Hebrew Pronunciation Update

NNic

Is there anybody there??

4 December 2013 and still no results. Come on Logos, why, given the time delay in producing this, do you not give updates??? Surely we deserve to be kept in the loop!!!!  Quando quando

Vincent Setterholm

Last month I received a large sample from the voice talent of approximately 25% of the total forms to be recorded, which gave a sizable sample to development to make sure the implementation was done on their end. I'm not ready to hazard a release date; given the rate of recording, I think we're still several months away.

I've been told there are two things you don't want to know how they're made: laws and sausages. I'm tempted to add Hebrew Pronunciation Add-ins to the list. But to give some historical background: when the project was first posted on Pre-Pub, the Hebrew Bible that was at the center of our infrastructure was the AFAT. It's what was behind all the Reverse Interlinears, the users of which are certainly a primary market for this tool. This created some serious design challenges, since AFAT didn't have vowels points in its lemmas, but relied on an increased number of homograph indicators to differentiate one word from another. This was before we'd merged lexical forms and homograph indicators into lemma data type references. So it was obvious that the code and tools used for the Greek recording, which were based entirely on how the word was spelled, weren't going to work. (E.g. without vowel points or homograph numbers, how are you going to tell when DBR should be 'davar' or 'dibber' or 'dover' or 'dvir' or 'dever'?)

Right about the time we got the new recording tool ready to help work around these issues (but no code written to ship the recorded results), we made the decision to develop our own in house Hebrew Bible (LHB) and replace AFAT in the reverse interlinears. Hindsight being 20/20, perhaps we should have pressed ahead with the original tool, but I was loath to go through the recording process twice when we had a new list of lemma data type references coming, especially since the new lemmas would solve a lot of problems when it came time to write the code for delivering the data, so the project was put on the backburner during the initial stages of LHB development.

Fast forward to the point where LHB was far enough along to begin recording, and the tools originally written for the project didn't work anymore (they needed to be re-coded to function with Logos 4 instead of Libronix DLS). It took a while to get the development time to solve that problem and then there were delays in contract negotiations with the fellow doing the recording, followed by another round of code updates to work with Logos 5 and then some period of time over the summer when we (the voice talent and I) were hashing out how to handle verbs in particular, where lexicons like HALOT and most databases still don't use vowel points, which ended up restructuring the project data a third time.

So now recording proper is well under way and I'm optimistic about getting this delivered, but not quite ready to name the day or the hour.