Should we create a 2.2.1 branch?

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
23 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Should we create a 2.2.1 branch?

dxli
Since we have added lots of features since 2.2.0.x:

1. create a freature freeze branch towards a stable 2.2.1 release;
2. Make the master branch Qt6 and drop 32bit windows support.

Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

cadtec
This post was updated on .
Good morning dxli

dxli wrote
1. create a freature freeze branch towards a stable 2.2.1 release;
2. Make the master branch Qt6 and drop 32bit windows support.
I can agree with both points.
But I have found one more thing:

Clicking on the following button crashes version 2.2.1_alpha-424-g806e61d0:


Furthermore, the round line ends of individual lines do not yet work:


This also applies to polylines, polygons an text, where the line ends are also angular.

I work under Windows 11 23H2.


Kind regards,
André

Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

cadtec
In reply to this post by dxli
Select a text in a drawing.
Right-click on the text -> Edit Block ...
The block is opened.
Click on Current Drawing Options -> LibreCAD crashes.
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

jona
In reply to this post by dxli
As 2.2.0 is two year old (dec 2022) and that a lot of development occurred since then, I can't agree more that a release would be welcome ! :-)
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dellus
In reply to this post by cadtec
As the link dxli gave from QT documentation shows settings for line ends and for joins are separate. So obviously in alpha 424 only joins have been set to round. If you draw a couple of lines as a polyline (like rectangle) you actually get rounded joins, if you draw these as normal single lines, you get squared off joins. In both cases open line ends are always square.
Probably something dxli just has missed, easy fix.
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
In reply to this post by cadtec
Added a quick fix in master, will do some refactoring later on.

cadtec wrote
Select a text in a drawing.
Right-click on the text -> Edit Block ...
The block is opened.
Click on Current Drawing Options -> LibreCAD crashes.
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
In reply to this post by cadtec
The pen widgets are currently work-in-progress with two large PRs from sand1024.

Both Cap-style and Join-style are now fixed to "Round" in master.

cadtec wrote
Good morning dxli

dxli wrote
1. create a freature freeze branch towards a stable 2.2.1 release;
2. Make the master branch Qt6 and drop 32bit windows support.
I can agree with both points.
But I have found one more thing:

Clicking on the following button crashes version 2.2.1_alpha-424-g806e61d0:


Furthermore, the round line ends of individual lines do not yet work:


This also applies to polylines, polygons an text, where the line ends are also angular.

I work under Windows 11 23H2.


Kind regards,
André
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
In reply to this post by jona
Great!

I will fork a 2.2.1 after the PRs from sand1024.

The PRs are quite solid in quality, so they should go to 2.2.1

jona wrote
As 2.2.0 is two year old (dec 2022) and that a lot of development occurred since then, I can't agree more that a release would be welcome ! :-)
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

cadtec
dxli wrote
The PRs are quite solid in quality, so they should go to 2.2.1
A big compliment!
I almost can't wait ;-).

Kind regards,
André
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

flywire
In reply to this post by dxli
In the next release I'd like to see:

* snap point Start/Middle/End fixed on Line commands for Orthogonal and Relative Angle

* fix the relative angle display rounding issue that often displays a 30° angle as 29.99999999999999999999999996°



Version: 2.2.1_alpha-412-g9ccea6fb
Compiler: GNU GCC 13.2.0
Compiled on: Apr 17 2024
Qt Version: 5.15.2
Boost Version: 1.75.0
System: Windows 10 Version 2009
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
The angle display issue is easy.

Could you explain more on the snap point part?

It's useful to have an orthogonal line at intersection points, for example.

flywire wrote
In the next release I'd like to see:

* snap point Start/Middle/End fixed on Line commands for Orthogonal and Relative Angle

* fix the relative angle display rounding issue that often displays a 30° angle as 29.99999999999999999999999996°



Version: 2.2.1_alpha-412-g9ccea6fb
Compiler: GNU GCC 13.2.0
Compiled on: Apr 17 2024
Qt Version: 5.15.2
Boost Version: 1.75.0
System: Windows 10 Version 2009
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

cadtec
In reply to this post by dxli
Found a little Bug:

If the diameter symbol is activated when dimensioning a circle, it is also active in the aligned dimensioning, angular dimension, Arc measure  and cannot be deactivated there.

Deactivation is only possible if you go back to the diameter dimensioning and deactivate the diameter button there.

Best regards,
André
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

flywire
In reply to this post by dxli
dxli wrote
Could you explain more on the snap point part?
Please draw a line then try the line types and you will notice the options are inconsistent: https://docs.librecad.org/en/latest/ref/tools.html#line

2. Angle: Angle, Length, Snap Point (Start,Middle,End) options work as expected
3. Horizontal: Length, Snap Point (Start,Middle,End) options work as expected
4. Vertical: No options displayed (bug) - should have Length, Snap Point (Start,Middle,End)
12. Orthogonal: Length - should have Snap Point (Start,Middle,End) option too
13. Relative Angle: Angle, Length - should have Snap Point (Start,Middle,End) option too

Version: 2.2.1_alpha-412-g9ccea6fb
Compiler: GNU GCC 13.2.0
Compiled on: Apr 17 2024
Qt Version: 5.15.2
Boost Version: 1.75.0
System: Windows 10 Version 2009
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
In reply to this post by cadtec
Fixed

cadtec wrote
Found a little Bug:

If the diameter symbol is activated when dimensioning a circle, it is also active in the aligned dimensioning, angular dimension, Arc measure  and cannot be deactivated there.

Deactivation is only possible if you go back to the diameter dimensioning and deactivate the diameter button there.

Best regards,
André
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
In reply to this post by flywire
The snap point is a feature implemented at individual action level.

Our current action design makes it difficult to share features across drawing actions.

To copy the implementation is an obvious solution here, but we will see whether a better design can help

flywire wrote
dxli wrote
Could you explain more on the snap point part?
Please draw a line then try the line types and you will notice the options are inconsistent: https://docs.librecad.org/en/latest/ref/tools.html#line

2. Angle: Angle, Length, Snap Point (Start,Middle,End) options work as expected
3. Horizontal: Length, Snap Point (Start,Middle,End) options work as expected
4. Vertical: No options displayed (bug) - should have Length, Snap Point (Start,Middle,End)
12. Orthogonal: Length - should have Snap Point (Start,Middle,End) option too
13. Relative Angle: Angle, Length - should have Snap Point (Start,Middle,End) option too

Version: 2.2.1_alpha-412-g9ccea6fb
Compiler: GNU GCC 13.2.0
Compiled on: Apr 17 2024
Qt Version: 5.15.2
Boost Version: 1.75.0
System: Windows 10 Version 2009
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

cadtec
In reply to this post by dxli
dxli wrote
Fixed
I can confirm that it works on Windows 11.

Thank you very much!


Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

flywire
In reply to this post by dxli
dxli wrote
...we will see whether a better design can help
That would be good. Many of those features are the same command but the angle is know so not required from the interface.
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
In reply to this post by dxli
Done:

1, created a 2.2.1 branch (fixed macOS building), tagged 2.2.1_rc1;
2, moved the master branch to qt6, dropped 32bit win32 support, with 3 architectures left: Linux, macOS, win64.


dxli wrote
Since we have added lots of features since 2.2.0.x:

1. create a freature freeze branch towards a stable 2.2.1 release;
2. Make the master branch Qt6 and drop 32bit windows support.
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

flywire
@dxli Did you see the comment?
> Looks good except Relative Angle still does not have Snap Point.
Reply | Threaded
Open this post in threaded view
|

Re: Should we create a 2.2.1 branch?

dxli
I mean we will figure out a way to do it better in 2.2.2.

The current implementation is at individual action level, so new implementation is needed for additional actions.

It's probably too late for 2.2.1.

flywire wrote
@dxli Did you see the comment?
> Looks good except Relative Angle still does not have Snap Point.
12