From: Takashi Sakamoto Subject: Re: [RFC] libhinawa: a light-weight I/O library for status/transactions to ALSA FireWire devices Date: Tue, 30 Sep 2014 11:21:24 +0900 Message-ID: <542A13A4.8080607@sakamocchi.jp> References: <54215BF6.5090600@sakamocchi.jp> <20140930014919.GE5605@marvin.atrad.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: Received: from smtp303.phy.lolipop.jp (smtp303.phy.lolipop.jp [210.157.22.87]) by alsa0.perex.cz (Postfix) with ESMTP id 7EBD52606AB for ; Tue, 30 Sep 2014 04:21:30 +0200 (CEST) In-Reply-To: <20140930014919.GE5605@marvin.atrad.com.au> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: alsa-devel-bounces@alsa-project.org Sender: alsa-devel-bounces@alsa-project.org To: Jonathan Woithe Cc: "alsa-devel@alsa-project.org" , "ffado-devel@lists.sf.net" List-Id: alsa-devel@alsa-project.org Hi Jonathan, On Sep 30 2014 10:49, Jonathan Woithe wrote: > The situation to avoid is the one we had with libraw1394; when the new stack > came about the preferred way to interact with it switched back to the native > kernel interface since (as I understand the situation) it was felt that > libraw1394 didn't really provide a worthwhile API abstraction. I'm sorry but I don't understand what you assume. What is 'the new stack'? > Out of interest, is there any signficance in the library's name "hinawa"? I require unique name for this shared library. Then, for this RFC, I select 'hinawa' as surely-unique name, from Japanese word according to my association from 'FireWire'. ('hi' = 'fire', 'nawa' = 'wire'). The 'hinawa' is just my convinience. I don't mind to use another word for its name. >> Dice notification is an actual example. Dice based devices transfer >> notification to a specific address on host controller. The address space >> is exclusive resources on an system. For Dice notification, ALSA Dice >> driver gives a way to utilize it for applications As well as Fireworks >> situation, some userspace stuffs are needed to use this functionality. > > For sure. One thing to keep in mind is that while this is an issue for DICE > devices it may not apply to others. The library API should be structured to > keep in mind that not all features will be needed across the board. We > don't want to burden all subsequent firewire-audio streaming drivers with > baggage which is only required for a small subset. My intention of this library is to write applications for device control with LL (however currently it's C only...). I have no plan to add other functionalities except for helping transactions and notification because I want to reduce my maintaining effort. In short, I have no will to make alternatives of libraw1394, libiec61883, libavc1394 and libffado. Regards Takashi Sakamoto o-takashi@sakamocchi.jp