home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.sys.acorn:10562 comp.sys.acorn.tech:1423
- Newsgroups: comp.sys.acorn,comp.sys.acorn.tech
- Path: sparky!uunet!spool.mu.edu!agate!doc.ic.ac.uk!warwick!pavo.csi.cam.ac.uk!idg10
- From: idg10@cl.cam.ac.uk (I.D. Griffiths)
- Subject: Re: Monty Pythons most used sentence
- Message-ID: <1993Jan26.125507.5538@infodev.cam.ac.uk>
- Sender: news@infodev.cam.ac.uk (USENET news)
- Nntp-Posting-Host: stonea.cl.cam.ac.uk
- Reply-To: idg10@phx.cam.ac.uk
- Organization: U of Cambridge Computer Lab, UK
- References: <1k11shINNgps@info2.rus.uni-stuttgart.de>
- Date: Tue, 26 Jan 1993 12:55:07 GMT
- Lines: 94
-
- In article <1k11shINNgps@info2.rus.uni-stuttgart.de>, zrzs0111@helpdesk.rus.uni-stuttgart.de (Helmut Springer) writes:
- |> "And now to something completely different"
- |>
- |> Has someone build a module with a "save, easy useable, not crunshing"
- |> method to request memory (for RISCOS 2.00)?
- |> If you request memory with OS_Module, the memory can't given back, if a
- |> other module gets loaded.
- Yes and no. If you give it back, and something else asks for some space in
- the RMA, then it will get the hole you left rather than allocating a big
- chunk, so it's not completely stupid. Just fairly stupid.
-
-
- |> If you request memory with WIMP_SlotSize, you may run into programs which
- |> sits already in the memory (cause of BBC 6502 compatibility ?) and twice
- |> this method has no garbage collection ...
-
- Yerwhat? There is no way that memory allocation via Wimp_SlotSize will
- allocate you memory already owned by another task (unless it has illegally
- stolen a page of memory by going direct to the MEMC) since it only gives
- you memory that was previously sitting in either the Free or the Next slot.
- And what do you mean "(cause of BBC 6502 compatibility ?)"? I don't understand.
-
- Well garbage collection only applies if you're using a memory management scheme
- where the client requesting blocks doesn't later de-allocate them. The
- answer is to make sure you release all memory as and when you finish
- with it...
-
- |> It seams, that good old MSDOS v. 2.00 has better memory allocation routines
- |> than this one on the archimedes 8-(
- Oh yes? What was that then?
-
-
- |> "And now to something completely different"
- |>
- |> Has someone find a method to replace OS_ReportError with a WIMP-tasking
- |> errorviewer ?
- |> With the old (RISCOS 2.0) version of a errorviewer the machine freeze ....
-
- Well, write your own? If you're writing in C, it's not to difficult to
- produce a dbox which does the same thing. (The only thing you need to do
- is split up the text into several lines.)
-
- Come to think of it, isn't there an option on the ReportError swi to return
- immediately so you can just keep polling it? So why not use that?
-
-
- |> "And now to something completely different"
- |>
- |> Has someone find a method for a "forkviewer" ? A program that shows text
- |> but do not block the calling obeyscript.
- |> Example:
- |> obeycommand1
- |> ;eg. running a compiler
- |> forkview file
- |> ;eg. viewing a file with warnings
- |> obeycommand2
- |> ;e.g. running a linker
- |>
- Do you mean like *Type?
- I don't see the problem. Obey files don't get stopped by running applications.
- It's a well known problem that if you stick any lines (empty or not) after the
- Run <Foo$Dir>.!RunImage in a !Run file then that file remains open until the
- application terminates, so presumably the Obey mechanism is quite happy to
- nest stuff.
-
- Or do you mean that it should fork a wimp application to view the text?
- Well if you have a wimp app that views text, e.g. !ViewText, then try
-
- obeycommand1
- WimpTask !ViewText file
- obeycommand2
-
- (I'm pretty sure WimpTask is the command to launch another task)
- Actually, the problem with this is that it will need a wimppoll to happen before
- the new task is run, so this will only work if your obey file is being run
- in a taskwindow.
- The way to get around this is to write a small wimp app which starts off,
- executes the command using WimpTask, does a wimp poll asking for a null and
- then exits.
-
-
- |> "And now to something completely different"
- >
- |> so long
- |>
- |> "And now to something completely different"
- |>
-
- Oh PLEASE! The phrase is:
- "And now for something completely different"
-
- (The third word was wrong in every instance in your message.)
-
- Ian Griffiths
-