PureBasic Forums - English

Fred wrote:

StarBootics wrote:Hello everyone,

I have some questions about some stuff since C compilers are capable to deal with :

• 1. unsigned long and unsigned long long (sorry to ask again)
  2. Static arrays outside structures
  3. Multi-dimensions static arrays inside and outside structures
  4. The possibility initialize static arrays like : ArrayName.f[4] = {0.0, 1.0, 2.0, 3.0}
  5. The ternary operator

Will all this be introduced into the PureBasic language

Best regards
StarBootics

No new features will be added to PureBasic. It's just a new back-end to be able to mainly support new CPU (I wasn't even sure to support Windows at first).

Thanks for the answer.

Best regards
StarBootics

Features will only be removed

Well.. hope there will be some syntax evolution (not talking about libraries). A (lack of a) few core language features are really painful and twisted and should be prioritized to get fixed/added IMHO (time/money/complexity).

@Saki & dpi: so true. You have to fight both the win32 api and the gadget library.

Demivec wrote:

normeus wrote:

Fred wrote:(I wasn't even sure to support Windows at first).

Using google translate it sounds like he was thinking about changing the back end for Linux and Apple and leaving windows with the same back end, but don't quote me on that.

@normeus: I think Fred was referring to the distant past when PureBasic existed on the Amiga only.

Note, please excuse me for quoting you. In my defense I didn't notice you wrote about that until after I had finished responding.

Sorry my statement wasn't clear enough, I meant: "I wasn't sure to ship the C back-end with the Windows version as there is no real need for it". After thoughts, it will make more sense to have all the PB versions to have access to this backend, if you write inline C code for example, and also to get more optimized output.

My question was not answered.

useful wrote:Very interested in the answer to an important question.
It is especially important for the linux ecosystem.
Will it be possible to outsource the generated C source to be built with gcc without purebasic?
p.s. I mean the ability to generate standalone deb-src (rpms)

Fred wrote:I wasn't sure to ship the C back-end with the Windows version as there is no real need for it

https://docs.microsoft.com/en-us/windows/arm/

Fred wrote:"I wasn't sure to ship the C back-end with the Windows version as there is no real need for it".

At this moment SB's gadgets are limited (both in range and functionality) to maintain compatibility with PB.
Is it possible this could lead to a similar dynamic between PB[C] and PB[ASM]?

If so maybe PB[ASM] should be left as an entry level product with its lifetime license and PB[C] introduced as an entirely new product with a different licensing model and no obligation to backport new functionality to PB[ASM].

@ Rinzwind
Yes, it is a horror with DPI aware
The problems really become visible on 4K monitors.
On a standard monitor, people usually don't use more than 125%, so you often don't see when something doesn't fit.
Also, if you don't use floats, you'll just round and it will be 1 instead of 1.25.
On a 4K monitor people often use 200% or more.
You can see problems brutally.
If a user sets intermediate values and the developer neglects to use floats, nasty and hard-to-locate errors can result.
I once had someone point out that he gets problems with 4K.
He said he didn't care about my 4K monitor.
That's the same as saying I don't care if my software runs on the users or not.
I had already given up on porting an older, very graphics-heavy code to DPI aware.
Another problem is also overlooked with 4K usage.
Graphics become much larger internally and animations quickly go down the tubes. Such applications become hardly usable or there is misbehavior when working with timers, things then can no longer be done in defined time spans.

Hi Fread and Team,


Now that PB will be closer to C, is it possible to add some useful basic C functions like formatting and string analysis.
(i.e: printf, sprintf, fprintf, scanf, sscanf, fscan

Also some system functions would be welcome for example on process management (like Tasklist and Taskkill tools (windows) ps/kill on linux)

Marc56us wrote:Now that PB will be closer to C, is it possible to add some useful basic C functions like formatting and string analysis.
(i.e: printf, sprintf, fprintf, scanf, sscanf, fscan

like so?

Somehow it's like a coop of startled chickens here.

Marc56us wrote:
Now that PB will be closer to C, is it possible to add some useful basic C functions like formatting and string analysis.

For me, Fred's statement is not that difficult to understand:

Fred wrote:No new features will be added to PureBasic. It's just a new back-end to be able to mainly support new CPU

It will surely take some time until PureBasic runs as stable under C as it is now. This should be the first and most important step, after that there is still enough time to think about extensions (in my humble opinion).

Josh wrote:Somehow it's like a coop of startled chickens here.

Marc56us wrote:
Now that PB will be closer to C, is it possible to add some useful basic C functions like formatting and string analysis.

For me, Fred's statement is not that difficult to understand:

Fred wrote:No new features will be added to PureBasic. It's just a new back-end to be able to mainly support new CPU

It will surely take some time until PureBasic runs as stable under C as it is now. This should be the first and most important step, after that there is still enough time to think about extensions.

Let me speak please, this is just an idea that was proposed before and can be more easily implemented now.

I know that Fred is doing it his way and the new functions will come later.

This suggestion can therefore be moved up several levels in the Todo list (IMHO)
We use these functions a lot in C, it can bring new customers to PB

First of all, it has to run stably.
How is that supposed to be debugged if buckets of new functions are to be added now?
That all will take its time.
By then, all those wishes will have sunk into the depths of the forum here.
So you should really wait until that time comes.

Saki wrote:First of all, it has to run stably.

Comments like this make me a bit scared/worried to update. I might have to stick with v5.73 for a very long time.

Although, I wonder if the number of virus false-positives may suddenly reduce/disappear due to the new code base?

At a minimum, I expect inline C to be a new feature.

BarryG wrote:Although, I wonder if the number of virus false-positives may suddenly reduce/disappear due to the new code base?

It wont fix broken AV software thats for sure.