Fix: [Network] Reading beyond the length of the server's ID when hashing password

Under normal circumstances the server's ID is 32 characters excluding '\0', however this can be changed at the server. This ID is sent to the server for company name hashing. The client reads it into a statically allocated buffer of 33 bytes, but fills only the bytes it received from the server. However, the hash assumes all 33 bytes are set, thus potentially reading uninitialized data, or a part of the server ID of a previous game in the hashing routine.
It is still reading from memory assigned to the server ID, so nothing bad happens, except that company passwords might not work correctly.

Fix: Crash when extra viewport height is zero with sign in view. (#9175)

If a viewport sign straddles the top of a viewport, a crash will occur if the viewport height is zero. This is resolved by simply not attempting to draw the viewport in this situation, consistent with other widgets.

Codechange: Warn if randomaction2 group count is not a power of 2.

Previously noted by a comment, this does not need to be guarded against as non-powers of 2 will not cause issues beyond the choice of results being reduced.

Fix #6598: Prevent invalid memory accesses when abandoning a join from within a network game

One could join a network game from within an already running network game. This would call a NetworkDisconnect, but keeps the UI alive. If, during that process the join is aborted, e.g. by cancelling on a password dialog, you would still be in your network game but also get shown the server list.
Solve all the underlying problems by falling back to the main UI when (re)connecting to a(nother) server.

Fix #6598: Do not disconnect before company number validation

NetworkClientConnectGame already does a NetworkDisconnect, so no reason to do it here

Codechange: Scale sprite font height once on init instead of every call to GetHeight().

Scaling is not expensive, but it does not change either, and this avoids the need for a virtual method call. This cascades back to all GetCharacterHeight(FS_xxx) and FONT_HEIGHT_xxx calls.

Cleanup: Use GetDefaultFontHeight() call instead of direct access.

This makes this part of font size setup in FreeTypeFontCache consist with OSX and Windows variants.

Fix: Scale cargo lines in industry chain window.

Replaces constant pixel values with values scaled based on font size.
This allows the industry chain to maintain a consistent look across
different sizes. Previously all except cargo line height were fixed.

Fix: Improved scaling and spacing of sign list window.

Both company icon sprite and text now centred within each row, and extra
padding added to avoid the sprites running into each other.

Codechange: Use C++ features for train wagon overrides. (#9141)

This removes the need for C-style array management and allows use of iterators to perform wagon override lookups.

Remove: performance measurements in YAPF

YAPF was constantly measuring its performance, but only at
certain debug-levels this information was shown.

Now after years, I sincerely wonder if anyone still knows about this
feature and who still use it. Especially with the new framerate window,
this detailed performance is not as meaningful anymore as it once
was.

Change: use TCP for everything except for master-server and initial server scan (#9130)

This means that pressing Refresh button and adding servers manually
now uses TCP.

The master-server and initial scan are still UDP as they will be
replaced by Game Coordinator; no need to change this now.

If we query a server that is too old, show a proper warning to the
user informing him the server is too old.

Fix: String validation could leave invalid Utf8 encoded strings (#9096)

In case a character was encoded in multiple bytes, but required fewer bytes to be encoded, the first byte would be copied to the output leaving an invalid Utf8 encoded string. Later uses of the validated string would use the same decode logic, which would yield a question mark and just read a single byte, so nothing dangerous happened.
Furthermore, because the next byte would not be a first byte of an encoded Utf8 character, the last few valid characters could be removed by the validation as well.

Codechange: use NetworkAddress instead of two host/port variables where possible

This also means we no longer need last_host/last_port, but can
just use a single last_joined setting.

Fix: 'Cache' top and bottom lines of textfile viewer to avoid overdraw. (#9131)

* Fix: 'Cache' top and bottom lines of textfile viewer to avoid overdraw.

The text file viewer calculated the number of lines required to set the scrollbar, but did not retain this information, so this was recalculated on every draw operation. This includes overdrawing text outside the bounds of the current scroll position.

With this change the top and bottom lines for each line of text are remembered, and reflowing is avoided where possible. Text outside the current scroll bounds is not drawn.
Additionally the scroll interval is now based on text lines instead of pixel lines, which increases the text capacity depending on the font size.

* Fix: Limit text viewer to showing 64k lines.

Text files with more than 64k wrapped lines would exceed the scrollbar capacity and cause an assert. This is harder to reach now that the scrollbar counts lines instead of pixels.

Update: Translations from eints
norwegian (bokmal): 16 changes by Anolitt
finnish: 2 changes by hpiirai

Fix: Incorrect vertical alignment of icon and text in DropDownListIconItem. (#9133)

This happens if the bounding dimensions are changed so that each item is the same size, as happens on the railtype/roadtype dropdown lists, as the vertical offset was calculated before this dimension is changed.

Fix b3003dd1: swap SERVER_GAME_INFO with CLIENT_GAME_INFO (#9129)

The idea is that if you query an older server that does not support
this packet yet, the client receives an error. The assumption was
that on every "illegal packet" the connection would be closed. This
turns out to be false.

Now CLIENT_GAME_INFO aligns with the old PACKET_CLIENT_NEWGRFS_CHECKED,
which does a pre-check (which fails), and an error is sent back
and the connection is closed.

This is not a nice solution, but it is the best we got.

Update: Translations from eints
spanish (mexican): 40 changes by absay
english (us): 1 change by 2TallTyler
korean: 3 changes by telk5093
german: 1 change by danidoedel
finnish: 1 change by hpiirai
catalan: 1 change by J0anJosep
portuguese: 45 changes by azulcosta
portuguese (brazilian): 44 changes by Vimerum

Change: no longer use UDP when entering the lobby of a server

The lobby of a server requested some parts via UDP and some via
TCP. This is strictly seen fine, but for future extensions it
is a lot easier if just one protocol is used.

Codechange: refactor CheckGameCompatibility() from existing function

Later commits use this function in other places too.

Codechange: move all NetworkGameInfo related functions to a single file

It currently was a bit scattered over the place. Part of
NetworkGameInfo is also the GRF Identifiers that goes with it.

Change: [Network] lower TCP connect() timeout to 3s (#9112)

Currently we use default OS timeout for TCP connections, which
is around 30s. 99% of the users will never notice this, but there
are a few cases where this is an issue:

- If you have a broken IPv6 connection, using Content Service is
first tried over IPv6. Only after 30s it times out and tries
IPv4. Nobody is waiting for that 30s.
- Upcoming STUN support has several methods of establishing a
connection between client and server. This requires feedback
from connect() to know if any method worked (they have to be
tried one by one). With 30s, this would take a very long time.

What is good to mention, is that there is no good value here. Any
value will have edge-cases where the experience is suboptimal. But
with 3s we support most of the stable connections, and if it fails,
the user can just retry. On the other side of the spectrum, with 30s,
it means the user has no possibility to use the service. So worst case
we annoy a few users with them having the retry vs annoying a few
users which have no means of resolving the situation.

Update: Translations from eints
japanese: 26 changes by scabtert
catalan: 43 changes by J0anJosep

Change: [Network] Safeguard from using errno/strerror for handling network errors

They are likely not working as expected on Windows, so prevent their usage.
Winsock does not set errno and strerror does not return anything useful for Winsock error numbers.

Update: Translations from eints
english (us): 39 changes by 2TallTyler
korean: 44 changes by telk5093
german: 43 changes by danidoedel
russian: 7 changes by Ln-Wolf
finnish: 39 changes by hpiirai

Fix: Count engine details text in lines rather than pixels. (#9107)

This allows the details panel to scale correctly for different zoom levels.

Update: Translations from eints
norwegian (bokmal): 4 changes by Anolitt
spanish (mexican): 3 changes by absay
japanese: 60 changes by scabtert, 38 changes by Azusa257
english (us): 3 changes by 2TallTyler
korean: 3 changes by telk5093
russian: 4 changes by Ln-Wolf
finnish: 3 changes by hpiirai
slovak: 20 changes by FuryPapaya