AspeQT release 79 (1.0.0 preview-7) Source Code

[Joey Z]

So a while back, I said I had the release 79 source of aspeqt, and that I'd post it as soon as I reverted the changes I had to make to get it to compile on a weird setup. I have finally done this using diff on this source and r77. I remembered approximately what kind of changes I made, but not where they were, or exactly. Using diff to compare to r77, I was able to refresh my memory, and I believe this is basically r79 the way it was when I downloaded it originally.

 

 

On a related note, I will be forking this version at some point, and I will do my best to 'de-Ray' it. That is, fixing things I consider mistakes by Ray. I will also be fixing bugs and adding features when I have time to do so.

 

List of stuff to do for fork right away, when I get around to it:

Add ApeTime support back in (which should never have been removed in the first place)

Implementing a workaround for a bug I found while test compiling this source which can cause the menu bar to not show up. This is not a bug in AspeQT, rather it is a bug in some QT library. See Below for more details.

Remove reference/links to Ray's website for a couple reasons. First, Ray apparently continues to develop Aspeqt, so I wouldn't want to confuse users by directing them to Rays site for a fork, which Ray wouldn't be supporting of course. Second, he was using the software to advertise his own commercial motives/interests, which I personally believe should have no place in open source software, even if Ray was within his rights in doing this.

 

 

***THE BUG***

So this version had a bit of an issue for me when I compiled it actually. I am using a bit of a weird setup, a recent version of Ubuntu (I forget which) that has been kind of converted into Xubuntu by uninstalling various ubuntu packages and installing various xubuntu packages in their place. AspeQT's menu bar was not showing up in the window. Some googling revealed the problem is related to the appmenu-qt and appmenu-qt5 packages. If you remove these, the problem is solved. However, there is a better workaround than that, one which requires modifying the AspeQT source, more specifically mainwindow.ui.

<property name="nativeMenuBar">
<bool>false</bool>
</property>

If you add that snippet of XML just after this spot in mainwindow.ui:

<widget class="QMenuBar" name="menuBar">
   <property name="geometry">
    <rect>
     <x>0</x>
     <y>0</y>
     <width>684</width>
     <height>21</height>
    </rect>
   </property>

it seems to fix the problem as well. If you have issues, you can perform this fix yourself, or just uninstall the appmenu-qt related packages. This workaround will be included whenever I get around to forking r79 of aspeqt.

aspeqt-code-79-trunk.tar.gz

[morelenmir]

Many, many thanks Joey Z!!! This is exactly what AspeQT needs - a shot of fresh life and enthusiasm to get beyond the... unpleasantness... which led to a year-long stagnation. Hopefully one or two of the other genius programmers around here will come on board and 'Aspeqt' will be transformed. I am really looking forward to it!

 

In many ways "AspeQT" running on a cheap tablet is the ultimate floppy-drive replacement for the A8 - and these days a real bog-standard, HongKong knock-off tablet really doesn't cost that much more than a cased SIO2PC!

[RodLightning]

Agree 100%. Thanks for your efforts which are much appreciated!

[+DrVenkman]

This will be an interesting thread to follow. I previously compiled a version last summer but it had some weird Qt-related bugs (I think) that rendered performance abysmal. It ran but it suffered from stutters, halts and very slow net transfer rate. Oddly, the Windows version (running under WINE!) ran perfectly on the same hardware and using the same SIO2USB device. *shrug*

 

Anyway, I tried to compile the r77 version posted in the prior thread for OS X yesterday but ran into an immediate user-error on my part. I've wiped my drive since last summer and had to re-download all the Qt tools and libraries. I opened the project up in Qt Creator but get an immediate error as soon as I try to build it. I'm sure there's something missing from the project file or some supplemental library I missed. If anyone else wants to try, I'd like to hear what I screwed up; I shamelessly admit I know zilch about Qt, and it's been decades since I coded anything significant from scratch, in any language on any platform.

Edited June 29, 2015 by DrVenkman

[Joey Z]

This will be an interesting thread to follow. I previously compiled a version last summer but it had some weird Qt-related bugs (I think) that rendered performance abysmal. It ran but it suffered from stutters, halts and very slow net transfer rate. Oddly, the Windows version (running under WINE!) ran perfectly on the same hardware and using the same SIO2USB device. *shrug*

 

Anyway, I tried to compile the r77 version posted in the prior thread for OS X yesterday but ran into an immediate user-error on my part. I've wiped my drive since last summer and had to re-download all the Qt tools and libraries. I opened the project up in Qt Creator but get an immediate error as soon as I try to build it. I'm sure there's something missing from the project file or some supplemental library I missed. If anyone else wants to try, I'd like to hear what I screwed up; I shamelessly admit I know zilch about Qt, and it's been decades since I coded anything significant from scratch, in any language on any platform.

what error did you get? I know there *are* some errors in ui_xxx.h files from r77. These can be safely deleted as they will automatically be rebuilt from their respective .ui files.

[+DrVenkman]

what error did you get? I know there *are* some errors in ui_xxx.h files from r77. These can be safely deleted as they will automatically be rebuilt from their respective .ui files.

 

Huh. Like I said, user error. I just tried again with the r79 code you provided. I don't know what I did differently but this time I got literally dozens of potential issues (and about half of them cautionary warnings) but damned if the thing didn't build a proper executable. I need to dig out one of my 8-bit machines and my SIO2USB device to test it, but the compiled app at least loads and runs.

 

EDIT: Additional comments and observations. I still haven't tested full functionality but I did dig out my SIO2USB device and this build does recognize it and open up a communication channel to it, so that's promising. I'll have to do some actual I/O tests with it to a real Atari to se if it's affected by the weird issues with hesitation and spiking CPU usage I was seeing last summer with the prior codebase (r66?). For what it's worth, I've now proven I can build OS X versions which at least load and run and can communicate with an SIO2USB device. I'll be following the thread as the real coders make changes and discuss them, and I'm happy to periodically produce builds and post them; no need for everyone and their brother to download gigabytes of Xcode and then Qt tools and such unless they want to. Last summer I also spent far too much time making a pretty icon (a 1050 with a 410 on top, like I had as a kid). I'm happy to contribute that "art" so the OS X version at least has a pretty icon instead of the ugly default.

Edited June 30, 2015 by DrVenkman

[+DrVenkman]

The forum editor ate some of my other comments:

 

The About page for this code still has some references, links and verbiage you may want to clean up and update. To be clear: I own one of Ray's devices and it works perfectly and is very well-made, so nothing in this suggestion should be taken as any kind of criticism in that regard. It's just that if the code is to move forward, that might be one area that needs to be fixed.

 

Lastly, we might also want to consider a name change. AspeQt carries a lot of baggage at this point, and there's no need for it. I don't have any suggested alternatives (yet!) but perhaps others will.

[Joey Z]

The forum editor ate some of my other comments:

 

The About page for this code still has some references, links and verbiage you may want to clean up and update. To be clear: I own one of Ray's devices and it works perfectly and is very well-made, so nothing in this suggestion should be taken as any kind of criticism in that regard. It's just that if the code is to move forward, that might be one area that needs to be fixed.

 

Lastly, we might also want to consider a name change. AspeQt carries a lot of baggage at this point, and there's no need for it. I don't have any suggested alternatives (yet!) but perhaps others will.

Yes, well, this isn't the fork just yet, I'm working on 'de-ray'ing it as we speak, converting references to AspeQt into the name of the fork, removing references to Ray's website, etc. The source code here is an actual version of AspeQt I had downloaded sometime last year, and figured I'd release it here so people have it until I get this fork going...

[RodLightning]

Lastly, we might also want to consider a name change. AspeQt carries a lot of baggage at this point, and there's no need for it. I don't have any suggested alternatives (yet!) but perhaps others will.

Maybe a modifier to differentiate anything done by Joey Z or others? AspeQt-e, AspeQt+ or even AspeQt-JZ ?

 

More suggestions anyone...

[+DrVenkman]

The source code here is an actual version of AspeQt I had downloaded sometime last year, and figured I'd release it here so people have it until I get this fork going...

 

I appreciate you doing so. And I also appreciate the fact that it piqued my curiosity enough to actually see if I could compile it successfully.

[Joey Z]

Maybe a modifier to differentiate anything done by Joey Z or others? AspeQt-e, AspeQt+ or even AspeQt-JZ ?

 

More suggestions anyone...

I was thinking: RespeQt