[sane-devel] Sane API Suggestion
Major A schrieb: The SANE API is designed to be simple and portable, and any change towards object orientation or functional programming would only bloat it unnecessarily. Sorry, if you want to interface Scheme to SANE, you'll have to adapt to the way things are. Well, changing the current API is indeed not a good idea. For newer versions, we should be a bit more open, IMHO. Adding a callback data argument to both sane_init() and SANE_Authorization_Callback would cause a number of problems, in particular that of network transparency. Scheme or C++ programmers would certainly like to use that parameter as a pointer to a structure, but that only works if the size of the argument is at least that of void*. Since the latter varies across platforms, this would be ugly and non-portable. I don't think that network transparency introduces a problem for user data pointer sizes. Since Sane itself does not deal with this pointer, the net backend does not need to send and receive the correct value of this pointer to/from the Sane server. Instead, the net backend can use the pointer value as passed by the application in the callback. Since the only callback in SANE is the one responsible for authentication, I'd say it wouldn't be worth changing anything. A Scheme frontend would either have to relay the authentication request to the user, in which case the callback data is not used, or look up username and password in a database, in which case the resource string supplied is sufficient. Well, some time has passed sine my last look into the Sane 2 API docs, but IIRC it has at least one more callback. Abel
[sane-devel] Sane API Suggestion
I don't think that network transparency introduces a problem for user data pointer sizes. Since Sane itself does not deal with this pointer, the net backend does not need to send and receive the correct value of this pointer to/from the Sane server. Instead, the net backend can use the pointer value as passed by the application in the callback. OK, you're right. Back to the original question: I just realized that the callback function is set when you call sane_init(), thus it must be the same for all scanner instances. Therefore I can see little point in having any kind of user_data since even a Scheme interface to SANE would only use sane_init() once so it could store the user_data itself just the same. (Unless, of course, the Scheme interface would allow for more than one backend to be used at the same time, but with dynamic loading etc. this would soon become a mess anyway.) Well, some time has passed sine my last look into the Sane 2 API docs, but IIRC it has at least one more callback. I can't remember any. The only thing that is similar in a way is the file descriptor returned by sane_get_select_fd(), but it's not a callback. Andras === Major Andras e-mail: and...@users.sourceforge.net www:http://andras.webhop.org/ ===
[sane-devel] Sane API Suggestion
The initial version of my based application (written in scheme) does not use the sane api directly, but utilizes scanimage and scanadf to perform scanning. In a future release, I would like to use the sane api directly for better control. Although I have a number of years of experience programming C/C++/Java, I now prefer to use scheme for most of my programming. After briefly looking at the sane api, I noticed that callbacks are sometimes used. I am aware of a number of ways and various ffi libraries to handle callbacks from non-C to C, but it sure would be nice if the sane API was a little more non-C friendly. Specifically, when a callback pointer is required, why not also require a user data void pointer (or perhaps an integer id) that would be returned untouched to the callback as a parameter? This would allow non-C languages to interface more easily to the sane api. A thin C interface layer could then use one callback dispatcher to handle all C callbacks. The callback dispatcher could use stdarg.h to process the other args or perhaps they could be passed to the callback using a count and array. In the future, I would be happy to contribute a scheme-sane interface library to the sane project, but it sure would be nice if the sane api was more non-C friendly. Peter
[sane-devel] Sane API Suggestion
Hi, On Sun, Jul 20, 2003 at 08:48:54AM -0400, Peter Santoro wrote: Although I have a number of years of experience programming C/C++/Java, I now prefer to use scheme for most of my programming. After briefly looking at the sane api, I noticed that callbacks are sometimes used.
[sane-devel] Sane API Suggestion
Hi, I forgot to add: On Sun, Jul 20, 2003 at 08:48:54AM -0400, Peter Santoro wrote: In the future, I would be happy to contribute a scheme-sane interface library to the sane project, but it sure would be nice if the sane api was more non-C friendly. I don't want to discourage you: A scheme-sane interface and interfaces to other languages are higly appreciated. I rememebr someone writing a C++ interface and there is a (extremely outdated) java interface. Bye, Henning
[sane-devel] Sane API Suggestion
A thin C interface layer could then use one callback dispatcher to handle all C callbacks. The callback dispatcher could use stdarg.h to process the other args or perhaps they could be passed to the callback using a count and array. Well, maybe the other developers have more experience with other languages. My first impression however, is that this makes the interface more complicated and complex. However, I'm no expert in interface design so if the other developers say it's better that way, I'll accept that. Scheme is a full-blown functional language, and thus very different from a pure procedural language. You can see from GObject (loosely part of GTK+) how difficult it is to approach object-oriented programming with C, and emulating functional programming in C would certainly be totally impractical. No C API will ever approach the full closure capabilities of Scheme, so some sacrifices on the Scheme side must always be made when interfacing C code. In the future, I would be happy to contribute a scheme-sane interface library to the sane project, but it sure would be nice if the sane api was more non-C friendly. Well, it's a C interface so beeing C-friendly is a feature :-) The SANE API is designed to be simple and portable, and any change towards object orientation or functional programming would only bloat it unnecessarily. Sorry, if you want to interface Scheme to SANE, you'll have to adapt to the way things are. Adding a callback data argument to both sane_init() and SANE_Authorization_Callback would cause a number of problems, in particular that of network transparency. Scheme or C++ programmers would certainly like to use that parameter as a pointer to a structure, but that only works if the size of the argument is at least that of void*. Since the latter varies across platforms, this would be ugly and non-portable. Since the only callback in SANE is the one responsible for authentication, I'd say it wouldn't be worth changing anything. A Scheme frontend would either have to relay the authentication request to the user, in which case the callback data is not used, or look up username and password in a database, in which case the resource string supplied is sufficient. Andras === Major Andras e-mail: and...@users.sourceforge.net www:http://andras.webhop.org/ ===
