My wishlist for PureBasic 4.50

Kale · Post by **Kale** » Mon Jul 27, 2009 6:24 pm

My wishlist for v4.5

True OOP catered for! yay!

IceSoft · Post by **IceSoft** » Mon Jul 27, 2009 6:25 pm

Kale wrote:My wishlist for v4.5

True OOP catered for! yay!

LOL. Please wait...this is on my whishlist for 5.0

Kale · Post by **Kale** » Mon Jul 27, 2009 7:00 pm

IceSoft wrote:
Kale wrote:My wishlist for v4.5

True OOP catered for! yay!
LOL. Please wait...this is on my whishlist for 5.0

cool!

IceSoft · Post by **IceSoft** » Mon Jul 27, 2009 7:19 pm

Kale wrote:
IceSoft wrote:
Kale wrote:My wishlist for v4.5

True OOP catered for! yay!
LOL. Please wait...this is on my whishlist for 5.0
cool!

Ooops...sorry I forgot a '0'...wrong wishlist version... I meant v50.0

Post by **freak** » Mon Jul 27, 2009 7:52 pm

Actually there is no real need to drop support for these old OS so far. So long as the code we write keeps working on them without too much extra effort, i do not see why we should not support them.

Droping OS support just to update to a new linker? Its not like the process of linking executable has evolved much in the past years, so as long as there are no problems with it its hardly enough of a reason to throw support for these OS out of the window.

IceSoft · Post by **IceSoft** » Mon Jul 27, 2009 8:15 pm

@freak,
Good spoken.

And that is what I want to add to your statement:
As long PureBasic has 'up-to-date' third party library versions as long there is no reason to break the support for any (old) OS.

'Up-to-date' means not only the LATEST feature version...it means more a stable and good working version of a third party library which is used on the LATEST PureBasic release.

Post by **freak** » Mon Jul 27, 2009 8:18 pm

IceSoft wrote:'Up-to-date' means not only the LATEST feature version...it means more a stable and good working version of a third party library which is used on the LATEST PureBasic release.

Do you have any particular problems with the current linker version ?

IceSoft · Post by **IceSoft** » Mon Jul 27, 2009 8:38 pm

freak wrote:
IceSoft wrote:'Up-to-date' means not only the LATEST feature version...it means more a stable and good working version of a third party library which is used on the LATEST PureBasic release.
Do you have any particular problems with the current linker version ?

No. Not with the linker (In the past I used mostly the latest version from the PellesC web page)

Rescator · Post by **Rescator** » Wed Jul 29, 2009 5:03 pm

freak wrote:Actually there is no real need to drop support for these old OS so far. So long as the code we write keeps working on them without too much extra effort, i do not see why we should not support them.

True, but I certainly would be ok to see PureBasic 5.0 supporting only NT 5.x (W2K) and later, this would mean no more Ascii and only Unicode as well.

iostream · Post by **iostream** » Thu Jul 30, 2009 12:56 pm

For me a important wish:
Support ExamineFTPDirectory() for servers with Windows OS.

swhite · Post by **swhite** » Tue Aug 04, 2009 1:19 am

I vote for OOP with multiple inheritance as well.

Simon

PB · Post by PB » Tue Aug 04, 2009 11:22 am

Kale's OOP request was a joke, from what I can tell from using a smiley,
because it's been discussed over and over before and it's not happening.
So newbies: forget it. Don't hold your breath for it. Thanks for reading.

Kale · Post by **Kale** » Tue Aug 04, 2009 3:35 pm

PB wrote:Kale's OOP request was a joke, from what I can tell from using a smiley,
because it's been discussed over and over before and it's not happening.
So newbies: forget it. Don't hold your breath for it. Thanks for reading.

he he... in all honesty fully implementing OOP is the next logical step for PB but like you say i can't see it happening.

I am hoping Fred is going to verge off and start another language embracing the OOP paradigm.

I know everyone will take this the wrong way but OOP is the only choice for large serious programs.

...and before the flames, i agree, i can't see PB going OOP in the near future. In fact i think i read somewhere in another thread that Fred wouldn't even know where to begin implementing OOP in PB (me neither), it would take a massive redesign and a huge amount of learning on his part.

srod · Post by **srod** » Tue Aug 04, 2009 4:08 pm

Aye, must admit that having recently finished a PB application consisting of about 50000 lines of code; I really couldn't imagine having done that without using OOP. My procedural code generally begins to look like the bottom of my kettle after a few thousand lines!

The truth is though that I like the very simple way of using basic OOP through PB's interfaces etc. and whilst I wouldn't complain about having a few more OOP features added (constructors and the like) I really don't think the absence of such features prevents the effective use of OOP at all. Having to create the VTs ourselves just makes things so much clearer to me and avoids a lot of the unnecessary features I see added to many OOP implementations. But it is like a lot of programming in general; you like a particular language enough then you will always find ways of accomplishing whatever it is you need etc.

Still, horses for courses and all that. This topic has of course been done to death.

Kale · Post by **Kale** » Tue Aug 04, 2009 4:38 pm

srod wrote:The truth is though that I like the very simple way of using basic OOP through PB's interfaces etc. and whilst I wouldn't complain about having a few more OOP features added (constructors and the like)

Me too, i kinda like the way it is but i do wish for stuff like better encapsulation, some sort of constructor and inheritance.