Sorry to raise this thread from the dead but it is pertinent. Quite a few years ago I posted about the death of Yahoo Widgets, RocketDock and Xwidgets. I then inquired about the suitability of Winstep widgets to do some of what I wanted, in that chat you recommended that I create my own engine...

Well, I haven't quite done that but I have done some steps towards building a replica environment of what I used to have... bear with me.



That is a high res 4K image loaded from imgur, so forgive if it takes a while to load.

That is my current desktop, windows 7, 10 or 11, it is the same.

PS. There is a possibility that being discrete binaries they could be Winstep widgets...

I have:
Replicated Rocketdock using VB6.
Created a VB6 template for creating simple widgets, the planets.
Created a VB6 template for complex widgets, the gauges.
There are a three remaining .js widgets there that are being converted now.

I have embraced good old VB6 because I have found a graphic framework based on Cairo that can put transparent PNGs on the desktop. In addition there is now TwinBasic, a new product that is VB6 compatible and can compile 64bit binaries. It also has modern language features. In time it will be multi platform. Currrently in alpha. I will be migrating to TB64 when it is production ready.

Hope you like what you see. Not in competition with Winstep! All my code is FOSS and most is unfinished. It all works though! When I have something for you to test I will drop you a line.

When you suggested this, I didn't feel competent back then to begin such a task. I am several steps down the path and I have something that works though it is some way from being a widget engine. Each of the gauges and planets is a discrete binary, slimline and using very little CPU (<0.01%).

Thankyou for your suggestion and in my way I have taken your advice!

Wow! That's impressive! Nice Steampunk, too. Just one question arises, though. Do all those 'widgets' actually have a function and if so what function?

Normally, I'd come down on posts added to old threads, but in view of the relevance I think we can say, all is forgiven.

That's quite an effort. Although, given what MS are doing with Windows 11 (and f*** knows what's going to happen in Windows 12!) especially, it might have been wiser to stay clear of VB in any way, shape or form - something like C# might have been the better choice perhaps and would have had the added advantage of being easily portable. (It also looks relatively easy to learn, going by what I have seen.) And also good to see that you've decided to have it all FOSS! (As a mainly Linux user I'm very much into the FOSS principle, of course. Even if it's not free of cost.)

All together, I have to say "well done"!

Oh yes, they all have a function, they all work. The gauges display the usual sysmetrics, there are several different types. The desktop hole is a pit where you can drop desktop icons for automatic sorting, the trinkets (planets) are merely for eye candy but they can be assigned a function. The volume control does what it says on the tin. The weather gauges display current weather characteristics. The stopwatch and alarms function as expected. The dock works as per Rocketdock. Nothing is static, all is dynamic.

Coding in VB6 makes better sense than it ever has due to the forthcoming replacement TwinBasic. None of my knowledge is lost, the same code compiles using TB64 and is now out of Microsoft's hands. They cannot disable it.

Go have a look at TwinBasic.

Do me a favour and let Jorge know I responded to his post. I think he'll find it interesting.

Re: portability, Wayne Phillips the dev of TwinBasic has a roadmap of making the product multiplatform too so that avenue is covered. In my humble opinion taking one's eggs out of Microsoft's basket is always a good idea before they scramble them...

I mentioned the desktop hole earlier but I realise it is hidden under my comms program in that image.

Here it is on another desktop.

Great stuff, well done!


Unfortunately, I have to disagree there. Under Windows 11, part of Winstep is struggling with work-arounds as the original code is no longer working due to MS having completely changed the taskbar and esp. the system tray, all for no good reason except cosmetics. Ask Jorge what a PITA the whole business has become. And there's always a possibility that MS could change Win32 to be only available to .net, or indeed to scrap Win32 completely. In the final analysis, MS can do whatever they want, alas. (And I haven't even mentioned the issue of VB being single-threaded only, also a major pain, but one that wasn't an issue when Winstep was started back around 1998.)

Jorge lately tends to be absent from the forums around the weekend, but I'm sure he'll see this thread during the week.

Nice!

Oh yes, I see now what you are referring to. Microdolts taking away control and screwing with things just for the sake of it.

As you say, the taskbar screw up is one such problem. I have always run the windows taskbar at the top of the screen and placed my own dock at the bottom. It has become increasingly hard to manually move the windows taskbar and it appears it still cannot be done programatically, certainly not without a registry change and a reboot, which is super rubbish. They seem to be locking down customisation of the UI which is, for me a big reason to move from Windows, but - I can't of course... as all my skill resides within Win platform environment.

Luckily, I am not writing commercial code and so I don't give a flying poo what MS are going to do next - as it is going to be feck-up regardless of whether I like it or not. Instead I just keep coding, learning, producing good stuff and ignoring Microdolts and their latest shennanigans. I am lucky that it has no affect on me commercially.

As an alternative o/s I contribute in a small way to ReactOS and test my programs there. I have a large tranche of testing to do there shortly. Many of my programs work under wine on Linux too.

I appreciate your response and why your preferred environment is Linux. When TB64 goes Linux, I have no doubt my direction will shift there too.

In general, I no longer use anything from MS except their o/s and VB6, re: oses, I have my favourites: Win XP (loved it), Win 7, a very good 64bit o/s, everything since a bit sh1t, Win 10 bearable, just, requiring a lot of customisation with openshell, tweakUI, retrobar and trying to avoid using the settings screens where possible.

Re: VB6 single-threading.

Not being able to perform tasks in new threads has always been very limiting, some of the tasks my dock has to perform require multiple threads and without this capability it sometimes on some tasks performs rather slowly. TB64 by contrast has multiple thread capability. It is a different and more advanced product with 100% compatibility.

I am not its dedicated salesman but I might as well be.

re: win32 - that is not going to disappear for a decade yet and by that time ReactOS may be usable - always another ten years away.

Don't bet on Win32 being around for another decade. Don't even bet on Windows still being around for another decade as a desktop OS. MS will go where the bread is.

As for VB/TB64, multi-threading or otherwise, it's an irrelevant language already (and the last thing anyone in Linux would still want - heck, we've got a huge amount of different languages, inc. compiled scripting languages, which is what VB essentially is, too.)

If the only thing you're still using Windblows for is coding, then jump ship already before it goes down. Hell, skill set or not, that's acquired very quickly elsewhere, OSes aren't *that* different from one another these days. In my time I learned to work with something like a couple dozen different OSes without a problem, starting with Commodore Basic (and of course coding it) for its PET, then their other 8-bit machines up to the C=64, then the Amigas, then Nextstep on NeXT hardware, Irix on SGI Indigo and Indigo Power, Solaris on Sparc/Ultrasparc and eventually on my first, self-built, PC (alongside Nextstep/Openstep x86, OS/2 Warp (without DOS/Windblows), Slackware Linux (didn't last too long at the time - too unstable), BeOS x86, and eventually Win NT4 - the first useable Windows.) And I used to code under several of them - Objective C under NextSTEP/OPENSTEP and the OPENSTEP frameworks for Win NT, Solaris and Java, early php under NT, and so on.

As for ReactOS, heck, that's sooo long in the tooth already and still perpetually will be another 10-20 years away. I've never really seen the point of it anyway I must admit. Why on earth try and reinvent the wheel? A bit like BeOS and Haiku - re-inventing the wheel in the form of another Tom Tit OS that went down the plughole.

At least the major Linux distros are mature OSes - as are the *BSDs nowadays - with a vast treasure trove of superb apps. Although, alas, even Linux is not without its problems. Hardware support can be one of them, albeit getting a lot better, and too many distros with far too many different DEs. Far too much choice for users. Not a good thing, too much choice, in anything. Eventually people get tired of it all - just look at what's happened to the whole Windows customisation and skinning scenes. Virtually dead, apart from Winstep and just about Rainmeter. And that's even without Microshite entirely making things impossible.

I can see you are a thorough sceptic. I was that way inclined too but then I forgot to care, settled down in my nice niche with a few chosen technologies and stopped worrying.

Linux is a fine idea but I'll wait until two things occur. 1. My chosen development environment migrates to it and 2. More users adopt it as a day to day OS (not just us geeks).

That's quite a list. Mine isn't quite as long nor varied.

I started out on ZX80 BASIC, then GWBASIC on the IBM PC, moved onto scripting with DCL on VMS, then QB45 on DOS, VBDOS, followed by VB5/6. Then a jump to HTML/.js/ jscript and PHP, a brief sojourn to typescript and then back to VB6.

Loved .js when coupled to a decent widgeting engine, otherwise it was a bit of a pain. Tried building some apps in VB.NET - succeeded but hated it. Returning to VB6 is like returning home. TwinBasic64 is a lifeline to my own personal obsolescence. I will be diving some more into typescript soon.

That's it.

1) Is unlikely to be practical in the first place - Linux is a very different beast to Windblows. There are multiple dependencies that VB would find very hard to cope with, also, even with efficient multithreading, it will be far too slow. But even ignoring all that, it would be very likely to be met with a pretty hostile reception among developers. Whether VB or TB64, anything of the sort never had a place in the *nix world. To seriously code in Linux, you need at least one member of the C family, plus Python, especially for the kind of thing that you're doing. (Not a sceptic, just a realist.) And -

2)This isn't the '90s or even '00s any more, Linux is far from geek-dominated - you want geek, you'd want Arch Linux. And Linux is getting pretty close to overtaking the Mac in terms of user base. More desktops and laptops than ever from major brands such as Lenovo and Acer come with Linux pre-installed, mainly Ubuntu and Fedora. The user base is not only growing but also more and more diversifying. I've personally helped quite a few people move from Windows to Linux, people who mostly would never before have thought about installing any OS and who now are happy as Larry. Linux also happens to be much more user friendly than Windows or macOS, on the whole. In ordinary day to day use, the ordinary user would almost never really need to use the terminal (cli). Even if he did, it would be easy enough to pick up the basics, especially as it's usually the BASH shell - powerful yet easy to use. (I've used one form or another of the *nix Bourne shell or its FOSS BASH brother since about '90,and picked it up in absolutely no time at all and quickly learned to appreciate it when needed and indeed to love using it.) Linux is a whole lot more than just a fine idea today. It's far and away the best desktop OS - never mind that somewhere in the region of 80-90% of today's servers and 90%+ of all super computers run on Linux - for the average user there is.

In conclusion I have to say I admire and love what you've done so far, and all that would certainly find a substantial fan base in Linux, without a doubt. But any form of VB is just flogging a dead horse, unless you've got decades invested in it and re-doing thousands and thousands of lines of code in another language just wouldn't be feasible. You're obviously very good and very talented at what you're doing, so my advice to you, if I may proffer such, would be, don't waste your talent, learn a couple of alternative, future-proof languages, and work in Linux from the start (or the re-boot! ) and maybe port to Windows.

No, I'm fine. I have plenty of Unix/linux experience going back decades so I'm happy there if and when I choose to make the move. For the moment and for the foreseeable future VB6 has an actual future and I'm going along with it.

With the sort of things I want to do it is the perfect vehicle. If there is ever another I will jump ship, I'm always on the look out but not yet...

Bear in mind that I drive a Triumph Spitfire, a BSA motorbike and my house is 17th century. I am quite used to being in and around obsolescent tools and quite enjoy the experience.