Let's organise a bounty to buy a new CR2032...

@trixie

Quote:

Me too. I get all sorts of tingly feelings when I think of the two development companies getting together.

I don't think that's right.

As I understand it, the ".recycled" drawer is a fixed size (on all of my partitions it's about 1% of the disk). It will never get bigger than that and the filesystem can not gain more disk space by deleting stuff from the ".recycled" drawer.

What you are describing is a "Trashcan", which is a conventional directory that expands and contracts as you add or delete files from it.

I tried but it won't accept my Visa card, it says it's "not suitable".

I then tried to use PayPal, which is linked to the same card, but PayPal then asked me for another bank account to transfer funds from. When I logged out of that, OWB crashed.

Sorry, too much hassle.

Maybe your calendar has fewer months than mine, but I don't see how 12 months is up yet.

How many "major new versions" of Firefox have been released since September last year?

The algorithms are merely the interesting bit in the middle. Most of the code relies on the interfaces to the outside world - how you get input from the user running AmigaOS, how you talk back to him/her.

Then there are the problems associated with trying to read stuff from disk, how to write it back. That's the stuff that is not written in any textbook, because it's unique the operating system you are using.

Once you get to that stage, you have all the Amiga documentation on the wiki. Some of it is still old, but it's gradually being replaced with up-to-date versions.

@K-L:

Thanks, that SetEnv works once I edited it suitably. As you quoted it (and used in ClipDown), the command SetEnv "---" ends up saving an ENV-variable as

>echo $OWB_USER_AGENT
Mozilla/5.0 (iPad

- which obviously doesn't work. I changed it to:
"SetEnv OWB_USER_AGENT Mozilla/2.0" and it works just fine.

Thanks for all the help, now I can read the news again.

Sadly, neither of those suggestions is usable. It does not support any tooltype that suggests spoofing, and the other utility is solely for editing saved bookmarks and hotlists.

I guess I'll just have to use MUI-OWB for the ABC and RA-OWB for everything else.

Our local independent news site (www.abc.net.au/news) has today started generating a "mobile" display by default. I now get a stripped-down "mobile" version on RA-OWB.

MUI-OWB has the ability to select "Spoof As..." and by selecting "FireFox 2.0 for Windows XP", I can get my display back, but is there any way I can spoof with RA-OWB?

It's a bit hard to take seriously. None of his "team" has any experience with AmigaOS (or else he would name them and we would recognise the names). If his team is a single person (him) then he's dreaming. If it's a number of people, then how are they going to be trained in the AmigaOS and how are they going to earn funds to keep them going?

Look at the problems that a newcomer like AlfKil has had trying to learn the AmigaOS and its quirks. Is "The Team" going to jump straight into coding at the kernel level (device drivers) and the highest application level (OOP) ?

It would be great if a new, enthusiastic team were to suddenly appear and join the OS4 dev team, but with no income, there's going to be no outcome.

I'm sure that Thore will answer your question.

In the meantime, though, are you dealing with someone else's application using MUI, or a program that you have written yourself, using MUI? The answer really is that the application has to be written so as to prevent the crash.

@Karlos:

Karl, I was not Removing the node in the snippet of code that I proposed. I do know better than that!

[edit]
Sorry, I didn't realise that "L_FreeFiletype()" might free the Node in the argument.
If it does, then the "Next" node must be saved before the call, as you so rightly point out.
[/edit]

You should not even be using the names "lh_Head", "ln_Succ" and others within the structures. The Exec provides foolproof calls to find the first/next/previous/last node in a List, and returns NULL safely if the item asked for does not exist.

What is going to happen when we change the contents of a list node? Your port is going to fail. Use the provided Exec calls like GetHead(), GetSucc(), etc, and remain future-proof.

The loop should read:

for (node = IExec->GetHead(filetype_list;
node != NULL;
node = IExec->GetSucc(node))
{
L_FreeFiletype((Cfg_Filetype *)node);
}

The filesystem does not need to know the owner of a Lock, only DOS does. As Chris says, I don't think it's possible without "inside" information.

It might be possible (depending on modern MMUs and what they can do - I'm no longer an expert in the field) to set up the MMU to handle the passing of the message from one address space to another, translating address pointers within the message as it does. If that task could be done in hardware, the additional latency could be reduced to a single ALU cycle.

The main problem (as you well know) is that some addresses have to be changed (those relative to the tasks concerned) and some don't (those relative to the system). Somehow the MMU has to be shown which addresses to change and which to treat as immutable data.

We used to do it back in the 80's with DEC hardware, so it should still be possible with hardware that is thirty years more advanced. However, a lot of features of the old CPUs have been dropped in modern microprocessors to save money, (eg multiple-level interrupts), so it may not be possible these days.

If you set the proper tooltypes in PlayCD, it will fetch the CDDB file for you.

I have a directory reserved for CDDB files and I use these PlayCD tooltypes:


DOSDEV=CD0
(DEVICE=scsi.device)
(UNIT=2)
(SKIP=???)
(PEEKTIME=333333)
(PUBSCREEN=<public screen name>)
(BUFMEMTYPE=1)
USECDDB
URL=http://freedb.org:80/~cddb/cddb.cgi
(PROXY=)
USER=xxxx@yyyy.com
CACHEDIR=Caches:CDDB
LISTSIZE=10
USEAHI

"USER" has to be an email address. It doesn't have to be real, any address will do. I use an old address from the dial-up days, but it still works, even though the company folded 15 years ago.

Hey, Ace!

How's it going?

Inline protos are OK, as long as they are defined in a way that prevents clashes between modules.

If you say "INLINE <type> MyFunc()" in a global header that all modules can read, then each module will export the name MyFunc() and you'll get multiple definitions. You have to change the inline definition to:

static INLINE <type> MyFunc(...)

Put that in your global proto header and no module will export the address to conflict with others.

Another frequent use of the tags is when you want to use a structure that starts with a MsgPort, but all the rest of it is private stuff.
eg:
struct PostOffice
{
struct MsgPort;
APTR ---
---
---
};

You can allocate such a structure as a MsgPort of a larger size, using the tag
"ASOPORT_Size, sizeof (struct PostOffice)". You will then get a clean, initialised MsgPort and all the space after it that you need. You can simply FreeSysObject (ASOT_PORT, xxx) when you're finished with it.

There is a Kickstart component called "diskboot.kmod", maybe that calls itself "diskboot.config" when running, or spawns a task that calls itself that.

If I weren't so busy (Read: lazy), I could look up the sources and find out. But look at it this way: when your program crashes in Exec, you don't blame Exec for the crash, you blame your program, don't you?

In this case, Quake is calling a newlib function that ends up in a crash. It's most likely that the crash is quake's fault. It is probably caused by some old coding trick that used to work back in the old days, but has been made illegal these days.