Re: Standardisation between Win libraries
- Posted by Derek Parnell <ddparnell at bigpond.com> Jul 25, 2003
- 569 views
----- Original Message -----=20 From: "Dave Probert" <zingo at purpletiger.com> To: "EUforum" <EUforum at topica.com> Subject: RE: Standardisation between Win libraries >=20 >=20 > Phew, I guess it's more difficult than I first thought. Like most things in life =20 > So, basically we're looking at many MS Windows (ignoring Linux at the=20 > moment - sorry) libraries, that each do much the same thing, but with=20 > differences (bits missed, etc). Each of them written to manage the=20 > tasks that the writer wanted to do at the time, but not complete = enough=20 > for a wider range of tasks. Each of them sits somewhere on the = Euphoria=20 > Path and there can, potentially, be many copies of them (Lots of = people=20 > supply a copy of win32Lib, ewin32API, win32R, etc; along with their=20 > application) - I already have 12 copies of win32lib that came in = various=20 > zips. Yep, that about sums it up. =20 > I agree with Derek on the sensibilities of not simply including ALL = the=20 > function even if not needed. I'm not sure about the absolute need for = > cross-platform All-In-One type of library development - shouldn't = there=20 > be a wrapper or different set of (same name) files to be included when = > going to another platform? Ahhh..the perfect world concept again...hmmmm. > A more modular and heirarchical approach to the library designs is=20 > probably what I'm trying to describe. One that would allow the base=20 > functionality to be there, but with the ability to extend it beyond = that=20 > in various ways without touching (ie modifying) the base code. = Nothing=20 > new there - been done before in many languages. Yep, it sure has. The Euphoria user base is so small and uncoordinated = (read: not-paid) that a central Eu standards body is not all that = feasible. And RDS does not want that role either. > The Standard Euphoria Library does seem to be dead, but it's a sound=20 > idea - especially for Windows development. Looking through many files = I=20 > see repetition of so many functions and slight variations on functions = > (eg. or_all() and or_all_bits() ) - that strikes me as simply a lack = of=20 > some further core libraries which we all could benefit from. Hear, hear! One of the first things that new Eu coders find, after = getting over the intro stuff, is that they have to go an reinvent many = things that are taken as normally available in other languages. How many = copies of abs() do we really need!? RDS should be a lot more proactive = in packaging the commonly re-invented library routines - even if it = means that RDS takes responsiblility in maintaining them. The continual = mess of amorphous sub-committees for "Eu Standard Library" is adding to = the frustration. > What are the benefits of each of the current Windows libraries anyway? Win32lib attempts to make coding apps for MS-Windows easier and faster = to do; trading execution speed for development speed. > What are the differences? =20 I think that other WIN libraries either go for faster execution speeds = and/or cross-platform functionality. >Which Library is the most used? =20 Win32lib, of course >What can be done to help improve (and/or merge ) the libraries=20 >(or preferably one) Nothing, as they have different (mutually exclusive) goals to achieve. > Still not an attack, just a discussion from a new Euphoria user who's = a=20 > little confused, having come from the C/C++/Java/PHP/ActionScript=20 > world!! :) And in spite of this lack of 'core' functionality, I still find myself = liking Euphoria more and more. 'Chaos Theory' in action. --=20 Derek