Re: Feasibility/Requirements/Wishes of xAL (was: Standard Library Interest?)

[Marin David Condic <nobody@noplace.com>]

Well, O.K. If what you're saying is that there would be some 
organization producing a "Common" Ada library (or whatever it is 
ultimately called) and you expect that there are people who would want 
to contribute pieces - such as bindings, etc. - then yes, I can see a 
need to manage the names. But I'd think that we'd want to first see some 
notion of general acceptance of the concept of a CAL (or insert name 
here) and how it would be produced. Whoever is doing the publishing 
could worry about name allocation.

MDC


Robert I. Eachus wrote:
> 
> 
> First notice that without a group working on CAL or whatever name is 
> chosen, this isn't much use at all. The bit about avoiding name 
> conflicts might be helpful in a vacuum, but that is about it.
> 
> However, if we do have a serious collaborative library building effort 
> going on, or several of them, then having the registrar in place and 
> working deals with several major problems and has one side benefit.
> 
> The major problem is that there are names and areas which are potential 
> minefields.  For example, GNADE uses ODBC.  ODBC is a Microsoft 
> standard.  It would be really nice to name the package that supplies the 
> thin binding to ODBC, something.ODBC, and have Microsoft's permission to 
> do so.  This is part of what the registry is there to accomplish, just 
> like the process for registering domain names on the Internet.
> 
> In addition there are some names which may be highly prized and others 
> which should be used for implementations of particular bindings or 
> libraries.  I don't care if CAL has a CAL.BLAS which has nothing to do 
> with the Basic Linear Algebra System. (Well, I think it would be silly, 
> and if I got a vote inside CAL I'd probably vote against it.) But I 
> think any "root" library named BLAS would only be given to a (complete, 
> working) BLAS implementation.  Similarly I can't see any legitimate 
> standards group approving the name GtkAda for anyone except the authors 
> of GtkAda.
> 
> Now comes the side benefit...  There are certain Microsoft bindings that 
> won't get into CAL or any such library without Microsoft's blessing.  To 
> the extent that Microsoft wants to control such bindings, they will. But 
> if there is a registry in place and well managed, it will be a lot 
> easier to get sensible Ada bindings to Microsoft (and other company's) 
> interfaces registered and with sensible copyright restrictions.  No, 
> that doesn't mean that we can talk Microsoft into going to open source 
> licensing for everything.  But it should mean that they will agree to 
> have the package spec free for copying without modification, even if 
> Microsoft wants you to buy their implementation.  (But I expect that 
> Microsoft will be willing to make any necessary Microsoft bindings 
> available from their website.  They have in the past.)
> 
> I used Microsoft to be specific.  But there are other companies such as 
> Oracle, Sun, and even ACT where before we are done we may have to deal 
> with trademark and copyright issues.  Do it right from the start, and 
> there may even be funding from someone other than the compiler 
> companies.  But if we don't start out with the right intentions--to do 
> our best to acknowledge and protect copyrights and trademarks without 
> making the libraries impossible to use--then we will fail.


-- 
======================================================================
Marin David Condic
I work for: http://www.belcan.com/
My project is: http://www.jsf.mil/NSFrames.htm

Send Replies To: m c o n d i c @ a c m . o r g

     "All reformers, however strict their social conscience,
      live in houses just as big as they can pay for."

          --Logan Pearsall Smith
======================================================================