Janek,
wifi passwords, reading history and maybe even more
is not affected by firmware update.
I do not know if website bookmarks, web cache and history, cookies, website passwords, music player settings and bookmarks are affected. I do not use it that much.
It is true that, also
time, date, timezone, user language (I am not sure if it is preserved or not) and
MAC address should be preserved.
To my opinion fonts and dictionaries do not have to be preserved as user may place them in user (non system) partition or even SD card (I think...), that is easily accessible from computer. No need to alter the firmware instalation file or to install OTFM file manager (for M92) to copy files to system partition.
It would be nice, if also sleep-mode screens could be placed in "user space", like dictionaries or fonts, so that one does not have to copy the files to system part. In my case it is 50+ files, so it takes time....
Janek, I agreee that it would be fine to have "bug free" firmware and than fine-tune it and add new features. But I can imagine that to replace old pdf reader by new one may be the best and fastest way to achieve this goal... We do not see into internal workings of the old one.
The firmware could be better and less buggy and there could be a better work flow... But for my style of using it, it does most of the things that I need. I do not have the hardware problems and I had less hangups (maybe 2-5 in 7 months of daily use) than in Windows. It is to my opinion acceptable and very near to stable. But maybe that parts of firmware I do not use are more buggy. I do not know... I use M92 mostly for reading pdf documents and scribling in it.
I must say, I expected the firmware development faster than it is in reality. Some things are fixed or improved very fast, some that seem to be also easy take long time. I do not know, if the reason is, that the programmers do not get the info or it is more difficult than I expect or it is not high priority for Onyx or ...
Anyway, I would recomed M92 to my friends even thought it is not perfect.
I do not know any better solution on the market... Maybe Pocket Book...
---
I didn't have much time to test a new firmware, namely beta release of a new pdf reader.
I like 2 collumns mode. It works as expected. There is a 'problem' if the 1 and 2 collumn texts are mixed on one page. That is standard on the first page (headline and annotation) and sometimes last few pages. Also tables and graphs/pictures may be one-column while text is two-column. Hard to say, how this should be tackled. Probably to show parts that are one column as one column (and show it twice - every time the reader passes that section of left or right collumn...).
I do not know if those 'peculiar' parts can be detected. Generally everything is possible, but the question is if the pdf engine interface allows to capture this info in +- easy way, or at least the hard way. Even with scanned documents it could be handled. Just the detection, if there is the white central margin, is needed...
Sorry for long post...