@AmigaOldskooler

Yes, they both worked at the point I uploaded them

However, my weatherstation was generally connected to a Raspberry Pi running PyWWS, and hasn't actually been connected up since I moved.

I hit the "3" on the front page and it took me to the most recent unread post. I was expecting it to take me direct to page 3 (although actually most recent unread is more useful)

@trgswe

There's code in the SDK for building keymaps. I think the format on OS4 differs from OS3, but it's a long time since I've needed to create an OS4 keymap.

On OS3 I don't think there's any source code available, so you have to use a keymap editor!

@rjd324

Back-tick should be the top-left key under Esc - that's what's printed on the key. When I tested, it's Alt and that key for the back-tick.

I don't know what Commodore were thinking - I believe the British keyboard on all Amiga models have the back-tick on the top-left, but the keymap itself has the apostrophe there!

On the A600 and newer there are two blank keys they could have used to separate them out, and then corrected the keycaps.

@NinjaCyborg

If AmigaOS used XML across the board and we had an xmlparse.library which was simple to use and included with the OS - properly encoding and decoding strings and setting types etc - then I'd be less against it.

But instead we have a mishmash of open source libraries, none of which are particularly easy to use, and developers who aren't familiar with them or XML (I include myself in that*) but want to use XML as apparently it was cool around the early 2000s, even though it isn't really suitable for the intended task.

* I've never been much keen on XML, and every time I've had to try to deal with it it has been difficult to work with.

@Raziel

As an official component a Prefs editor is pretty essential. But it should probably have written the config in IFF PREF format like most of the other Prefs editors (excluding some of the newer ones which also use XML as NinjaCyborg noted).

The bonus is IFF parsing through iffparse.library is quite similar to how XML parsers work, but simpler, and probably both less memory intensive (the parser alone is smaller) and quicker.

@NinjaCyborg

XML is horrendous for parsing, it doesn't surprise me it causes problems. Why people insist on using it for config files instead of something more appropriate is beyond me.

Hint: if you don't have a complex application which necessitates nesting configuration options (and even if you do there are other solutions), you shouldn't be wasting your time using XML. There are plenty of other options, such as IFF, ToolTypes, INI, JSON, YAML. The first two in particular are preferred and easy to parse for Amiga applications.

@LyleHaze

I don't really understand why that solved your problem, but...

Quote:

Yes, you can connect the wired interface of a Pi to the Amiga and tell it to route traffic through the WLAN. I did this for years - it works really well, and much faster than the ancient 802.11b cards that we have drivers for.

@Raziel

Mount it as a CIFS share in /etc/fstab.

You won't get a drive letter - Linux has no such concept - but you will get a known mount point.

This looks like quite a good guide:
https://linuxize.com/post/how-to-mount-cifs-windows-share-on-linux/

@ChrisH

Nothing significant that I remember.
You can trawl through the git logs if you like, it hasn't been massively active: http://git.netsurf-browser.org/netsurf.git/log/

@328gts

I don't think it will; guigfx seems to have some endianness issues (probably my fault). Having said that, I think I re-enabled using P96 under OS3, and it wasn't working with that either, but I can't remember if the colours were wrong or the graphics were completely corrupt.

@utri007

Probably, and I don't even know what caused the speed issues.

@alfkil

If you want monotonic time you have to use the EClock. Something like: (this is in ms)

@alfkil

Are either of them on wi-fi? I've had weird effects with all sorts of different computers with communication randomly not working on wi-fi, although I think pings are failing then too.

@328gts

A cross-platform GUI toolkit.

As it's you, yes!

I didn't even realise Qt had a v6!

Open users/<your username>/Choices and add the line:
log_filter:level:CRITICAL

@Raziel

If it works as nbache says, then the sync might be blocking the drive and then the reboot happens, but due to timing the inhibit may not be completely done. If Ringhio has added a reset-handler which saves prefs on reboot then that could cause it to half write the file before the drive is protected. Result - corrupt file.

I don't see the need for sync, take it out.

@Raziel

I'm not sure why you need a sync - just wait for disk access to finish and it should be fine. Presumably the sync is what is screwing it up.

For example use a different reboot command and something like this (don't know if it works on OS4 but the source is there): http://aminet.net/package/util/boot/ResetHandler

@Raziel

Is something installing a reset-handler which maybe writes to the file but doesn't finish writing quick enough (or it ends up in memory cache rather than on disk and reboots before sync)

Not sure if it's possible to find out which reset-handlers are installed.