I tried something more:
As I said, in the windows client all folders and their content is shown and accessible using the seafile-browser. So, I tried creating a download link:
a) A download link works, when it was created for a folder that is shown in the web interface
b) A download link does not work, when it was created for a folder that is not shown in the web interface
So, my guess was that there might be some data corruption and I run seaf-fsck. It did not find any error.
Any more idea what I can do (or try) in order to be able to access all folders also from the web interface?
I have to correct one point: the download-link created in the seafile desktop client works now, even if dealing with a folder that is not shown in the library in the web interface.
[11/24/17 23:14:33] ../common/peer.c(950): Local peer down
[11/24/17 23:24:32] ../common/session.c(409): Accepted a local client
[11/24/17 23:24:33] ../common/peer.c(950): Local peer down
[11/24/17 23:34:33] ../common/session.c(409): Accepted a local client
[11/24/17 23:34:33] ../common/peer.c(950): Local peer down
in EDGE: The javascript console does not show any error or warning.
What might be helpful:
If I scroll down very, very slowly EDGE loads once more folders. FireFox doesnot.
FireFox gives a warning in the javascript console: Diese Website verwendet anscheinend einen scroll-verknüpften Positionierungseffekt. Dies könnte mit asynchronem Verschieben (Panning) schlecht zusammenspielen; siehe https://developer.mozilla.org/docs/Mozilla/Performance/ScrollLinkedEffects für weitere Details und nehmen Sie an der Diskussion über damit verbundene Werkzeuge und Funktionen teil!
Which translates to something like: This website seems to use a posititioning effect based von scrolling. This could cause problems with asynchronically moving (Panning); see https://developer.mozilla.org/docs/Mozilla/Performance/ScrollLinkedEffects for more details and discuss about linked tools and functions!
Chrome’s console might also be helptful:
main.ce0533b25c52.js:212 [Violation] Added non-passive event listener to a scroll-blocking 'touchstart' event. Consider marking event handler as 'passive' to make the page more responsive. See https://www.chromestatus.com/feature/5745543795965952
add @ main.ce0533b25c52.js:212
(anonymous) @ main.ce0533b25c52.js:212
each @ main.ce0533b25c52.js:212
each @ main.ce0533b25c52.js:212
on @ main.ce0533b25c52.js:212
delegateEvents @ main.ce0533b25c52.js:212
t.View @ main.ce0533b25c52.js:212
i @ main.ce0533b25c52.js:212
initialize @ main.ce0533b25c52.js:212
t.Router @ main.ce0533b25c52.js:212
i @ main.ce0533b25c52.js:212
(anonymous) @ main.ce0533b25c52.js:212
execCb @ require.640929dac3c2.js:1658
check @ require.640929dac3c2.js:874
enable @ require.640929dac3c2.js:1151
enable @ require.640929dac3c2.js:1519
(anonymous) @ require.640929dac3c2.js:1136
(anonymous) @ require.640929dac3c2.js:132
each @ require.640929dac3c2.js:57
enable @ require.640929dac3c2.js:1098
init @ require.640929dac3c2.js:782
(anonymous) @ require.640929dac3c2.js:1424
require.640929dac3c2.js:1413 [Violation] 'setTimeout' handler took 253ms
main.ce0533b25c52.js:212 [Violation] 'readystatechange' handler took 593ms
I think this is something you should look into @daniel.pan . It has been reported multiple times and I think with this information in mind a fix should be possible.