OS2 World Community Forum

OS/2, eCS & ArcaOS - Technical => Internet => Topic started by: David Kiley on January 04, 2024, 10:12:18 am

Title: Arca Noae should be directly involved in web browser development efforts
Post by: David Kiley on January 04, 2024, 10:12:18 am
I think Arca Noae should be prioritizing the browser over hardware support at this point and should be directly involved in the efforts to develop a functional browser either with development work or funding..
Users of OS/2 can live with slightly or even much older hardware for years but an OS is dead in the water without a functional browser.

I could see a purpose of paying to  upgrade ArcaOS if they did this but otherwise I kind of don't.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Roderick Klein on January 06, 2024, 03:49:53 pm
I think Arca Noae should be prioritizing the browser over hardware support at this point and should be directly involved in the efforts to develop a functional browser either with development work or funding..
Users of OS/2 can live with slightly or even much older hardware for years but an OS is dead in the water without a functional browser.

I could see a purpose of paying to  upgrade ArcaOS if they did this but otherwise I kind of don't.

Like in the Mensys days the community does a portion of the OS development. If AN wants todo browser development is part two. Send an email to AN about this. I know they do not read the forums allways.

BWW and OS/2 VOICE (me) started working on the QT web browser. This project has proven to be (despite making an estimate) a very big pile of work. But that is what you get with one FTE.

The primary reason that the browser was never finished is because Dmitry is from Russia. To put it simple in a recent newspaper interview with Norwegian newspaper he compares Russia to facits state. Dmitry was in 2014 involved with starting the independed  the independed newswebsite https://media.zona. Which has been declared a foreign agent (goverment sanctions). At this moment Dmitry has left Russia also because he does not support Putin his genocidal war in Ukraine and because of his activity with independed press. Dmitry has for 2 years not found a new country to stay. Effectively he moved to Turkey and is now in one of the Baltic states. He will most likely never be able to go back to Russia as long as the country does not turn into a democratic direction.

I just explain this above for people who did not see my Warpstock presentation I did in Orlando in November last year.
I am currently working on an updated newsletter about the browser development. VOICE has not forgotten about the browser, The panful issue is that without this stiupid war we would have had a browser.

Roderick Klein
OS/2 VOICE
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Sergey Posokhov on January 07, 2024, 11:48:46 am
Quote
as long as the country does not turn into a democratic direction
Nooooo, please, never!!!  ;)
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Lars on January 07, 2024, 01:32:36 pm
I have never understood why Russia never managed to make more out of their abilities and possibilities.

They have capable sports guys, they have capable and talented artists (some of the world best pianists among them), they have smart people (I have a big maths book written by a Russian mathematician in the 1950's or so), they have resources/raw materials and they have money.
But the result is just plain meagre.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Davorin on January 07, 2024, 01:43:37 pm
The primary reason that the browser was never finished is because Dmitry is from Russia. To put it simple in a recent newspaper interview with Norwegian newspaper he compares Russia to facits state.
Bandera is a democrat and Putin is a fascist??!
" O tempora, o mores"
:'(

We definitely need a functional browser and I am ready to financially help the new team again, which will not run away before the job is done.

Regards
Davorin
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Andrew Stephenson on January 07, 2024, 02:48:18 pm
(Photo of sign board in snow:)  Er, sub-titles?  (Please.)   ???
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: ivan on January 07, 2024, 05:05:30 pm
My answer to the problem, since I will not run a browser without NoScript addon, is to have a second monitor connected to one of my Linux computers and run firefox esr on that.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: mauro on January 07, 2024, 05:43:02 pm

I just explain this above for people who did not see my Warpstock presentation I did in Orlando in November last year.
I am currently working on an updated newsletter about the browser development. VOICE has not forgotten about the browser, The panful issue is that without this stiupid war we would have had a browser.

Roderick Klein
OS/2 VOICE

Hi Roderick full respect for those ones who experience a hard period because of....we all know what/who.

But this post has "Arca Noae" as subject. It is a profit Company who sells a pc operating system at 139€ as personal edition including support for the first 6 months, then pay 70€ if you need support the following year, not talking about the major costs for the commercial license.  This in the situation that "the community does a portion of the OS development"???  And this portion of development includes the realization of an actual browser. In both purchase cases (139 € personal, 249 € commecial) you will get an operating system "running on modern multi-core and multi-processor systems, as well as traditional BIOS or UEFI-based hardware" that can offer you nothing more than Firefox38/Seamonkey2.41 browsers, released in 2015, two years old by the year AN took in charge the OS2 development for selling purpose.
They did a lot, true, but telling that a modern/actual/reliable browser realization (or at least ported from an existing one) is not part of their tasks, is too much.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Sigurd Fastenrath on January 07, 2024, 06:41:41 pm
Arca Noae has the right to pick or choose the things, they want to develoep or make their target. That's true and OK.

Regarding the browser the "American/English" users may discover the feelings for waiting for years for something as Users from other languages, whom are waiting for their local release.

I.e.:

- the german Version was announced some years ago (Yes, allways with the hint that it "may" come and depends on the Translators)

Result today: still not available, should come with 5.1.1.



- the WLAN Drivers, have been "focused" or "considered" for at least 5 years now

Result today: no WLAN drivers available, no plan for a further developement in this direction (if I understand Warpstock 2023 correct).


Yes, you can say we have UEFI, GPT support and USB 3.0, that is what industrial customers need. And one can install XWP 1.0.16 with language support.

But for me, the result is the same: if I want to use OS/2 it makes definitly more sense to use a Virtualisation, where I can use a native language version, WLAN of the host, no RAM problem with a maximum of 4 GB usable, and support for the screen resolutions, the Laptop offers.

In this situation there is no need for a new browser as one can simply use the browser of the host OS,

And in addition, as a logical consequence, there is no need for another or newer Version of OS/2 than at least 4.52. There is no need for ArcaOS.

And so: it makes (sadly) no sense for me to spend more money for a modern browser nor to pursuade someone to do so.



Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: David Kiley on April 26, 2025, 10:57:37 am
Sorry for my late reply. I've been out of the loop for a while. Anyways I just wanted to say that I totally didn't mean any disrespect regarding the community efforts and understand the challenges there.

I just think, for me to be motivated to pay for ArcaOS I feel they should be funding or developing the browser too. I'm pretty much in the same place as Sigurd.. that right now with emulation being the best path there (for me) there really isn't any point to donating more money for a new version of ArcaOS.
They make a lot of their money from enterprise customers who probably don't care that much about a browser, so maybe it's not their priority. I know when serenity systems was involved with eComstation they pretty much only cared about enterprise customers so it might be similar.

Perhaps I should consider donating to the browser directly though. It would make more sense for me to pay the community instead of paying to upgrade ArcaOS.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Paul Smedley on April 26, 2025, 11:30:24 am
I donn't mean to be negative about the prospects of getting a working browser, but in the 2+ years since I first got Qt6webengine compiling, to the best of my knowledge, despite all code being published in githib, nobody except me has compiled the webengine.

A few others have built qt6base and some of the other dependencies, and have contributed patches, for which I'm grateful.

I've had some success getting it compiling, and through tenacity, have fixed some stuff, but the reality is that I'n *NOT* a programmer. I know literally none of the OS/2 API, and only have very basic knowledge of C/C++. I can hack existing code and maybe get it working, but when it comes to complex code and interpreting crashes, I'm truly not that good.

What I'm trying to say is that right now, I hold out little prospect of us getting a newer browser than the one based on Qt 6.2.6.

The other reality is, that even if we can somehow get a newer version, the 4GB memory limit is a REAL issue.....
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Martin Iturbide on April 26, 2025, 05:11:42 pm
Hello

I also agreed that Arca Noae should bring a developer to the scene to finally got a modern web browser on ArcaOS. Arca Noae is the only one with resources that can help with this, the rest of the community does not generate any incomes to be directly invest on the OS.

I applaud Paul's effort, but as he says, we need someone with understanding of the OS/2 API that can connect the webengine. Even if it is younger developer that does not knows OS/2, at least hire him permanently to learn the OS/2 API in a few months.

Arca Noae strategy at first was to support OS/2 on modern hardware, but it has to change and evolve to include key software development to help people use the OS for their everyday activities.

Regards
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Remy on April 26, 2025, 06:58:46 pm
Hi !

I agree, Paul did a very big and great job on QT6, webengine and the browser is very near to be usable all day.
I'm retired with some health problems and have missing knowledge in the language used for QT6 and OS2 API   

I'm sure that the free low shared memory issue, some silent crashes can be quickly resolved by a developer from ArcaNoae.
May be some issue could be connected with usb or audio drivers (interrupt handing?) - it mostly occurs when moving my usb mouse pointer or viewing youtube videos.   
   
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Dariusz Piatkowski on April 26, 2025, 08:39:28 pm
Martin, everyone...

...I also agreed that Arca Noae should bring a developer to the scene to finally got a modern web browser on ArcaOS. Arca Noae is the only one with resources that can help with this, the rest of the community does not generate any incomes to be directly invest on the OS...Even if it is younger developer that does not knows OS/2, at least hire him permanently to learn the OS/2 API in a few months...

Umm...folks, we are a DEAD platform...I hate to say it, but that's just the pragmatic part of me screaming in my head!!! LOL

So Martin, as much as I would love to contribute ($$$) again to this effort, the reality is that you can't simply go fishing for a resource, especially someone younger that needs to do a WHOLE STACK worth of learning only to develop a skill that's basically completely unusable anywhere else.

Yup, that's that.

The problem I have always seen as the proverbial "brick wall" of our platform is the fact that we seemingly still do not hava a ready 'out of the box' DEV enviornment. This is a significant mountain to climb for anyone that even attempts to do this.

Look, I've been professionally employed writing code (albeit Oracle PL/SQL DBMS stuff, and some other stuff) but the point is, that's sort of like learning how to ride a bike, and once you figure that out it's just a matter of getting used to different gear sets, handling, etc.

I will tell you this: having tried to upgrade our DISKIO utility (and that thing is a tiny, itsy-bitsy piece of functionality) it's been nearly impossible to find all the pertinent references that allow us to account for how things have changed between the '90s where OS/2 was still a viable platform to how things are today.

Don't get me wrong, I'm not a coding Pro, but I'm no newbie when it comes to DEV environments, IDEs, libraries, etc...so why does it always feel like getting anything off the ground on our platform is like climbing Mt. Everest???  :o ::)
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Dave Yeo on April 27, 2025, 01:24:16 am
As pointed out, we're running an ancient system in PC terms, 20 years since IBM dropped all support, 30 years since most active development was stopped, 35 years since OS/2 2.x was basically designed.
We have a mickey mouse tool chain, Paul has done an excellent job of keeping our GCC fork up to date, but it is a fork and parts of it are ancient. Aout object format as an example with no one having the skills and interest in updating to a more modern object format such as ELF or even what Windows currently uses.
We're also missing other tools to lint our code to find obvious bugs like perhaps a missing free(), and those tools probably would require more then 4GB's of ram, especially to check the Qtwebengine.
Browsers are supported by large organizations, Google, Mozilla and Apple are about it. Even MS switched to being based on Google's work. Development happens fast and just keeping up would take a team of people. We saw with Mozilla, one person trying to keep up leads to burn out.
Our platform is missing modern features. Getting IPC working has proved to be hard, we have to run the browsers in --single-process mode, which is unsupported and may be the source of memory leaks. Seems the problem is our TCPIP stack just craps out on local socket operations that are too big.
For a developer to do much, they would have to be a guru, an OS/2 guru, of which there is a severe shortage of. And they would have to be full time, perhaps with a team. Expensive.
I hate to say it, but a modern browser is not likely. A semi-modern one might be possible. Pale Moon could perhaps be done by a decent OS/2 programmer as an example. All the code is in the Mozilla tree and it is a fork, though lots of changes there. Even then a lot of web sites would  not work.
We're an ancient platform. I'm lucky as most of what I do, our SeaMonkey is good enough with work arounds like my Credit Union where clicking an icon doesn't work but using the menu does.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Dave Yeo on April 27, 2025, 01:37:31 am
The other side is financing. I don't know much about Arca Noae's internals, not even sure of how many people are involved though I have a better idea then most here. They have to pay IBM some amount for every copy of AOS sold or given away. No idea of how much but I'd imagine IBM wants enough that the profit margin on AOS isn't that high. I don't think they have many enterprise customers and the number of users has been steadily dropping. I think it is as much hobby and side hustle for the owners and not something that they are getting rich on and I doubt they have the resources to pay for browser development besides supplying a bit of infrastructure and similar. I also don't think the community is big enough to support much either.
All the above is my opinion as I don't know much about the internals of Arca Noae.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Michael on April 27, 2025, 06:15:41 am
What might be a better option would be a kind of XP Mode, like in Windows 7. A lean and fast virtual environment for running a modern browser. And as if that weren't dreamy enough, perhaps even seamless on the WPS.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Dave Yeo on April 27, 2025, 06:58:36 am
What might be a better option would be a kind of XP Mode, like in Windows 7. A lean and fast virtual environment for running a modern browser. And as if that weren't dreamy enough, perhaps even seamless on the WPS.

I don't know about a virtual machine. We do have Odin (think WINE). It hasn't been worked on in a while but you can run some Win32 programs, and can compile some Win32 programs to look native. In the past, Adobe Acrobat, OpenJDK for example have had their Win32 code recompiled on OS/2. Last time we went through this in the IE days, there was actually an official version of Opera built this way.
It would take syncing a recent version of WINE into Odin, probably a big job by skilled people, but a possibility.
Problem is that Win32 is going away. Less and less support it and I think Qt6 doesn't support it and not sure about others.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Michael on April 27, 2025, 08:02:42 am
I was thinking more along the lines of a lean Linux or BSD. And while I'm dreaming, I wanted it to run as a 64-bit system in the virtual environment and be able to use memory above the 4GB limit.

All of this is purely theoretical, of course, and without knowing whether it would even be technically possible.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Roderick Klein on April 27, 2025, 03:15:44 pm
What might be a better option would be a kind of XP Mode, like in Windows 7. A lean and fast virtual environment for running a modern browser. And as if that weren't dreamy enough, perhaps even seamless on the WPS.

It can work, however the big issue is that if you look how much memory the VM needs, plus the memory to run the OS + the browser. With the limited memory ArcaOS has it will pretty quickly run out of gas.

All discussions about PAE memory is sadly a big fat none starter. Idea's to make any program on OS/2 use PAE memory is a pipe dream. Now the current ArcaOS loader enables you to put a ramdrive in PAE memory. However the amount of addressable memory is not increased. The PAE code is extremely cool bit of codingin the ArcaOS loader. However it has a hidden cost in terms of performance. When you want to access memory above the 4 GB address range (for the ram drive) the loader effectively holds the CPU, switched CPU to PAE mode safe stack again and switch back. I am not describing it technically 100% accurate but PAE support via the UEFI loader has performance cost. There is BTW no other way to implement this.

Patch the OS/2 kernel to support PAE would possibly take about 6 to 12 months of full time. It would be a gigantic job. But then the drivers  need to be updated for this.
I looked at this with Steve Levine years ago and its to much work that it will mostl likely never work.

One of the things that needs to be looked at is if the memory consumption of the QT webkit engine can be reduced. But this is currently under research. I will have more news on QT development. Tune into Warpstock Europe video stream, details on wwww.warpstock.eu.

Roderick
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: JTA on April 27, 2025, 04:57:42 pm
OS/2 *is* dead ... too many problems in too many technical areas to really do what you want, unless lots of trade-offs are made ... in the "old days", we typically abandoned these OS's, and the user base went into decline. The trend of use is generally downward ...

But, with recent technology, all dead OS's can be revived now, and uses found for it that suit the interested end-user's goals of putting it to work again. With a HostOS (Windows/Linux), virtualization (Virtualbox), and Winflector (pipe 64-bit apps directly into OS/2), all running on a single modern hardware platform (laptop/desktop, i7, 32gb ram, ssd's), there are no OS/2 problems of any kind. The trend of use can be reversed, and made upward again ...

AN does their part, because generally ArcaOS is much better than Warp 4.52 ... use it standalone if you must, but it slots into AToF and improves many things over 4.52 use. Rather than beat up on AN to solve all those remaining problems with OS/2 in general, just use AToF to bypass every single one of them.

AToF just one of your modern laptops or dekstops (see the virtualization folder for the process), and you will be enabled to put OS/2 back to work again. You can use OS/2 and have all the modern ram, browsers, device support you can eat. I demonstrated AToF at Warpstock 2023 ... it just works, and I cannot see any other short-term solution to putting OS/2 back to work again (not without all those trade-offs).

It is very hard to get folks to try new things ... AToF languishes because (I think) there isn't a comparable solution to the Linux LiveCD/USB process. But, I think I've found all the pieces to build that, such that one can boot up a USB key and have a live AToF environment to test. Don't know if it'll be ready for Warpstock 2025.

Long-term, I think I even see a path for OS/2 to get to 64-bits (because it's been enabled in the past), but as the Beatles say ... "it's a long and winding road".

Getting back to the OP issue ... just AToF a box, and you are good to go with OS/2 or ArcaOS, and modern browsers.

Hope this helps ...
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Dariusz Piatkowski on April 27, 2025, 06:10:37 pm
JTA!

...With a HostOS (Windows/Linux), virtualization (Virtualbox), and Winflector (pipe 64-bit apps directly into OS/2), all running on a single modern hardware platform (laptop/desktop, i7, 32gb ram, ssd's), there are no OS/2 problems of any kind. The trend of use can be reversed, and made upward again...

Great post...I'm particularly interested in Winflector as I've been using Thinstuff for a while now which relies on the RDP protocol to share individual apps. On our end I use FreeRDP, and that works very well indeed.

Having said that, Thinstuff does not support Win10, and so there is a definite end-of-the-road there, and by the looks of it Winflector certainly does.

Looking at Winflector site though they talk about using "client" to connect to the hosted Apps, so what does that look like on our OS/2 platform? Is that RDP still?

Thanks,
-Dariusz
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: JTA on April 27, 2025, 07:29:38 pm
Winflector has server running on the AToF HostOS (as a service), and the native client running on some other OS where you want the windows apps to be displayed, in a native OS window frame. Don't know what the internal protocol is, but it works very well on AToF solutions, whereas RDP implementations caused me varying levels of grief.

Where there isn't a dedicated OS native client, Winflector also supports any browser with minimal capabilities (html 5, iirc), and our old OS/2 firefox fits this need. So, you can send any Window app, even 64-bit, over to the OS/2 vm.

Luckily, Winflector offers a free license, so it's pretty much available to all. I've asked them to build an OS/2 native 32-bit client, via their forums. AN could possibly help with this, and I've asked unofficially for them to push from their side, but ... no visible progress at either end.
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: Craig Miller on May 03, 2025, 12:22:05 pm
For whats it's worth, I have ArcaOS (5.1) running on an IBM T43 and for the first time in a long time, its running near flawlessly because of the upgrades Arca Noae have added over the years.

With that, the major issue I have is email and browser usage.

I think for a typical user, these two internet apps ARE the apps that have to work.

We are a small group and I know even if we got a new/updated browser, we will always have to live with compromises but that is ok.

What is not ok is I am almost to the point where I can't use Firefox  45.9.0 ESR for 80% of the sites I want to go to. I can't use email at all.

I don't expect miracles but I do hope there is a day to get these two apps working closer to where I don't have to look into remoting into another computer.

On a side note, I am REALLY hoping AI moves forward to the point it can help with programming. My bother is a programmer right now and uses it a bit to help him.

People that take their time to port/bring new apps to Arca OS are amazing but we have only have a few of them. I can't program, just don't have the mind set for it but if AI can help people like me contribute that would be a game changer.

I don't know if that will ever happen but I can dream =)
Title: Re: Arca Noae should be directly involved in web browser development efforts
Post by: JTA on May 03, 2025, 05:28:17 pm
AN does an amazing job with ArcaOS (AOS), and for me, running it in a VM, it is next to flawless (replacing Warp 4.52 in much of my OS/2 use).

Modern apps are indeed the only major remaining problem (*along with* 32-bit OS and certain device driver issues), but modern apps are a very intractable problem, as these have multiplied in various 64-bit code bases and feature sets. If you want to do something, there's a modern application for it ... but only if you are on a modern (64-bit) OS.

AN has said the "modern app" (browser) problem is not theirs to solve ... their focus is AOS. Their (commercial) base must not need a browser? And, their individual user base AOS sales aren't grossly impacted with the lack of such modern apps? They do their best to support those developers who are working on app availability in many ways, plus they've added ANPM to bring over more apps when available from porting. That's impressive ...

For those not wanting to run AOS in a VM, but rather directly on hardware, much of the entire set of modern apps will remain out of reach, unless you utilize RDP/RDC (to another machine), or Winflector (from another machine). Or wait for porting of a (dwindling?) small subset of all modern apps.

AToF is my solution to the modern app problem on OS/2, and running the entire AToF scheme on a laptop like your T43 should be no problem (if you can wrap your head around the AToF scheme). AToF is a simple combination of modern tools from our existing computing toolbox (beefy laptop/desktop, HostOS, Virtualbox, Winflector). As a bonus, you'll have little to no remaining 64-bit, device driver, or modern app problems, while booting into and working entirely in your ArcaOS desktop.

It must be quite the experience to run OS/2 (directly) on T43 hardware or similar, and perhaps it harkens back to the good ol' days, with IBM "suits" looking over both shoulders as we used their stuff, but I'd have to agree that you need a modern browser and email to live entirely in OS/2 on one machine in today's world. I personally can't live with most computing problems, and having experienced them just once, I immediately look for any solution that gets me out of said problems, vs the trade-off of "living with the problem(s)." Programming folks program their way out of problems, however long that takes, but us (old?) sysadmins (not being programmers) integrate thru or otherwise work around problems ... trying to get to a solution "now".

AFAIK, only AToF gets us 99% of the way there on a single machine like your laptop, potentially solving ALL problems. I hope to have an updated AToF at one of the upcoming Warpstock events ... I last presented it at Warpstock 2023.

But, you can try it right now ... wipe your machine (after backing up your data), load a HostOS, Virtualbox, and Winflector. Then put AOS into a vm ... done ... no more intractable problems; a modern browser and email app running on the AOS desktop, within the Winflector client. And, let me know what I can do to get AToF to 99.9% ...