Home » Community » U++ community news and announcements » 2024rc2
|
|
|
|
Re: 2024rc2 [message #61063 is a reply to message #61060] |
Tue, 05 November 2024 08:54   |
cbpporter
Messages: 1427 Registered: September 2007
|
Ultimate Contributor |
|
|
After some more testing, not all the awkwardness of the editor is done.
Example, open up AddressBook, go to a function where "array" is defined, type array, then '.', autocomplete popup, press down arrow to navigate the list, press tab, the editor add a tab instead of auto-complete. Other scenarios involving tab work.
|
|
|
Re: 2024rc2 [message #61064 is a reply to message #61057] |
Tue, 05 November 2024 09:32   |
Oblivion
Messages: 1209 Registered: August 2007
|
Senior Contributor |
|
|
There is also another problem I've noticed:
SQLite doesn't seem to work in release mode (in Windows). Examples, such as SQL_SQLite, fail to create tables with the following log:
* C:\Users\user2\Desktop\upp\out\reference\CLANGx64.Blitz\SQL_Sqlite3.exe 04.11.2024 16:38:05, user: user2
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (0, 'Joe', 'Smith', 20000101)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (1, 'Mike', 'Smith', 20000102)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (2, 'Jon', 'Goober', 20000103)
ERROR no such table: SIMPLE_TEST1(1): Preparing: delete from SIMPLE_TEST1 where ID > 3
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (5, 'wrongname', 'wronglastname', 20010604)
ERROR no such table: SIMPLE_TEST1(1): Preparing: update SIMPLE_TEST1 set NAME = 'rightname', LASTNAME = 'rightlastname', BDATE = 20010604 where ID = 5
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: select ID, NAME, LASTNAME, BDATE from SIMPLE_TEST1
ERROR no such table: SIMPLE_TEST1(1): Preparing: select * from SIMPLE_TEST1
Am I missing something here, this shouldn't be happening in release mode, right? The examples work fine (read: can succesfully create and query tables) in debug mode.
Best regards,
Oblivion
Github page: https://github.com/ismail-yilmaz
upp-components: https://github.com/ismail-yilmaz/upp-components
Bobcat the terminal emulator: https://github.com/ismail-yilmaz/Bobcat
|
|
|
Re: 2024rc2 [message #61065 is a reply to message #61064] |
Tue, 05 November 2024 09:56   |
 |
mirek
Messages: 14260 Registered: November 2005
|
Ultimate Member |
|
|
Oblivion wrote on Tue, 05 November 2024 09:32There is also another problem I've noticed:
SQLite doesn't seem to work in release mode (in Windows). Examples, such as SQL_SQLite, fail to create tables with the following log:
* C:\Users\user2\Desktop\upp\out\reference\CLANGx64.Blitz\SQL_Sqlite3.exe 04.11.2024 16:38:05, user: user2
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (0, 'Joe', 'Smith', 20000101)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (1, 'Mike', 'Smith', 20000102)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (2, 'Jon', 'Goober', 20000103)
ERROR no such table: SIMPLE_TEST1(1): Preparing: delete from SIMPLE_TEST1 where ID > 3
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(ID, NAME, LASTNAME, BDATE) values (5, 'wrongname', 'wronglastname', 20010604)
ERROR no such table: SIMPLE_TEST1(1): Preparing: update SIMPLE_TEST1 set NAME = 'rightname', LASTNAME = 'rightlastname', BDATE = 20010604 where ID = 5
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: insert into SIMPLE_TEST1(name,bdate) values(?,?)
ERROR no such table: SIMPLE_TEST1(1): Preparing: select ID, NAME, LASTNAME, BDATE from SIMPLE_TEST1
ERROR no such table: SIMPLE_TEST1(1): Preparing: select * from SIMPLE_TEST1
Am I missing something here, this shouldn't be happening in release mode, right? The examples work fine (read: can succesfully create and query tables) in debug mode.
Best regards,
Oblivion
Uhm, it is intentional, although for sqlite3 example maybe a bit misplaced. If you check the example, schema code is in #ifdef _DEBUG blocks...
The explanation is that in production (that is "release mode"), you really do no want to run schema updates in your application. So traditionally, we were using schema updates just for development, then pick generated scripts in debug mode and use them to prepare upgrade packages (actually, these were debian packages) for production environment.
Mirek
|
|
|
Re: 2024rc2 [message #61066 is a reply to message #61063] |
Tue, 05 November 2024 09:59   |
 |
mirek
Messages: 14260 Registered: November 2005
|
Ultimate Member |
|
|
cbpporter wrote on Tue, 05 November 2024 08:54After some more testing, not all the awkwardness of the editor is done.
Example, open up AddressBook, go to a function where "array" is defined, type array, then '.', autocomplete popup, press down arrow to navigate the list, press tab, the editor add a tab instead of auto-complete. Other scenarios involving tab work.
Uhm, and what should have happened? Should tab insert auto-complete now? (I have no problem with that if that is considered standard in other editors, but it was never this way).
Mirek
|
|
|
Re: 2024rc2 [message #61067 is a reply to message #61066] |
Tue, 05 November 2024 14:55   |
cbpporter
Messages: 1427 Registered: September 2007
|
Ultimate Contributor |
|
|
Yes, it should insert the autocomplete. In fact it does, if in the scenario I described, you do not press down arrow, but go directly for tab. Interacting with the drop down removes the tab behavior.
Also, is it normal for the Automatic build methods, when not dealing with clang, to not pick up MSC. Unfortunately, of fortunately if you will, our tools have inherited TheIDE style autodetection and both seem broken.
I'm working on a new solution that autodetects and uses "c:\Program Files (x86)\Microsoft Visual Studio\Installer\vswhere.exe" . That should be more robust and avoid the extensive searches employed today? Not even MS can figure out where they put VS...
PS: How on God's earth have you managed to integrate Clang. I'm not using lib-clang, but lib-llvm and it is impossible to link with it. And trying to download a precompiled version that matches what I have is a nightmare...
|
|
|
|
|
Re: 2024rc2 [message #61070 is a reply to message #61068] |
Tue, 05 November 2024 22:28   |
cbpporter
Messages: 1427 Registered: September 2007
|
Ultimate Contributor |
|
|
Klugier wrote on Tue, 05 November 2024 21:01Hello cbpporter,
Thank you for your feedback! However, it is more about features request rather than fixing existing problems. So, we shouldn't address it right now and create next rc candidate after that.
For vswhere. Thansk for providing that information it is nice to know that something like that exist. I agree that we should launch this binary and then analyze the output in the first place. It looks like the installation path is there:
resolvedInstallationPath: C:\Program Files\Microsoft Visual Studio\2022\Community
However, what about if you install Visual Studio Installer to the diffrent path. Then launching vswhere will be problematic. Also, I am afraid that after some time M$ could change the path to Program Files.
For 'tab' and autocomplete in Assist++ dialog. As Mirek noticed, we didn't support it in the past. However, after verifying with other IDE's I agree that should normalized the behavior and allow tab to act as enter.
Klugier
Mine is installed in weird locations and the tool does report the correct paths. First we should check all possible program files. That's why I want to prototype a new detect using vswhere. DirFinder solution no longer works.
Edit: with how iffy VS is, doesn't hurt to have a fallback or two.
[Updated on: Tue, 05 November 2024 22:40] Report message to a moderator
|
|
|
Re: 2024rc2 [message #61071 is a reply to message #61070] |
Tue, 05 November 2024 23:05   |
 |
mirek
Messages: 14260 Registered: November 2005
|
Ultimate Member |
|
|
cbpporter wrote on Tue, 05 November 2024 22:28Klugier wrote on Tue, 05 November 2024 21:01Hello cbpporter,
Thank you for your feedback! However, it is more about features request rather than fixing existing problems. So, we shouldn't address it right now and create next rc candidate after that.
For vswhere. Thansk for providing that information it is nice to know that something like that exist. I agree that we should launch this binary and then analyze the output in the first place. It looks like the installation path is there:
resolvedInstallationPath: C:\Program Files\Microsoft Visual Studio\2022\Community
However, what about if you install Visual Studio Installer to the diffrent path. Then launching vswhere will be problematic. Also, I am afraid that after some time M$ could change the path to Program Files.
For 'tab' and autocomplete in Assist++ dialog. As Mirek noticed, we didn't support it in the past. However, after verifying with other IDE's I agree that should normalized the behavior and allow tab to act as enter.
Klugier
Mine is installed in weird locations and the tool does report the correct paths. First we should check all possible program files. That's why I want to prototype a new detect using vswhere. DirFinder solution no longer works.
Edit: with how iffy VS is, doesn't hurt to have a fallback or two.
Proper PR with better VS detection will be accepted fast.. That said, not a high priority for me - I mean, I do not plan to work on this personally.
|
|
|
Re: 2024rc2 [message #61072 is a reply to message #61071] |
Wed, 06 November 2024 14:42   |
cbpporter
Messages: 1427 Registered: September 2007
|
Ultimate Contributor |
|
|
mirek wrote on Wed, 06 November 2024 00:05
Proper PR with better VS detection will be accepted fast.. That said, not a high priority for me - I mean, I do not plan to work on this personally.
That will take some time...
I have 2 PCs, same VS, one is detected, the other not. Mysterious. Think will be very hard without making a VM with WIN so I can rollback installs...
Until then I can test what the machine that detects VS says if I feed it the vswhere input...
Edit: I spoke too soon.
One doesn't detect, the other detects:
"c:\program files (x86)\microsoft visual studio\2019\professional\vc\tools\msvc\14.29.30133\bin\amd64 \cl.exe "
This folder does not exist. Probably the code should not detect files that do not exist.
The correct path is:
"c:\Program Files (x86)\Microsoft Visual Studio\2019\Professional\VC\Tools\MSVC\14.29.30133\bin\Hostx 64\x64\cl.exe "
vswhere.exe reports:
installationPath: C:\Program Files (x86)\Microsoft Visual Studio\2019\Professional
Think I'll stop talking about this here, not to hijack the thread and open a new thread.
Edit 2: now it finds the good path... Yeah, new thread .
[Updated on: Wed, 06 November 2024 14:59] Report message to a moderator
|
|
|
|
|
|
|
|
|
Goto Forum:
Current Time: Sun Jun 01 16:15:09 CEST 2025
Total time taken to generate the page: 0.02743 seconds
|