OS2 World Community Forum

WebSite Information => Article Discussions => Topic started by: Martin Iturbide on December 31, 2019, 08:56:25 pm

Title: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on December 31, 2019, 08:56:25 pm
Hi

I had been wondering what can be do as a community on 2020 to help supporting our platform.
OS2World is going to be 20 years old on 2020 and we usually work on helping each other on the forums, publishing news, trying to consolidate documentation and knowledge (wiki) and in general supporting the OS/2 platform.

But what do you think we should try to do different and/or improve  for the next years?

Any draft, idea, thought is welcome for the brainstorm.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Andreas Schnellbacher on December 31, 2019, 09:57:04 pm
The main answer is simple: Do that, what our main producer, Arca Noae, doesn't and what is valuable for us.

About the details:
Additional thoughts:
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dariusz Piatkowski on January 01, 2020, 04:23:42 pm
Martin, everyone...: Happy New Year to all!

Hmm, excellent question you pose. Here are a few wishes I have (some of this I've voiced in the past as well):

1) Development
- can we pick-up from where EDM/2 left-off?
- I know some effort has been made, but does anyone today really know "who knows what" (from a technical/dev perspective) on our platform?
- Yes, there are a couple of core groups (BWW and AN), but who does what, who knows what?
- Along with this really comes the notion of "can anyone build/write-up instructions to create a robust OS/2 development environment?" we can all use?

2) RPM
- it feels like we would do ourselves a favour by migrating as much of what we currently have to RPMs, so is there a way for all of us to engage in the RPM-Migration effort?
- if this is doable, how do we support that effort with the right platform? I'm thinking an actual central repository of OS2World activities, say os2world-rel & os2world-exp? These should not double up on what's already being done elsewhere, instead these should be anciliary

3) Testing
- a bit of a follow-up to the RPM above, but if you take a look at everything that's sitting out there on netlabs-exp right now you've got a list of about 15 packages I think...many of them have been in the netlabs-exp repo for quite some time...so is that b/c the testing is slow to move these to -rel?
- perhaps we could help with the testing efforts?
- right now I (personally) have absolutely ZERO idea how the current team does the testing and what actually causes a package to be pushed from -exp to -rel status
- this is a lot of manual effort, which means body count matters, and that's precisely where we could all help

-Dariusz

EDIT
====
From a "who done it an how" perspective I'd be happy to sign-up for a "How to configure RSYNC" article. I am currently going through this on my machine, both LOCAL and NETWORKED runs, so I think I should be able to pull some stuff together. Do we just toss this out into OS2World Wiki => HowTo pages (https://www.os2world.com/wiki/index.php/Category:How_To (https://www.os2world.com/wiki/index.php/Category:How_To))?
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 01, 2020, 09:41:38 pm
From a "who done it an how" perspective I'd be happy to sign-up for a "How to configure RSYNC" article. I am currently going through this on my machine, both LOCAL and NETWORKED runs, so I think I should be able to pull some stuff together. Do we just toss this out into OS2World Wiki => HowTo pages (https://www.os2world.com/wiki/index.php/Category:How_To (https://www.os2world.com/wiki/index.php/Category:How_To))?

Hi Dariusz.
Free free to toss it to the "HowTo pages" on the wiki, I will also review it and make any format change so it can be correctly displayed. Let me know if you need any help with the Wiki.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 01, 2020, 10:03:46 pm
Happy New Year !!!

Thanks for sharing your ideas.  I really like some of those, and I would like to help in some.
I would not like to reply on each idea yet, maybe each one require a new thread.

Let's keep brainstorming.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on January 02, 2020, 12:20:16 am
Hi Dariusz, Happy new decade.
For number one, there is some stuff on Netlabs that could be copied/formatted for edm. Porting, RPM usage comes to mind.
What do you mean by a development environment? There's so many depending on what is being developed. One thing that would be nice, if RPM supports it, are meta-packages, so you install the GCC or Unix porting develp meta-package and it pulls in all needed stuff as dependencies.
For number two, seems like a good idea as long as avoiding RPM hell, so RPM's have to be tested and maintained.
For number three, testing is always a good idea. some things are simple and some aren't. For example, the new NSPR and NSS are stuck in limbo as they seem to make the Mozilla apps unstable and the updated NSS breaks some HTTPS sites with old versions like FF 38ESR. Ideally is not just testing but fixing. Silvan seems to be doing most of the work on a volunteer basis and while he has been doing a good job, he only has so much time and energy. Things like the browser just silently vanishing are hard to debug too.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Doug Clark on January 02, 2020, 06:16:33 am
I am coming to appreciate the power of RPM more and more. But what we really have are two "standard" installation techniques that don't talk to each other: WarpIN and RPM. WarpIN in many ways is the the friendlier of the two, but its two level database (OS2 INI files) really isn't powerful enough to handle large applications and packages.  With RPM/APM I don't know how to find out where stuff gets installed, and since so much is automatically installed I have found myself installing something with WarpIN that turned out to be already installed via RPM. And as far as I know there is no uninstall in RPM/APM.

The other issue with RPM is how to package the product and where to place the RPM packages so that they can be found. With WarpIN you just create an archive and upload to Hobbes.

Recognizing that I don't know much at all about RPM or its database, or how realistic this is, perhaps for 2020 we get a modified WarpIN that uses the same database RPM uses for storing info about its applications and packages.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on January 02, 2020, 06:38:44 am
RPM does do uninstalls, yum uninstall package or right click in ANPM and choose uninstall. Can also downgrade much the same way. The problem at times is dependencies, try to uninstall something and RPM will also uninstall its dependencies after asking. Generally if something is available as an RPM and Warpin package, it's likely the RPM one is what you want to use.
There's directions on packaging on Netlabs and it is possible to install an RPM file or create your own repository, not for the faint of heart though.
There is ongoing work to also handle Warpin with ANPM, the beta I'm using is capable of it to a degree. The problem is that there are so many assorted Warpin packages that they'll never all be supported and I believe Arca Noae's plan is to have a Warpin repository for subscribers. The actual databases will likely never be merged.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dariusz Piatkowski on January 02, 2020, 04:20:07 pm
Hi Martin,
Free free to toss it to the "HowTo pages" on the wiki, I will also review it and make any format change so it can be correctly displayed. Let me know if you need any help with the Wiki.
Regards

OK, I will get cracking on that within the next day or so. I do after all want to have a successfull RSYNC run here before I claim to "know it all" and proceed to share my wisdom with the crowd eh? LOL
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dariusz Piatkowski on January 02, 2020, 04:41:32 pm
Hi Dave!

Let me split up the response into multiple sub-sections.

For number one, there is some stuff on Netlabs that could be copied/formatted for edm. Porting, RPM usage comes to mind.

So do we still own/maintain EDM/2? I know the site is still up and chugging away, but it's unclear who actually has the current responsibility, is it netlabs? On the 'Contribute - EDM2' page (http://www.edm2.com/index.php/Contribute (http://www.edm2.com/index.php/Contribute)) you can find the following section:

Code: [Select]
Access to EDM/2 Wiki

You can write to this wiki with a netlabs.org user ID, to get that visit http://www.netlabs.org/site/member/

If anything is wrong please contact ktk@netlabs.org.

...but based on what I found on netlabs.org (following that provided link) I did not see anything specific to EDM/2 contributions.

...
What do you mean by a development environment? There's so many depending on what is being developed. One thing that would be nice, if RPM supports it, are meta-packages, so you install the GCC or Unix porting develp meta-package and it pulls in all needed stuff as dependencies.

Dave, you bet, that was absolutely my thinking. The idea being that we could basicall build a DEV-environment-install-template for each different configuration. So for example: I have VACPP here, along with Watcom and GCC stuff. As I periodically attempt to do things I manage with that type of template based definition which is built around the Visual SlickEdit project functionality. However, that only allows me to deal with PATHs, environment variables, etc...that does not handle the spelling out of what actual toolsets are required.

So in the case of VACPP, or Watcom that's fairly self-explanatory: you install the application and run from there. But GCC...oh boy, that's another thing altogether. I do not mean just the plain vanilla install either. I am talking about all the pieces needed to attempt to migrate anything bigger than the basic Linux type codebase.

...For number two, seems like a good idea as long as avoiding RPM hell, so RPM's have to be tested and maintained...

Well, I completely agree. To make this happen however it feels like we basically need good ol' traditional Project Management. A way to assign different folks to different tasks. Since I have no visibility to how teams like BWW or even AN do this type of transactional detail, I have no idea if there are some best-practice type of approaches we could take. Heck, I'd love for us to be able to come in with OS2World testing 'army' that could take on some of the less complex testing efforts. This is where the beauty of community based engagement would really pay off.

...For number three, testing is always a good idea. some things are simple and some aren't. For example, the new NSPR and NSS are stuck in limbo as they seem to make the Mozilla apps unstable and the updated NSS breaks some HTTPS sites with old versions like FF 38ESR. Ideally is not just testing but fixing. Silvan seems to be doing most of the work on a volunteer basis...

Right, so the 'OS2World' testing army concept is how I see us being able to make a positive contribution in some of these efforts. We do something similar to this already on the AN Tester List, but that's mostly been unit-level testing with some sprinkling of integration testing (this comment is based only on my personal experience with what I have tested - I realize others are much more engaged and do things like full AN OS installs, etc.).
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 02, 2020, 04:59:23 pm
Hi

I want to give a further opinion on some of the post which are very interesting to me.

-
Quote
Sample lists of working and recent hardware. Do we already have more reports than working hardware components?
Yes, I like that a lot.   My doubt still remains "how"?
I had been trying to document hardware running no ArcaOS every time opportunity appears, but creating a wiki page on the OS2World wiki (some example (https://www.os2world.com/wiki/index.php/Laptop_Computers)). When someone post on the forum about a new machine I ask for the pci.exe report to try to enhance the description on the page. Sadly it is not very effective since it is a not formal campaign, i don't have the man power to focus on that, and it is very reactive.

I liked a past approach were eCo used to ask every week for different reports, but I wanted to do something more collaborative (under a Creative Commons license) and where you can add more details. People also tell me that a Wiki is not the best tool to do hardware reports, but it is simple enough for me.

So, the opening for leading the community on hardware reports is open  ;D

-
Quote
Sample lists of utilities that are missing in recent ArcaOS installations.
Good one. Os2World forums is always open for people requesting things or posting different ideas, but I think we lack the consolidation effort on what utilities should be on the top of the wishlist... It sound interesting too.

-
Quote
Also a documentation about...
Documentation on any topic related to ArcaOS (How To, Reviews, Bug workarronds, Tips, Tricks) are always welcome, and this time, the OS2World wiki is a good tool for it. Maybe more efforts are needed on trying to get more people writing and after that communicating the new stuff that can be read.

-
Quote
RPM
I also find interesting try to consolidate the RPM servers with more OS/2 software. I even made some tests to create rpm packages for some games, but since we still do not have a community RPM server (beside Netlabs) I found it discouraging to continue packaging this with RPM.
- I think we need a friendly (more relaxed) rpm server on where to upload packages like games, tools, etc with a simple process on how to upload things to it.
- Also, we need a backup place that replicates and preserve the rpm packages in different places, just to be sure no more files get erased with time.
I think it will be worthy to put more efforts on creating more rpm packages for software that is already available to simplify the installation process.

Very interesting brainstorming, please keep posting any ideas of what the community can help within the platform.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 02, 2020, 05:03:01 pm
So do we still own/maintain EDM/2? I know the site is still up and chugging away, but it's unclear who actually has the current responsibility, is it netlabs? On the 'Contribute - EDM2' page (http://www.edm2.com/index.php/Contribute (http://www.edm2.com/index.php/Contribute))...

Hi
About EDM/2. The final responsibility on it is on Adrian from Netlabs.  I have some administration rights and I work on that wiki weekly trying to consolidate the documentation. But registering new users can only do it Adrian.

Let me know any idea you have for the EDM/2 wiki, I will be happy to provide help/advice/moral support, etc.
There used to be some monthly news I used to post about EDM/2 news and updates, but I have failed to continue announcing that.  Maybe because I reached the level when I had republished and linked as much as I was able to find and now I'm more focus on trying to complete the "OS/2 API Project" (which tries to describe all the functions of the OS/2 components).

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 19, 2020, 05:38:54 pm
Hi

Maybe it is time to sent an open letter to Ginni Rometty (IBM's CEO) to try again to request some OS/2 components to be open sourced.
We can turn it into a habit that every CEO since Lou Gertner should get an Open Source OS/2 related letter.  :D :D

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Doug Bissett on January 19, 2020, 06:56:23 pm
Hi

Maybe it is time to sent an open letter to Ginni Rometty (IBM's CEO) to try again to request some OS/2 components to be open sourced.
We can turn it into a habit that every CEO since Lou Gertner should get an Open Source OS/2 related letter.  :D :D

Regards

Have you ever heard the expression "Don't annoy the lion"? IBM's upper management, today, probably never even heard about OS/2, and annoying them could easily bring it to their attention, and they would likely kill it completely to get rid of the annoyance. Open sourcing OS/2 would be an expensive project, for IBM's lawyers, and IBM isn't interested in spending money that it doesn't have to spend.  Arca Noae has got permission to patch, and distribute patched versions of, IBM software. Not as good as open source, but better than we ever had before. Don't rock the boat.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 19, 2020, 07:34:06 pm
Like always, thanks for your positive words DougB, you are motivating me more to do it now.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: hakuchi on January 20, 2020, 12:08:50 am
For someone who isn't using OS/2 at the moment and you can say is rather new to the system anyways, (Did my brief look into it back in the 90's, but never really run it), but is going to get ArcaOS and install it and go from there I have found a bit hard to figure out software stuffs and such.

I do have killed my braincells with BSD and other Unix as well as toynixes 25+ years so I do have some understanding how things work and can be done very well or very badly. Brief relationships with BeOS and Commodore stuffs and such to mention some.

I really don't know how RPM is implemented in ArcaOS and such. But coming from FOSS world I can see big benefits in such thing. But of course if there is more than one way to do things, like I understood from the earlier posts, it will get messy. It has been tested back in the late 90's in the toynix-world. Anyway as I see it, it would be good, as a community, to achieve a consensus what would be "supported" way to do it and focus efforts to that. If someone want's to do it his/her way, then big boys/girls should take responsibility of their actions. Like there is different methods in use for different distros in toynix-world, but one can still go and break everything by doing stuff non-standard way.

About software one would benefit from the decent categorised list with some more details and comparison to software on other platforms Win/Mac/GNU/etc. than there is in the wiki at the moment. I don't say it's no use, but it's rather hard for someone who has no background to figure out what software would actually be something that one wants to check out. Some of those are nobrainers and need no explanation but there are some that would do with better description. Lists in the wiki are very good but just a bit not-so-new-user-friendly.

For example I tried to find, just now while writing this, Gopher client that I found but now cigar. Tried to search and no. Went through quite a lot, but didn't find it. Was going to point out that that is one of the things that made my day a bit better when I found it last year. And this isn't the first time I have tried to find it after my initial encounter. I came across of it when I just aimlessly wandered around, but since then it's been hiding. EDIT2: And this was it what I was looking for. I just happened to remember that I did bookmark it the second time I was looking for it. https://ecsoft2.org/gopher Could be useful to link these from wiki?

Kinda "suggested programs" list and also split them into "Up-to-date" "obsolete" in order to help new users to find software that they are looking for without going through all of the stuff that are there more or less only because of nostalgia's sake.

All this said, I still want to point out that I shoot out of the dark here. All this might be all good and sound in the ArcaOs at the moment, since I do not know.

These are just my thought of the matter from my quick looks into this stuff.

Anyway, keep up the good work! It's wonderful to see such a dedicated community and sincere will to keep Good Stuff(tm) alive!

H/T

p.s. I will record and document my install on my ThinkPad X60s + UltraBase when I get to that.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on January 20, 2020, 03:45:04 am
Gopher client was discontinued with OS/2 4.5.x, we had a very good client from IBM in Warp V3 and Warp V4 so no one ever bothered writing/porting any other. Now you need an older install to have it and play with ENDLIBPATH to run it as some of the needed DLLs are no longer included. Same with the old Webexplorer, an interesting browser which unluckily is mostly useless in today's web.
Title: Re: Community Ideas BrainStorm for 2020
Post by: hakuchi on January 20, 2020, 05:27:24 am
Gopher client was discontinued with OS/2 4.5.x, we had a very good client from IBM in Warp V3 and Warp V4 so no one ever bothered writing/porting any other. Now you need an older install to have it and play with ENDLIBPATH to run it as some of the needed DLLs are no longer included. Same with the old Webexplorer, an interesting browser which unluckily is mostly useless in today's web.

So you say one can't have that sweet looking GUI Gopher client working with ArcaOS? That's a big bummer I was so looking forward for that one. Oh, well Lynx will do for me then as it's done before and will continue to do so in the future.

I just hope this kind of things would be mentioned on the descriptions of software. I think this might be one good community thing to do, so new users like myself know where we are at the moment and what will works and will not.

If someone would step in and make this Gopher client thingie work, that would be awesome!
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on January 20, 2020, 06:00:54 am
It works fine but you need to get it and some DLLs from an older Warp v3-4 install.
Hmm, today I'm getting TCP/IP error number 10014, whatever that is.
Title: Re: Community Ideas BrainStorm for 2020
Post by: hakuchi on January 20, 2020, 06:13:56 am
It works fine but you need to get it and some DLLs from an older Warp v3-4 install.
Hmm, today I'm getting TCP/IP error number 10014, whatever that is.

Ah. And one can get those DDLs and such easily?

Answer to your question ...

Quote
WSAEFAULT
10014
Bad address.
The system detected an invalid pointer address in attempting to use a pointer argument of a call. This error occurs if an application passes an invalid pointer value, or if the length of the buffer is too small. For instance, if the length of an argument, which is a sockaddr structure, is smaller than the sizeof(sockaddr).
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 21, 2020, 09:21:38 pm
Hi Hakuchi

For someone who isn't using OS/2 at the moment and you can say is rather new to the system anyways, ....

I agree with you that we don't have organized documentation or articles that gives a hint on what software to use on OS/2. Sadly the community does not has as much members or a market that will facilitate to have people writing that kind of material. Also, the OS2World wiki looks more like my scrapbook than a formal wiki were you can find things. I really hope I can evolve the wiki to something more useful.

The best way for the moment is to ask on the forums things like "How I can do XYZ with OS/2 or ArcaOS?" or "What do you recommend me to use for xyx with OS/2?".  I know it is kind of problem for people that like self service but consolidating and generating new OS/2 knowledge is not an easy task.

Maybe this can give some ideas to people that want to collaborate on the community.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: hakuchi on January 21, 2020, 10:49:23 pm
I agree with you that we don't have organized documentation or articles that gives a hint on what software to use on OS/2. Sadly the community does not has as much members or a market that will facilitate to have people writing that kind of material.


I smell something I might be able to help with. As I go along with my install and all that, I'll write stuff up as I learn things. That might be best I can offer for now for the community at the moment. Too much things going on in my life one have to keep "extra" pile as low as possible!

Also, the OS2World wiki looks more like my scrapbook than a formal wiki were you can find things. I really hope I can evolve the wiki to something more useful.

The best way for the moment is to ask on the forums things like "How I can do XYZ with OS/2 or ArcaOS?" or "What do you recommend me to use for xyx with OS/2?".  I know it is kind of problem for people that like self service but consolidating and generating new OS/2 knowledge is not an easy task.

Having used mainly BSD's and such OS's for over couple decades now mailing lists and forums are not strangers to me when need some info about something I don't yet know. So asking Isn't that big problem, but this usually is relevant when there this problem/etc is at hand. Beforehand it's hard to know what types challenges one encounters.

For example this Gopher client thingie, I didn't have any idea that one need some old DLLs (which I don't know what they are and where to get them, still) in order to get it to work.

So I'll make note to myself to document my journey as I go down the OS/2 path in near future. Maybe next month I get to this. I almost got around to get ArcaOS already, but then life happened and I had to postpone it. I don't want to purchase my license before I really have time to spend on the matter.

Your work on wiki is very good and I do know how hard it is to maintain even "scrapbook" level of structure ... kudos for doing that!

Everyone keep up the good work!
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on January 21, 2020, 11:45:15 pm
I got the gopher client working here. The missing DLL seems to be SETLOC1.DLL, might need more if not a en locale. This was part of the Warp V4 install. Later when you need the exe and dll, ask.
Title: Re: Community Ideas BrainStorm for 2020
Post by: hakuchi on January 22, 2020, 08:43:18 am
I got the gopher client working here. The missing DLL seems to be SETLOC1.DLL, might need more if not a en locale. This was part of the Warp V4 install. Later when you need the exe and dll, ask.

This is wonderful news! Thanks for the heads up!

Sure I'll get back to you when I get into the install and tinkering!
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on January 31, 2020, 02:46:58 am
Maybe it is time to sent an open letter to Ginni Rometty (IBM's CEO) ....

Too late, Ginni was replaced (https://www.bloomberg.com/news/articles/2020-01-30/ibm-names-arvind-krishna-as-ceo-rometty-to-retire-at-year-s-end)   ;D
 
Any more ideas for 2020 ?

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Devid on February 01, 2020, 09:36:29 pm
Hi.
I've been using slackware for a long time so I am used to compiling and patching software by myself even though I'm not a programmer.
I think it would be useful a dedicated how-to section on how to install development environments: va3, va4, gcc, wc, ddk.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on February 01, 2020, 10:10:47 pm
Hi Devid.

That is a great idea.  I already have something about installing "IBM C and C++ Compilers, Version 3.6.5" on this EDM/2 article (http://www.edm2.com/index.php/Compiling_HWMan_WPS_Class_on_ArcaOS_5.x). But a lot more is required.

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: tthoms on February 05, 2020, 03:38:29 am
     I've had OS/2 on my mind quite a bit.  I have to say, it's always been difficult to work with.  I remember spending countless hours tweaking settings, hunting for drivers, fixes, and usable programs.  It was such a pain.
     I stuck with it, because once it worked, it worked well, and the multitasking was unbeatable. To this, it felt great to not be part of the Windows crowd.
     I stopped using OS/2 once XP came out.  It, for the most part, did everything OS/2 did with a lot less work.  IBM had long abandoned it's child, and I needed something that would continue to grow.
Title: Re: Community Ideas BrainStorm for 2020
Post by: tthoms on February 05, 2020, 03:53:48 am
(accidentally posted, this is a continuation of my previous post...)
Now with virtualization, I've been able to walk through the halls of yesteryear with OS/2, and while I remember it fondly, its still a pain.
     I guess the point of my addition to the brainstorm is that I can't see where the OS/2 experience can go.  As long as no one has access to the main code, nor cares to renovate it, it is not sustainable.  Since IBM likes to keep OS/2 in the rearview mirror, our only option is do do something like Haiku has done for the BEOS.

I've heard of OS/4, but I honestly don't understand what it is.  Is it a kernal replacement?  That would be a good start.  But I'm also wondering out loud what we'll be able to offer that other OS projects can't.  And if we do have a clear goal for OS/2, how can guys like me who can't code help us get there?

Thanks for letting me chime in. 
Title: Re: Community Ideas BrainStorm for 2020
Post by: OS4User on February 05, 2020, 06:49:47 am
I've heard of OS/4, but I honestly don't understand what it is.  Is it a kernal replacement?

Here is some info. https://www.os2world.com/wiki/index.php/Phoenix_OS/4
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on February 05, 2020, 04:25:52 pm
Hi Travis

....Since IBM likes to keep OS/2 in the rearview mirror, our only option is do do something like Haiku has done for the BEOS.

I could not agree more and it had been an ongoing discussion on the past.

I saw BEOS and ReactOS progress and I also dream about that for our community. Finally ditching IBM and embracing open source will finally reduce the risk of getting dump again by another vendor and will allow everybody to download, modify and improve OS/2 (or the OS/2 clone) for profit and non-profit, just like Linux OS is doing today.

But sadly we don't have enough resources like money, people and skills (or even a business case) to pull a full OS/2 open source clone. Resources are basically set on Arca Noae creating drivers to allow OS/2 run on today's hardware. But there is not a longer term strategy to became an open platform and try to clone it to have a high level of compatibility with native OS/2 software.

My personal opinion about OS/4, is that is a research, educational and non-profit project, but it is not open source. The people behind OS/4 has very good skill level understanding the OS/2 kernel, but they do not believe in open source software and copyleft software. They think that if they open source OS/4, they will allow other people make money with it. I see it in a different way, selecting a copyleft license (like GNU GPL) will force the big corporate developer to share the source code with the little developer and both will work with in the same ground.

What about a different crazy dream... what if HaikuOS, ReactOS and an OS/2 clone can share the same microkernel someday and share drivers between them? too crazy :)

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Sandra Asja Eickel on February 05, 2020, 08:26:27 pm
But sadly we don't have enough resources like money, people and skills (or even a business case) to pull a full OS/2 open source clone. Resources are basically set on Arca Noae creating drivers to allow OS/2 run on today's hardware. But there is not a longer term strategy to became an open platform and try to clone it to have a high level of compatibility with native OS/2 software.

Don't forget Bitwiseworks http://www.bitwiseworks.com/ (http://www.bitwiseworks.com/) - without them we wouldn't have much current software, like OpenOffice 4.1.7, our older FireFox and now QT 5.13 in beta (and GCC 9.2 together with a lot of new or refreshed tools to be able to continue porting Linux stuff - see https://github.com/bitwiseworks (https://github.com/bitwiseworks)) with the goal of a new chromium based browser in the not too far future.

They still need a lot of donations to pay their developer to finish this QT5/Browser project (see the donation meter at the bottom of their homepage - 0.5% of the estimated amount is from me, sadly I couldn't afford more). Most of their software is free, but they ask for the OpenOffice for OS/2 binary money (this more or less allows for keeping the port current, but doesn't generate otherwise income to finance other tasks). All money (except the Austrian VAT) goes into development, the owners have other jobs to generate income for their living.

And unlike some other companies in the past, Arca Noae and Bitwiseworks DO deliver, even if the progress is not as fast as we all wish, but with tight budgets and very few developers, their output and progress is astonishing. Who would have thought it to be possible to run our old Operating System on UEFI systems without BIOS compatibility, /2?

See https://www.arcanoae.com/what-arca-noaes-upcoming-uefi-support-means-for-arcaos/ (https://www.arcanoae.com/what-arca-noaes-upcoming-uefi-support-means-for-arcaos/) - I hope that USB 3 (XHCI) support will make it into 5.0.5, because that is even more pressing (USB 2.0 ports on modern board are in reality speed-crippled USB 3 ports with enhanced compatibility for problematic USB devices, not real EHCI ports which would work with our older drivers). Both features work on at least some of the tested environments.

Greetings,
Sandra-Asja
Title: Re: Community Ideas BrainStorm for 2020
Post by: David Graser on February 05, 2020, 10:54:06 pm
But sadly we don't have enough resources like money, people and skills (or even a business case) to pull a full OS/2 open source clone. Resources are basically set on Arca Noae creating drivers to allow OS/2 run on today's hardware. But there is not a longer term strategy to became an open platform and try to clone it to have a high level of compatibility with native OS/2 software.

They still need a lot of donations to pay their developer to finish this QT5/Browser project (see the donation meter at the bottom of their homepage - 0.5% of the estimated amount is from me, sadly I couldn't afford more).

Hi Sandra

Five percent is good.  I thought I was doing good at 2-1/2 percent of the $48,000.  My plan was to donate a $100 a month.  Unfortunately, my plant blew up right before Thanksgiving turning my life upside down.  I am still waiting to see what the company does about my truck which is still in the plant.  I still have a job, but my income has been reduced by half.  I have started donating again, but may not be able to donate every month.  I am grateful for Bitwise continued support.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on February 06, 2020, 12:52:26 pm
I think it is time for a good old fashion summary.
These are the ideas on what the OS/2 community can do to support and help with the platform on 2020

1) Support Arca Noae and Bitwise works.
2) More reports and list of supported hardware.
3) Articles, reviews and reports of useful OS/2-ArcaOS applications. Like "Suggested Programs for ArcaOS" article.
4) More documentation on installing ArcaOS (USB stick) and features of every new release.
5) Keep improving EDM/2. More software development related articles.
6) More Instructions, articles on how to build a development environment.
7) RPM installers for OS/2-ArcaOS software. Maybe a campaign to also distribute OS/2 software on RPM packages on available or new servers for the community.
8 ) Supporting any open source cloning of OS/2 close source components.
9) Porting open source software to OS/2-ArcaOS.

I just included 9, since it was kind of obvious.   :)

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: Sandra Asja Eickel on February 06, 2020, 09:20:48 pm
Hi David!

[Bitwiseworks]
They still need a lot of donations to pay their developer to finish this QT5/Browser project (see the donation meter at the bottom of their homepage - 0.5% of the estimated amount is from me, sadly I couldn't afford more).

Five percent is good.

I thought I was doing good [...]

Five percent would be REALLY good (that's about a month of development time). It's five per mil, a tenth of that, so you are doing good, even when you had to slow down in the last few months. Sorry about your truck and the companys plant ...

It really wish, that a lot more people would donate regularly (or could spend a larger sum once), because Dmitriy needs to be paid for his full time development job. We can't afford him to have to look for another job to finance his family, as that would drastically reduce the amount of time he could spend on developing for us.

And with the new GCC 9.2 port including many things in the toolchain, porting (or maintaining) other tools and applications has become a lot easier, because now it's "only" needed to fiddle with OS/2-specific things instead of having to do a lot of hacks to get the newer sources to compile at all with the old version 4 of GCC.
A compiler is not an end user tool like OpenOffice, but it's essential for providing such applications - you can buy a sponsor unit dedicated for GCC if you want.
And a library like QT5 is also not an end user tool, but a lot of applications are built using it, so it's essential, too.

I'll go to miss SeaMonkey in the future, but even on other platforms it massivly lags behind FireFox, as there are so many breaking changes coming in ever faster release cycles, without huge manpower it simply wasn't feasible to try to keep somehow up with them.
To be able to provide a new browser for OS/2 without so much hassle, a webkit/chromium based browser like falcon became the new target. As this required a current QT5 (we had QT4) and also could provide us with a lot more current end user software, the port was started.

For an IT project it is good in time, but there were several other things to do beside just porting QT5 itself, like porting current versions of GCC and other tools, as they are prerequisites for being able to finish QT5. More time means more money is needed, best in form of a steady flow.

So I absoluetly agree with Martins #1: Support Arca Noae and Bitwise works.


With the upcoming USB 3 support (and later UEFI) from Arca Noae, a lot more modern hardware can be used, so there will be enough stuff to add or update Martins #2: More reports and list of supported hardware.

I like #6: More Instructions, articles on how to build a development environment.
It's probably never been easier to turn an end user system into a development machine.
Install all those development stuff via ANPM onto your PC.
Do some magic ...
And with luck, it is just a git clone and a make ...
That would be a dream and #9: Porting open source software to OS/2-ArcaOS.

Sandra-Asja
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on February 06, 2020, 10:10:50 pm
What would be nice is meta-packages, which I see on deb based systems. Could have a gcc-devel-meta package that consists of a bunch of dependencies which when installed would pull in the whole development environment.
Title: Re: Community Ideas BrainStorm for 2020
Post by: tthoms on February 06, 2020, 11:50:01 pm
But sadly we don't have enough resources like money, people and skills (or even a business case) to pull a full OS/2 open source clone. Resources are basically set on Arca Noae creating drivers to allow OS/2 run on today's hardware. But there is not a longer term strategy to became an open platform and try to clone it to have a high level of compatibility with native OS/2 software.

I understand that the lack of resources is a real obstacle, and I think the fact that we can't make a business case for OS/2 at this point says a lot.  If we hope for growth I think the OS/2 community has to decide what niche OS/2 is going to fill, and how it is going to be better (or at least more enjoyable) than the alternatives.

Random thoughts:
How does the Amiga crowd do it?  I was watching some footage of a conference they had, and there was a surprising amount of young faces there, considering how old the platform is.

I am in awe of the dedication that the programming efforts currently ongoing to make OS/2 useful, and I think the idea to move a lot of the current software to yum is a great idea, as are the efforts to solidify the development packages for the platform.  Thank you all!

If I was to advocate for conversion of the os to an opensource format, I'd do it the same way the haiku folks did it.  Start with the Desktop.  Tracker was rebuilt as Open tracker.  Heaven knows the WPS could use some updates.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Martin Iturbide on February 07, 2020, 01:24:58 am
Random thoughts:
How does the Amiga crowd do it?  I was watching some footage of a conference they had, and there was a surprising amount of young faces there, considering how old the platform is.

I also think it is good to make friends with other communities, like Amiga, Qt communities, or even general open source development communities. Like when Roderick went to REXX LA to find a common ground between the two communities (Since in OS/2 we use REXX a lot). Maybe we should always try to get a little space to talk on different communities on how we use OS/2 and ArcaOS.


If I was to advocate for conversion of the os to an opensource format, I'd do it the same way the haiku folks did it.  Start with the Desktop.  Tracker was rebuilt as Open tracker.  Heaven knows the WPS could use some updates.

I like both approaches, if someone want to make a kernel and try it with the IBM's close source compoments on top (like PM, SOM, WPS) is great (Like from bottom to top). Or the other approach, trying to clone things from the WPS and going down (something like a top to bottom approach). Both ways are good to me while they produce open source things.

Maybe it can apply some crazy idea to try evolve XWorkplace into a full WPS class replacements instead of a child of IBM's WPS classes. (or at least go little by little with that goal in mind)

Regards
Title: Re: Community Ideas BrainStorm for 2020
Post by: David Kiley on February 14, 2020, 07:14:12 am
Honestly, I think emulation is probably the most viable future for os/2.
There is only so much hardware that a company can make compatible and it takes a lot of resources - not to mention having to constantly take software like firefox to make a new os/2 version.
Running OS/2 in a virtual machine on a linux host seems best to me - since linux is stable and has very good native support for things like Java and web browsers, (not reinventing the wheel) but I can still enjoy the OS/2 WPS and my favorite OS/2 native/Dos/ and Windows 3.1 programs.

For example I use a java program called MuCommander - it runs flawlessly without bugs on my version of Suse, whereas the last time I tried it on the OS/2 Java it would crash constantly.
Doesn't make sense to put more work into an inferior solution when you can just use all the work already done by the large linux community to make the best java experience.

I think Arca Noae should be offering a VirtualBox image with the os pre-installed as a new product.
VirtualBox has a very intuitive import/export system - so they could sell a  pre-made VirtualBox image with the optimal settings preset.
And the focus of development should be working with future versions of VirtualBox on multiple platforms instead of reinventing the wheel on software like firefox, java, openoffice etc.
A bunch of current effort used to make separate os/2 "open source" apps could instead be put directly contributing to virtualbox opensource project - but a sub community dedicated to just OS/2 compatibility.
Those same resources could be, for example, put into making the best future os/2 virtualbox additions to support things like shared folders, usb, video drivers, sound etc - to be most compatible to the latest ArcaOS .

It would also fit well with the "enterprise" market - since if a company has old OS/2 software that would be expensive to re-purpose - they can run a virtual machine on any modern hardware and they can also move the VirtualBox image to any other future host hardware that comes out (effectively becoming a forever OS) since you can just import and export at will on any new host OS. Therefor ArcoOS could focus more resources on enhancing the experience of the OS with new features instead of patching old software to work on new hardware.

I'm not trying to cast shade on any work Arca Noae has done - just the way I see the best path forward for os/2.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Dave Yeo on February 14, 2020, 07:56:11 am
My understanding is that a lot of the Enterprise customers, or at least the ones who haven't already moved to a virtual system, are tied to real hardware and they're the main customers of ArcaOS rather then us regular users.
Title: Re: Community Ideas BrainStorm for 2020
Post by: David Kiley on February 14, 2020, 08:03:37 am
My understanding is that a lot of the Enterprise customers, or at least the ones who haven't already moved to a virtual system, are tied to real hardware and they're the main customers of ArcaOS rather then us regular users.
There's no reason Arca Noae couldn't instead sell them on new hardware with a stable enterprise host OS installed and a virtual machine instead. They could make money on three fronts - selling the hardware, reselling an enterprise linux like Suse Enterprise Desktop, and selling the ArcaOS license. More profit for them since they would have more products to sell and more resources were it would count for the best ArcaOS.
Title: Re: Community Ideas BrainStorm for 2020
Post by: David Kiley on February 14, 2020, 08:30:21 am
Although I do have to admit maybe some enterprise companies need it to run natively - like if they integrate with specialized hardware or something.
Title: Re: Community Ideas BrainStorm for 2020
Post by: Mathias on February 15, 2020, 11:32:12 pm
True indeed. ; ) Getting hardware connected is hard enough on today`s hardware (in the lack of hardware connectors... like RS232 or Centronics connectors onboard (or even hard to get find expansion cards for PCIexpress).. but connecting these hardware into a virtual machine can be quite a challenge.

First of all the hardware itself needs to be connectable via bare metal, then needs to be forward`ble into such a virtual machine.. and that in a way the virtualised software can understand the input.

Currently our electrican has the problem that his old printer died.. physically. Now he could of course virtualise his computer and try to get a new USB printer and ``somehow`` via USB forward connect this new printer into a virtual machine... then somehow translate and inject this in the virtual LPT1... but WHO could possibly do such a thing... not an electican. And I also do not know how to solve this.

Or of course he could try to remain on bare metal.. and get a used old school printer (maybe even the same model as his old one)... but as time goes on, even these might show first sings of failure.. atleast this is not a solution that works forever, as the Centronics port already died many years ago.. and hardware makers do not implement that stuff anymore.. nor drivers for his old DOS 6.22/Win 3.1 machine.
So I thought, he could maybe change to ArcaOS, since OS/ 2 has better DOS than DOS, and a Windows 3.1 also! :D  -But  still.. not sure how DOS software would talk to today`s USB printers in ArcaOS... so maybe(?) also not a solution for him..

So.. maybe brainstorming for what`s needed in 2020... are translation drivers (wrappers?) to translate between the old and the new world.. sound cards, printers, scanners, etcetc..
For retro gaming I`d wish for something for OS/2 DOS sessions to (software based) fake an SB 16 or AWE32 (used to be ISA cards) and translate the stuff to the actual installed audio driver which connects to the actual sound chip. Luckily Lewis said, such a thing is already in development... atleast for audio..