PureBasic Forums - English

RASHAD wrote: Sat Jul 20, 2024 10:44 pm Hi Randy

Good to see you and netmaestro are still kicking around on Purebasic.
You've both been a great help to me. Thanks!!

Randy Walker wrote: Sun Jul 21, 2024 6:55 am

BarryG wrote: Sun Jul 21, 2024 3:33 am No, it's a real issue. Lots of items from pre-6.10 need updating from ".l" to ".i" type.

Here's my example from today -> https://www.purebasic.fr/english/viewto ... 96#p624596

Maybe that's why I gave up trying to use 6.xx
27000+ lines in my main program alone, was just way too much to troubleshoot.

.l/.i is not a matter of PureBasic version, it's a problem of programming style. .i has been the choice of pointers since there were 64 bit versions. I don't think it's good to stay with an older version just because it's comfortable.

BarryG wrote: Sun Jul 21, 2024 3:33 am

jacdelad wrote: Sat Jul 20, 2024 10:47 pmThis seems to become some kind of running gag

No, it's a real issue. Lots of items from pre-6.10 need updating from ".l" to ".i" type.

Here's my example from today -> https://www.purebasic.fr/english/viewto ... 96#p624596

I know it's a real problem...for the ones who used the pointers wrong since there were 64 bit versions introduced.

PB 6.1x just highlight a real programming error which should be corrected and not ignored. The fact it worked before doesn't mean it was right.

Let's just remove the long data type for a year or so and give it back afterwards.

Using long for addresses in 64 bit processes was wrong beforehand, just luckily didn't lead to an error due to memory positioning.

If you copy codes from 200X it is up to you to check how hardware and software evolved since then. Also if you switch the compiler from 32bit to 64bit. If you used the codes from back then it could potentially be that the developers only had 32bit computers.

jacdelad wrote: Sun Jul 21, 2024 11:02 am I don't think it's good to stay with an older version just because it's comfortable.

Appears unlike yourself, I am not a Neophiliac. I don't need it just because it's new. I need it because it's tried and true. BTW -- My best machine is still a Windows 7

Randy Walker wrote: Sun Jul 21, 2024 8:25 pm

jacdelad wrote: Sun Jul 21, 2024 11:02 am I don't think it's good to stay with an older version just because it's comfortable.

Appears unlike yourself, I am not a Neophiliac. I don't need it just because it's new. I need it because it's tried and true. BTW -- My best machine is still a Windows 7

PB 6.11 contains a ton of bugfixes which also affect 5.40.
Windows 7 doesn't get updates for ages now. Good luck securing it...

mk-soft wrote: Sat Jul 20, 2024 11:11 am I think it's more likely that it's the Windows API you're using that works differently under Win 11.
Which API did you use? It will not be due to the PureBasic functions.

I did NOT suggest or in any way imply it was the fault of anything Purebasic, per se. But I was only seeing this behavior in my main PB project, and after scouring the web I could find no clue as to what had happened or how to fix it, so I came here thinking/hoping maybe one of the PB enthusiasts might have run across a similar problem and worked out a solution they might share. Whatever the case, Thank you all for reading my OP.

DarkDragon wrote: Sun Jul 21, 2024 7:35 pmLet's just remove the long data type for a year or so and give it back afterwards.

Nope, some structures still require an actual long and not integer type.

@BarryG
Absolutely you are right
It is Not a GAG

Nope, some structures still require an actual long and not integer type.

RASHAD wrote: Mon Jul 22, 2024 5:46 am @BarryG
Absolutely you are right
It is Not a GAG

Nope, some structures still require an actual long and not integer type.

It's not a gag but it's sarcasm.

Do a search for ".l" in your code, what's the problem?

For example
https://learn.microsoft.com/en-us/windo ... gdi-bitmap
This is MS and it is the same using PB Structure viewer
Now the PB Pro said it is a gag or sarcasm
2000 posts and more and nothing but Bla Bla

AZJIO wrote: Mon Jul 22, 2024 6:31 amDo a search for ".l" in your code, what's the problem?

See my post a little bit up. You can't just remove all ".l" to make everything work. I once thought that way, until I was corrected.

BarryG wrote: Mon Jul 22, 2024 6:47 am

AZJIO wrote: Mon Jul 22, 2024 6:31 amDo a search for ".l" in your code, what's the problem?

See my post a little bit up. You can't just remove all ".l" to make everything work. I once thought that way, until I was corrected.

Yeah you actually have to understand your own code. Or you could do unit testing. Wasn't there some PureUnit or so somewhen?

BarryG wrote: Mon Jul 22, 2024 6:47 am

AZJIO wrote: Mon Jul 22, 2024 6:31 amDo a search for ".l" in your code, what's the problem?

See my post a little bit up. You can't just remove all ".l" to make everything work. I once thought that way, until I was corrected.

I didn’t say that it needs to be removed or replaced. I said we need to find it and make a decision to fix it or leave it. I don't think you have a thousand variables. Notepad++ and AkelPad have functionality for displaying strings with found occurrences. You have received a list and click on a line to go to it and make edits.
After half an hour you get a working code
screenshot in AkelPad
screenshot in Notepad++
screenshot in FindAllReferences