Jump to content
ebiguy

RespeQt 5.2 and new Repository

Recommended Posts

4 hours ago, pusakat said:

You may need to add your user to the group that has rw permissions to the serial port.

 

On Ubuntu, `sudo adduser username dialout` should do the trick. On Arch, I'm not sure if adduser would do the trick.

The alternative command would be `sudo gpasswd -a username dialout`.

Be sure to logout and login again for the new group permissions to take effect.

 

To find out what group has rw permissions on the device, use `ls -g /dev/ttyS0` or `ls -g /dev/ttyUSB0`. It'll probably show `dialout`. If it isn't, change all references to dialout in the previous commands to the group that gets displayed.

In Arch Dialout is apparently UUCP.  I added myself to that.  I did ls -g /dev/ttyS0 and got the premissions, but ls -g /dev/ttyUSB0 returned no such file or directory.  Not sure what to make of that since it should be part of UUCP.

Share this post


Link to post
Share on other sites
2 minutes ago, Tempest said:

In Arch Dialout is apparently UUCP.  I added myself to that.  I did ls -g /dev/ttyS0 and got the premissions, but ls -g /dev/ttyUSB0 returned no such file or directory.  Not sure what to make of that since it should be part of UUCP.

 

/dev/ttyUSB0 is only for SIO2PC-USB devices. I was including it as an example.

 

Did that get it to work?

Share this post


Link to post
Share on other sites
34 minutes ago, pusakat said:

 

/dev/ttyUSB0 is only for SIO2PC-USB devices. I was including it as an example.

 

Did that get it to work?

I havent tried yet.  I cant get a working version of RespeQT on my box at the moment.

 

I do have a SIO2USB though, so I dont know why the USB group wouldn't exist.

Share this post


Link to post
Share on other sites
2 hours ago, ebiguy said:

Do you think you can help me identify the issue by checkouting previous commits (between 5.1 and 5.2) to determine when the issue appeared.

I'm not native english ... what does the word "commits" mean in this case? "Changes"? Or like "Changes by other developer's suggestions"?

If I had the time, I'd really like to help you out. May be it just would be usefull not to stop emulation caused by timeouts. All-in-all it timed out when it was not spoken to, so it wouldn't lead to an error. May be I had to switch to another handshake method (none in my case, but I don't see any differences between the other options).

 

On Ubuntu 18.04 LTS I get another error:

"assert" was not declared in this scope in printers/atari1029.cpp (in line 203) ...

Ubuntu 18.04 LTS uses QT 5.9; debian 10 (where this line compiles) uses QT 5.11

 

As "this shouldn't happen", I comment it out.

 

... ok ... updates are out. I had to make two changes in the source-code (see this very thread).

Edited by atarixle

Share this post


Link to post
Share on other sites
1 hour ago, atarixle said:

"assert" was not declared in this scope in printers/atari1029.cpp (in line 203) ...

I already reported this bug to Josh but had no answer on it up until now.

I had also to comment this line out to build the Windows version.

 

1 hour ago, atarixle said:

what does the word "commits" mean in this case

This is not a matter of English native speaker.

It is a technical word for version control system used to allow sharing and modifying sources among a distributed team.

The word commit comes from the Git and the way the sources are updated.

Did you use Git to get the source or did you download the whole zip from Github?

Share this post


Link to post
Share on other sites

Looks like printers/atari1029.cpp is simply missing an #include <assert.h> at the beginning.

 

BTW: tags between r4 and r5.2 are missing in the repo.

 

@atarixle with tags present you could do a git bisect (see "git help bisect"), eg "git bisect start r5.2 r5.1", then compile and enter "git bisect good" or "git bisect bad", depending on the test results.

 

Instead of the r5.x tags you could also use commit hashes, but we'd need to know which hash was r5.1

 

so long,

 

Hias

Edited by HiassofT

Share this post


Link to post
Share on other sites

Using qt59 and a very old ubuntu I couldn't compile from sources the new RespeQt 5.2. It fails in diskimage.cpp ambiguous overload of [] operator (probably an easy fix). However, last week I compiled the latest version from ebiguy github repository without problems...  

That is weird...

Share this post


Link to post
Share on other sites

 

23 hours ago, ebiguy said:

Did you use Git to get the source or did you download the whole zip from Github?

I always complie the .tar.gz sources from the release assets: https://github.com/RespeQt/RespeQt/releases

21 hours ago, HiassofT said:

Looks like printers/atari1029.cpp is simply missing an #include <assert.h> at the beginning.

In Debian 10 (QT 5.11) this was not causing any issues - only in Ubuntu 18.04 (QT 5.9).

 

1 hour ago, manterola said:

It fails in diskimage.cpp ambiguous overload of [] operator (probably an easy fix).

This error appeares only when building on 64 Bit systems. I just casted (quint16)size_of(...). (See my former posts)

Share this post


Link to post
Share on other sites
On 3/21/2020 at 8:26 AM, Mathy said:

Hello guys

 

With as many Mac users as there are in the Atari 8 bit scene, it puzzles me that there still is no new Mac version of RespeQt.

 

Sincerely

 

Mathy

 

Mac binaries were posted for the 4.3 beta here:

 

I've been using this without any problem.

 

Please, let's make sure this continues to be a multi-platform program and doesn't turn into a Windows-only solution.  New releases should confirm this continues to build successfully on all supported platforms.

 

 

  • Like 1

Share this post


Link to post
Share on other sites
7 hours ago, jamm said:

Please, let's make sure this continues to be a multi-platform program and doesn't turn into a Windows-only solution

I understand this sentence as a general call for contributions as the maintainer can not handle all platforms, if you see what I mean...

On 3/21/2020 at 10:46 PM, atarixle said:

Strangely it works in debian 10 32 bits. If I cast sizeof(...) to (quint16)sizeof(...) it builds to the end.

I fixed the compilation issues

On 3/21/2020 at 10:46 PM, atarixle said:

Oh I forgot this may be important information: it times-out when ever the Atari reads from the SIO2SD in High Speed.

@greblus made a great work discarding latest commits to find out which commit did bring a regression on divisor 0 speed.

He found the commit and proposed a solution.

 

I pushed all these modifications into develop branch of https://github.com/RespeQt/RespeQt

 

Can Linux users/developers try to compile the sources and report if the issue with the divisor 0 speed is fixed.

Thank you very much for your contribution.

If it actually works, I will merge with master and release a stable 5.3

 

  • Like 5

Share this post


Link to post
Share on other sites

As a Linux User (unfortunatelly neither a C++ nor a QT developer), I will try to compile the current brunch later this week ... Hopefully.

Share this post


Link to post
Share on other sites
On 3/26/2020 at 6:23 AM, ebiguy said:

I pushed all these modifications into develop branch of https://github.com/RespeQt/RespeQt

 

Can Linux users/developers try to compile the sources and report if the issue with the divisor 0 speed is fixed.

Thank you very much for your contribution.

If it actually works, I will merge with master and release a stable 5.3

 

Using Ubuntu 18.04 x86_64

 

commit 3a3fa0a1ac22b54a9f39b8bcaa64163a8d6459f7 (HEAD -> develop, origin/develop)
Author: ebiguy <[email protected]>
Date:   Wed Mar 25 23:02:41 2020 +0100

    Fixed divisor 0 speed

 

Here's the compiler error output:

moc_serialport-unix.o:(.data.rel.ro._ZTV15AtariSioBackend[_ZTV15AtariSioBackend]+0x100): undefined reference to `AtariSioBackend::forceHighSpeed(int)'
collect2: error: ld returned 1 exit status
Makefile:517: recipe for target 'RespeQt' failed
make: *** [RespeQt] Error 1

 

  • Thanks 1

Share this post


Link to post
Share on other sites
19 minutes ago, greblus said:

 

Try deleting the build directory and then compile again.

 

 

I get the same result from a freshly git cloned copy of [email protected]:RespeQt/RespeQt and checked out develop branch.

And I tried it several times with different directories just to be sure.

 

Share this post


Link to post
Share on other sites

Sorry about the error.

This is part of a Linux file I do not compile under Windows.

I think I fixed it (pushed in GitHub repository).

Could you please try again?

  • Like 1

Share this post


Link to post
Share on other sites

... just built the development branch for Ubuntu 18.04, debian 9 (32 bit) and debian 10 (64bit and 32 bit) ...

 

There were no errors while building. And there are no timeouts while using RespeQt.

 

On my very setup I can go as fast as pokey-divisor 6 (I know I also got 0 working, but my USB-serial needs good mood for that).

  • Like 2
  • Thanks 1

Share this post


Link to post
Share on other sites
8 hours ago, ebiguy said:

Sorry about the error.

This is part of a Linux file I do not compile under Windows.

I think I fixed it (pushed in GitHub repository).

Could you please try again?

Compiling fine again on Linux Mint 19.3 64 bit.

 

Thank you for resolving that issue!

 

  • Thanks 1

Share this post


Link to post
Share on other sites

Hi

 

Time to release the r5.3.

I changed the version to 5.3 and merged master and develop branches.

I already uploaded in GitHub the binaries for Windows (32bits and 64bits).

 

I would like to add binaries for other platforms.

Could someone build the official r5.3 release and send the binaries to me via PM, please?

 

I would prefer, if you don't mind, not having all binaries in the thread (as I did in the past) because people browsing the forum don't understand what is inside each binary.

I will put in this thread the links to binaries I receive so feel free to compile for any platform that you don't see here.

 

As of today, the links are:

For Windows 32bits: https://github.com/RespeQt/RespeQt/releases/download/r5.3/RespeQt-r5.3-32bits.zip

For Windows 64bits: https://github.com/RespeQt/RespeQt/releases/download/r5.3/RespeQt-r5.3-64bits.zip

 

Thank you for your contributions

  • Like 3

Share this post


Link to post
Share on other sites
13 minutes ago, ebiguy said:

Hi

 

Time to release the r5.3.

I changed the version to 5.3 and merged master and develop branches.

I already uploaded in GitHub the binaries for Windows (32bits and 64bits).

 

I would like to add binaries for other platforms.

Could someone build the official r5.3 release and send the binaries to me via PM, please?

 

I would prefer, if you don't mind, not having all binaries in the thread (as I did in the past) because people browsing the forum don't understand what is inside each binary.

I will put in this thread the links to binaries I receive so feel free to compile for any platform that you don't see here.

 

As of today, the links are:

For Windows 32bits: https://github.com/RespeQt/RespeQt/releases/download/r5.3/RespeQt-r5.3-32bits.zip

For Windows 64bits: https://github.com/RespeQt/RespeQt/releases/download/r5.3/RespeQt-r5.3-64bits.zip

 

Thank you for your contributions

Hello ebiguy,

 

 

Thanks for the 5.3 release, i will check it but first my sdrive max for test.

 

Greetings Marco.

 

P.s. dis6502 still in progress? version 3.0

Share this post


Link to post
Share on other sites

Now that SIO is working, could we fix printer emulation?

 

I mean, absolutely nothing works here in Linux ... it even makes RespeQt stop working (memory protection error or so ...).

Share this post


Link to post
Share on other sites
17 minutes ago, atarixle said:

it even makes RespeQt stop working (memory protection error or so ...)

I was not aware of that.

You should have tell me before I released the v5.3.

That's too late for this version.

I will have a look if this problem is also in the Windows version

Share this post


Link to post
Share on other sites

First, the printer emulation crashes also on Windows.

I am trying with a 1020 emulation with Text window as output.

I tried to revert to many different commits in the history (more than 20), recompiling from scratch each time.

The latest working version I can find for this configuration (1020 / Text window) is dated 04/11/2019 which is almost one year old (78 commits ago):

* fbcc13a9e06b0626216d27a6652768533bd9e133 More decoupling of dependencies

Between this commit and the current version, I have either a segmentation violation (at different places depending on commit) or a message telling me:

* You are not allowed to use the passthrough emulation without an raw output

I let Joschen look at the problem (or anyone else) as I am not comfortable with the printer code.

 

Share this post


Link to post
Share on other sites

I'm not a Mac developer, but since no one else has volunteered to build the current version on Mac, I'm trying... 

I get the following warnings and errors using the latest version just cloned from Github and with the latest Qt (5.14.1).  I would expect you'd see these on Windows and Linux, also, unless you're using a different version of Qt or have disabled some compiler behaviors?

 

Screen Shot 2020-03-28 at 11.21.59 AM.png

 

The first two warnings are simple enough (char values only go up to 127, and we're trying to set one to 155 here.

Edited by jamm
  • Thanks 1

Share this post


Link to post
Share on other sites
3 hours ago, ebiguy said:

First, the printer emulation crashes also on Windows.

Good to know at least it's not a Linux-only problem ("permissions" sounds UNIXish).

If you version RespeQt step-by-step, I am ok with that. So we're aware of where we are. We know that SIO works in 5.3; let fix printers in 5.4. You may fix text-only-printing in 5.4, then 1029-SVG-printing in 5.5, I am alright with that.

I mean, it's just numbers.

Share this post


Link to post
Share on other sites
13 hours ago, jamm said:

I'm not a Mac developer, but since no one else has volunteered to build the current version on Mac, I'm trying...

Thank you for trying and ... succeeding !!!

Now we have also the Mac distribution package: https://github.com/RespeQt/RespeQt/releases/download/r5.3/RespeQt-r5.3-Mac.zip

 

Anyone to test this package ?

Edited by ebiguy
  • Like 1
  • Thanks 2

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...