The Voynich Ninja

(17-10-2024, 04:23 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.Would it be a lot of work to make a version of this reader available with lower quality/size images? Now when I click a page, it needs to load for quite a some time, and it doesn't show the hits on the thumbnails like voynichese.com does.

That's not a lot of work, so added You are not allowed to view links. Register or Login to view. for low quality images.

(18-10-2024, 12:43 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.What I can make available, that would be input to an updated version of voynichese.com, are:

1) the new Beinecke scans all scaled down to a height of 1500 pixels, which is equivalent to what voynichese does with the old scans. Exception: the rosettes page is scaled down to 3000, because the actual page is also almost twice as high. Note that voynichese does not include this page. These images could also be used by @farmerjohn in case he considers Koen's request. Of course, it is not particularly difficult to create these. I do have all the scale factors from original to scaled down images, also not very difficult to create, of course

2) updated and completed transliteration data. This is the RF1 file expressed in Basic Eva. The source data is in the STA1 alphabet, so it would be possible (not too difficult) to translate this (ad hoc, online) to any existing alphabet. Even any user's own alphabet, as long as he/she presents a single translation table from STA1 to his/her alphabet. Many-to-one mappings are allowed.

3) the coordinate boxes for all words in the RF1 file. These are 'slanted' boxes, meaning that they are rotated along the direction of writing. They contain six values instead of the four of voynchese.com. They are not as accurate as the ones on voynichese, but serve the purpose of highlighting each word.

These last data are in csv files, so would require a simple script to convert to XML files.
The logic of voynichese.com would require an update to convert the box coordinates. This should be quite straightforward.
If the STA-to-any-alphabet translation is used, further additional logic would be needed for the user interaction, and the invocation of "bitrans" to do the translation.

I can provide more details if someone if feeling up to doing this (especially 'Job' of course).

EDIT: this is how the slanted text boxes are defined:

Yes, converting coordinates/transliterations from one format to another is a relatively easy task. Moreover, several transliterations can be run in parallel. That's a representational part.

But there is also a functional part.
When searching for a common word you normally expect for example "o" to match "ö" or "ō", or "t" to match "T". And in case of Voynichese I think it's normal to expect for example "o" to match "o with plume", so "qoky" should match You are not allowed to view links. Register or Login to view. or otherwise search results would depend too much on pedantry of a transcriber; if a transcriber decides one day that there are two kinds of "d" this shouldn't break simple search queries. 
Thus, some "symbol matching rules" are also required to provide an adequate search engine and these rules heavily depend on transliteration strategy used and probably are not universal.

(18-10-2024, 11:01 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.And in case of Voynichese I think it's normal to expect for example "o" to match "o with plume", so "qoky" should match You are not allowed to view links. Register or Login to view. or otherwise search results would depend too much on pedantry of a transcriber; if a transcriber decides one day that there are two kinds of "d" this shouldn't break simple search queries. 
Thus, some "symbol matching rules" are also required to provide an adequate search engine and these rules heavily depend on transliteration strategy used and probably are not universal.

Indeed, I do not know the exact matching rules of voynichese.com . However, I believe it is in everyone's interest to have clearly defined rules. There are easy ways to define whether the plumes (represented by a single quote) should be significant characters, or ignored. This can all be done with bitrans tables, but also in different, specific, ways.

If a transcriber (or researcher) decides that there are two kinds of 'd' (GC has at least three) then this is a change in the alphabet, and this can be communicated to the search engine by offering a new table (for example, but not necessarily from STA to the new alphabet).

By the way, if the transliteration at your web site includes characters that are not part of STA1, I'd be happy to extent STA in order to incorporate them. This can be at any later time of course.

(random example) I would argue to the hills these are "h" + flourish/plume whatever we want to call them, rather than R or S (I'm not educated in linguistics if you hadn't guessed..) 



Would a crowd function such as submitting suggestions not be beneficial in these sorts of cases? Just touching on Farmer Johns points about functionality and transcriber input. I haven't looked how these are dealt with in Rene's transcription, but just the querying the general idea. Is it not something that could work, even as a alternative option to using the original transcription?

More updates:

Auxiliary You are not allowed to view links. Register or Login to view. to view transliteration of a folio in human-friendly form.

You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. for alternative transliteration (just a couple of random pages).

(20-10-2024, 12:58 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.More updates:...

When I click on links, the browser tells me that my connection is not secure, is this normal?

(20-10-2024, 04:47 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.When I click on links, the browser tells me that my connection is not secure, is this normal?

Maybe because the SSL certificate expired Sunday, September 29, 2024 at 1:59:59 AM. I added a security exception long ago so I don't know if the connection was secure before this date.

(20-10-2024, 04:47 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.

(20-10-2024, 12:58 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.More updates:...

When I click on links, the browser tells me that my connection is not secure, is this normal?

Yes, the site has no certificate

(20-10-2024, 10:26 PM)farmerjohn Wrote: You are not allowed to view links. Register or Login to view.Yes, the site has no certificate

It's a shame !
I'm completely ignorant in this area.