provide download links of both versions of OCR add-ons, one using the latest Tessarac engine, and the other using the older Tessarac engine, du o significant difference in file size/footprint

[nvaccessAuto]

Reported by bhavyashah on 2014-12-17 14:51
Quite some time ago, on the NVDA Add-ons list I had asked the authors of the OCR add-on to update the version of the Tessarac engine being used by the add-on, in order to get support of the languages included in the newer Tessarac engine release. In reply, Jamie had said that the newer Tessarac engine would significantly increase the footprint of the add-on, to about a hundred megabytes (my figures may be wrong, apologies if they are).
There really are quite a few languages added in the latest release, and I really need one of them which is Hindi.
SO, I would like to request you to provide download links of both versions of OCR add-on, one with latest and one with the older Tessarac engine.

[nvaccessAuto]

Comment 1 by jteh on 2014-12-17 23:08
It's not the OCR engine which has grown. It's the fact that the older version chose specific languages to include, but I can find no justification for why those particular languages were chosen (and Rui can't remember how he chose them either). I can't justify the inclusion of some languages and not others, so we either include all or just English. All is way too large (can't remember exact numbers, but the download was several hundred megabytes and it extracted to maybe a gigabyte or so), but English means everyone else has to download languages separately. It's that last part which is a problem: downloading and installing additional languages is non-trivial for non-technical users.

Having two downloads is not going to solve this problem.
Changes:
Added labels: wontfix
State: closed