Preview

[quote]There are obvious problems with recognizing the text of this document (creating text version). I was just wondering if it is because of underlining/formatting, or itis also language-specific? Does anyone have similar problems with the documents in another languages?[/quote]

Yes,and it is strange that the tags are shown in the forum preview (like [/quote] in this particular example - when you click on "forums" button and see recently updated threads). Maybe, that'd be better to show the preview in final page layout format - if you know what I mean?

[quote]recognizing texts is especially difficult if the original document is a clipping, e.g. in case of the following: [document=484]484[/document] [/quote]
Oh yes, recognizing texts is difficult in any situation when it is not just a perfectly scanned clearly printed document without consuderable "impurities"; however, when the text is clear in the source, I'm deeply impressed with the quality of the recognized outcome.
I think, it would make sense to:
a) make the button "create text version" more "obvious" and visible;
b) drop a warning note a sentence or 2 sentences long about the possible problems with text recognition.

BTW, when I see the list of documents, I can't see any indication whether it has "edited" textual version or not.
Maybe, there is a sign, but I just can't see it? Or we need to indicate it in descriptions?

I have also problems with creating text version (in case of Hungarian texts), however, the document was scanned in TIFF format at 300 dpi, black/white. To my mind, recognizing texts is especially difficult if the original document is a clipping, e.g. in case of the following: [document=484]484[/document]

(I was to try out the "Upload" option, but I was not really meant to upload any pictures or documents. However, I cannot change this command unless I reload the whole page, or first submit the comment, and then modify it. Why is it so?)

As I can see, this is more about the file format and quality than the language...though I have not been the one testing this feature, I think my colleagues have had success in various languages and alphabets (specifically Cyrillic). We are still working on this...so we are happy for you to report your results...the file formats and languages you were dealing with!

There are obvious problems with recognizing the text of this document (creating text version). I was just wondering if it is because of underlining/formatting, or itis also language-specific? Does anyone have similar problems with the documents in another languages?