someone* more problem-resolution info, and ultimately help the user base resolve the issue.

If an end-user is having difficulty in understanding *how to use this thing* effectively, then a use case *from* someone might give the user base more info on ways to use it, and ultimately help them decide that this thing is useful!

More use cases in these forums, or the wiki, or wherever ... might spark more interest in, and more possible ways to get things done with OS/2.

I did see the 88watts site (& the testlog utility), as AN had it linked off of one of their dev pages (the 88watt dev kits) ...