- From: James P. Salsman <bovik@best.com>
- Date: Thu, 30 Mar 2000 23:42:43 -0800 (PST)
- To: Valdis.Kletnieks@vt.edu
- Cc: ietf@ietf.org, www-html@w3.org
Valdis, Thank you for your reply: > When was the last time you bought a microphone/audio card for > a system that didn't include at least basic software to do > [recording to files]? Not too many months. Try any Linux on any of IBM's PCs with one of their soundcard/modem combinations; you'll see. Sure, someone has a driver somewhere, but even top-of-the-line consumer Linuxes can't (or won't) auto-detect it, for IBM's most popular consumer models. But simply providing such applications is not the primary difficulty. How many clicks and keystrokes does the recorder app on your desktop take to save a file? How many to select that file in an INPUT TYPE=FILE widget? Doesn't that tell you why, with a growing market in speech recognition-based language learning software, nobody seems to be using web file upload for microphone data upload? > Well, the MIME spec came "out of the box" with audio MIME types. None of which were suitable for spoken language instruction until RFC 2586. The majority of them are still proprietary, and even the almost-state-of-the-art-and-wildly-popular MP3 format is owned by a (litigiously overwhelmed) German firm. >> in late 1996 some language instructors on one of the distance >> education lists (DEOS?) or newsgroups were claiming that >> voice-email presents more trouble than it is worth, at least >> for some students. > > There are those who find VCR's challenging. It isn't NTSC's or PAL's > fault... We are talking about serving a population of students. And a standards organization claiming to be dedicated to platform- independence and interoperability, while simultaneously claiming that non-normative aspects of the proprietary OBJECT and EMBED elements absolve it from complicity -- complicity not only in the promulgation of noninteroperable specifications that reinforce wintel dominance, but that expose ordinary browser users to the profoundly serious security risks of raw binary executable code. For that latter reason alone I believe there is justifiable cause for the IETF to suspend the W3C's HTML type registration. Plus, there are issues pertaining to the granularity of each recording. With a web-based asynchronous audio conferencing system using microphone upload, the task of grading a plethora of spoken phrases turned in from students could be made to be much easier than trying to take care of the same number of email attachments. Cheers, James
Received on Friday, 31 March 2000 02:44:07 UTC