It would be really helpful if it supported Office 2010/2013 clipboard information. So that it would retain bold/italics/font size/etc.

One step further would be to also detect and store font styles that would be usable throughout the database.

For Example:
I have a very large microsoft word document that is using font styles for every component of my norwegian notes.

The reason why I used this system is so that when I want to make a change to several thousand vocabulary words formatting I can do it by just adjusting the style profile.

There are a lot of card-sets offered here of them a lot with sound:
http://www.byki.com/

To import the files from there would be great!

Suppose an English learner was learning the word "toothbrush". It would be facilitate learning if this vocabulary item could be left pending until both "tooth" and "brush" had already been learned.

When you review a difficult card, with this skip the card option, you skip it temporarily, not let it block the way to review other cards, just as you can skip some tricky questions in an exam and go back later.
In my experience, thinking of the answer of a difficult card is time-consuming. But if we can skip it and go back later, the answer usually spring up, due to our magical brains in recalling.
The function has been implemented on Mnemododo. Why not implement it on Mnemosyne 2.x?

Latex and HTML code should have a different color than normal text. That way it would be much more readable.

I just found out about a feature, which I wanted to suggest. ^^ Ctrl+Tab indents text in the edit window.
Indenting multiple lines of text, could be nice feature. For example: select some lines and then use Ctrl+Tab to indent all selected lines. Well, this feature, will probably not be added before mnemosyne 20.1, but maybe the devellopers will get bored one day. ;)

My language cards use Front-to-Back, Front-to-Back-&-Back-to-Front, Cloze Deletion types (I don't use Vocabulary type). It is helpful to use a different background color for each language so they are readily distinguishable (e.g. light red for Indonesian, light blue for English, etc.) But that requires a clone of each of the three above card types for each language. It would be much easier if I could set a background color which would apply, for example, to any Indonesian card, regardless of the card type.

If "fast format" is used, searching for cards in the browser (in the field "Containing this text in the cards") requires that the format characters be included in the search string. But often a user cannot remember if formatting has been used, or which formatting options and in what order. This means that some text matches may be missed in the search if formatting is not entered exactly. (I don't know if this applies to HTML formatting as well.) A search should look for content, not formatting. Therefore it would be better if the search ignores formatting strings.

It would be very good if a person can decide how he want to learn a card, I mean a person can decide how many days for a grad take till program ask the card again.

Most of my time is usually spent on doing html formatting stuff. I've even wrote a simple app in C++ to simplify my job. But its still too time consuming. it'll be great if you could paste an already formatted text from an external app (which are in the format of rich text or RTF) and mnemosyne converted it to html.

it would be so good if a person can more customized his own interface

I would bet some money that (when you control for when the user switches away from the application) you could get some signal out of the delay before a user answers a question. I would even bet that it would be better than the users vote regarding how easy/hard it was.

It would be awesome to hook this up to an EEG monitor, although it is an expensive task. I might need to purchase a researcher grade Emotiv Epoch anyway, so if you want me to submit the raw data...

For example, if I press printscreen, or if I use a screen snipper tool, I'd like to be able to paste those images into Mnemosyne like I can with Anki. The current method of saving the image and inserting it from a dialog is far too slow.

An Apply button for the edit window would be useful when editing multiple cards from the card browser. PageUp / PageDown could then be used to move to the previous / next card instead of closing the window with the OK button, then moving to the next card and then opening the next card.

Would be nice to be able to reduce the grade your answer selections. I use this for my 6 year old and 6 choices are too many. 0=mised it, 1= kinda got it after thinking for a while, 2= know it.
Being able to reduce the number of grades (which should be configurable) would be great!

It would be lovely if TTS could be included though I suspect this would be v. difficult due to the availability of decent free TTS engines and voices. Perhaps someone out there knows more than I.

Add support for specifying card type in/from the file format when exporting/importing/into/out from mnemosyne.

For example map, front-to-back, vocabulary, etc.

The user will be able to specify which field in the card type each column in the txt file corresponds to.

Some columns could also be used to specify tags.