C++ Version of POI

2002-05-07 Thread KG, Chandrasekhar
People, I have been following the poi-dev mailing list over the last few days.. saw a post about interest in developing a c++ port of POI (TJ). And was generally excited about it.. I call it poi++.. I saw the subsequent posts about C being a better alternative.. I disagree to it.. since when

Re: C++ Version of POI

2002-05-07 Thread Andrew C. Oliver
I'm not particularly excited about it. C++ is a detestible language and will always be one and you'll never convince me otherwise. That being said, nice work. Consider me marie antonette. let them eat cake, even if they like bitter nastly dry brittle cake. So in short I'll commit this.

Re: C++ Version of POI

2002-05-02 Thread Nicola Ken Barozzi
From: Andrew C. Oliver [EMAIL PROTECTED] It would be great to have an abstract code generator that would have front-ends to generate Java, C, and C++. I haven't studied the code base enough to know if this is feasible, but I would imagine we could generate the code to access many of the

Re: C++ Version of POI

2002-05-02 Thread Andrew C. Oliver
Nicola Ken Barozzi wrote: From: Andrew C. Oliver [EMAIL PROTECTED] It would be great to have an abstract code generator that would have front-ends to generate Java, C, and C++. I haven't studied the code base enough to know if this is feasible, but I would imagine we could generate the

Re: C++ Version of POI

2002-05-02 Thread T.J. Mather
On 1 May 2002, Andrew C. Oliver wrote: We already DO do XML-Java code generation. So you do XML-Java using XSLT for HSSF and HDF. This is awesome!!! Much cleaner than other code generation scripts I've seen (e.g. the Perl script for wv2). How much of the code base do you anticipate will

Re: C++ Version of POI

2002-05-02 Thread Andrew C. Oliver
T.J. Mather wrote: On 1 May 2002, Andrew C. Oliver wrote: We already DO do XML-Java code generation. So you do XML-Java using XSLT for HSSF and HDF. This is awesome!!! Much cleaner than other code generation scripts I've seen (e.g. the Perl script for wv2). How much of the code base do

Re: C++ Version of POI

2002-05-01 Thread Marc
Rainer Klute wrote: I think its a waste of time to do a C\C++ version of POI... Maybe, maybe not. It would broaden POI's appeal if it could be ported to C or (gag) C++. I wouldn't mind doing a C port of POIFS, but I have taken a vow of celibacy wrt C++ -- I'll never get screwed

Re: C++ Version of POI

2002-05-01 Thread Nicola Ken Barozzi
From: Andrew C. Oliver [EMAIL PROTECTED] Rainer Klute wrote: I think its a waste of time to do a C\C++ version of POI... +1. Fiddling around with doing in C/C++ the same things we have done in Java already only distracts us from getting POI more feature-complete. (A much more

Re: C++ Version of POI

2002-05-01 Thread Avik Sengupta
Besides, a C\C++ implementation would be focused towards linux\unix community because M$ already exposes this functionality in Windows I dont have too strong an opinion on whether it would be a wasted effort or not re the other impl of ole2, but i would like to point out that some people

Re: C++ Version of POI

2002-05-01 Thread Andrew C. Oliver
Oh man thats to funny. If they do it lets try and engage them. Thats like great PR for the project. If they keep their implementation of POI open we can reference it from the home page. Of course I think I'd rather see them call the Java stuff from J# or what-have-you as that would

Re: C++ Version of POI

2002-05-01 Thread Avik Sengupta
Really I'm not sure coming to think of it why they'd wan't to convert it to J#, it would be easier to create a J# wrapper to call POI Many reasons, most having to do with getting people to install a JVM on a production server with a large .NET app running. Like most programmers, they would

Re: C++ Version of POI

2002-05-01 Thread T.J. Mather
On 30 Apr 2002, Andrew C. Oliver wrote: I think the strong points for C++ are: * Much easier to port from Java, since C++ is much closer to Java than C I disagree. An adequate code generator will make this irrelevant. Besides any such advantage will be greatly exceeded by increased

Re: C++ Version of POI

2002-05-01 Thread Andrew C. Oliver
It would be great to have an abstract code generator that would have front-ends to generate Java, C, and C++. I haven't studied the code base enough to know if this is feasible, but I would imagine we could generate the code to access many of the Excel and Word datatypes is this manner.

C++ Version of POI

2002-04-30 Thread T.J. Mather
Would there be any interest in having a C++ library developed in parallel with the Java POI libraries, including POIFS, HSSF, HDF? The idea would be similar to how Xerces C++ and Java work, sharing a similar API and code structure. This way we would use POI in the various open-source office

Re: C++ Version of POI

2002-04-30 Thread Nicola Ken Barozzi
From: T.J. Mather [EMAIL PROTECTED] Would there be any interest in having a C++ library developed in parallel with the Java POI libraries, including POIFS, HSSF, HDF? The idea would be similar to how Xerces C++ and Java work, sharing a similar API and code structure. This way we would use

Re: C++ Version of POI

2002-04-30 Thread Andrew C. Oliver
Hi TJ -- I got Ken's message first so I replied to you inline with it and him at the same time.. see it. T.J. Mather wrote: Would there be any interest in having a C++ library developed in parallel with the Java POI libraries, including POIFS, HSSF, HDF? The idea would be similar to how

Re: C++ Version of POI

2002-04-30 Thread T.J. Mather
On Tue, 30 Apr 2002, Andrew C. Oliver wrote: Which are largely written in C and not C++ Why C and not C++? C++ is object oriented, and is much closer to Java than C is. And proponents of the GPL generally feel that APL software is *legally incompatible* I hope we can work out the

Re: C++ Version of POI

2002-04-30 Thread Andrew C. Oliver
On Tue, 2002-04-30 at 16:27, T.J. Mather wrote: On Tue, 30 Apr 2002, Andrew C. Oliver wrote: Which are largely written in C and not C++ Why C and not C++? C++ is object oriented, and is much closer to Java than C is. C++ is NOT very object oriented but thats all philosophy.

Re: C++ Version of POI

2002-04-30 Thread Ryan Ackley
I think its a waste of time to do a C\C++ version of POI and this is why. Its been done before. Of course it hasn't been done as well as its been done by our group, nevertheless I think the need has been recognized and met by other groups. I appreciate TJ's enthusiasm for our product. I would

Re: C++ Version of POI

2002-04-30 Thread Andrew C. Oliver
On Tue, 2002-04-30 at 18:28, Ryan Ackley wrote: I think its a waste of time to do a C\C++ version of POI and this is why. Its been done before. Of course it hasn't been done as well as its been done by our group, nevertheless I think the need has been recognized and met by other groups. I

Re: C++ Version of POI

2002-04-30 Thread T.J. Mather
Well I think there are strong points for C here: * Compatility * Usage from C applications * Support from Andy et al. I think we can get around the compatibility by using a Portable subset of C++. This is what many good C++ projects (Xerces, KDE, QT) do. The usage from C applications is a

Re: C++ Version of POI

2002-04-30 Thread Ryan Ackley
values. I have never seen a C or C++ program that was easy to maintain. - Original Message - From: Andrew C. Oliver [EMAIL PROTECTED] To: POI Development [EMAIL PROTECTED] Sent: Tuesday, April 30, 2002 8:03 PM Subject: Re: C++ Version of POI On Tue, 2002-04-30 at 20:23, T.J. Mather wrote

Re: C++ Version of POI

2002-04-30 Thread Andrew C. Oliver
the Java class libraries. (Ask Sam Ruby about that and .NET) -Andy - Original Message - From: Andrew C. Oliver [EMAIL PROTECTED] To: POI Development [EMAIL PROTECTED] Sent: Tuesday, April 30, 2002 8:03 PM Subject: Re: C++ Version of POI On Tue, 2002-04-30 at 20:23, T.J. Mather

Re: C++ Version of POI

2002-04-30 Thread Rainer Klute
I think its a waste of time to do a C\C++ version of POI... +1. Fiddling around with doing in C/C++ the same things we have done in Java already only distracts us from getting POI more feature-complete. (A much more intellectual challenging project would be to write a Java - C/C++ compiler