January 2007 - Oberon - mailman3.inf.ethz.ch

Bluebottle and Intel Mac without VM
by Alexander Hofmann 15 Feb '07

15 Feb '07

Hi, Does the Bluebottle ISO image currently available for download support USB-Keyboards within the Bootloader? I'm currently trying to boot Bluebottle on my Intel Mac Book without emulation, but beside the Message "OBERON loading..." nothing serious happens, till the Screen turns black and the System hangs... I would like to change some of the config-strings, but can't get into the Editor, holding down [shift] doesn't work... (I remember some previous version of Bluebottle succeeded in displaying the config-strings) Does anyone else have an Intel Mac and is trying to boot (or even succeeded in booting) Bluebottle? best regards, Alexander Hofmann. PS: I've already sent this message some days ago, but didnt' get it back from the list, thats why I'm sending it again... if you already got it, please excuse me...

2 2

format of source.arc
by Brantley Coile 01 Feb '07

01 Feb '07

Can someone send me the module to pull apart the .arc files for the Native Oberon?

2 1

Parallel processing
by Douglas G. Danforth 30 Jan '07

30 Jan '07

"In an InformationWeek article entitled 'Where's the Software to Catch Up to Multicore Computing?' the Chief Architect at IBM gives some fairly compelling reasons why your favorite software will soon be rendered deadly slow because of new hardware architectures. Software, she says, just doesn't understand how to do work in parallel to take advantage of 16, 64, 128 cores on new processors. Intel just stated in an SD Times article that 100% of its server processors will be multicore by end of 2007. We will never, ever return to single processor computers. Architect Catherine Crawford goes on to discuss some of the ways developers can harness the 'tiny supercomputers' we'll all have soon, and some of the applications we can apply this brute force to."

2 2

N-O: is this quirk mentioned anywhere ?
by Chris Glur 28 Jan '07

28 Jan '07

Since the user-interface is the most important aspect of stuff, it's good that the N-O documentation tell early what to do when [the new user] has unintendedly closed all his text-frames. Many potential users simply abort their early efforts at ETH-Oberon, when they hit an unexpected quirk. Like 'ghost in the machine', when no one would tell him how to 'scroll' big images, especially with vga setting. And where such a big image appears early in the documentation. Here's another quirk: * the pasteing to an Edit.Open frame is active/available immediately after it's visible. * the pasteing to MenuFrame of an Edit.Open frame is similarly immediately valid. ! But the renaming of a Desktops.OpenDoc-type frame by pasteing a new ID tothe NamePlate is NOT valid until <cr> ! So when the user renames and thinks he has saved the newly-named file he is mistaken. That's also why the System.Log is good. But for a cluttered application [like FontEditor] where the System.Log is covered and not visible == unexpected problems. == Chris Glur.

1 0

HowTo get Scanner's position at found token ?
by Chris Glur 27 Jan '07

27 Jan '07

Since the Texts.Scanner is a descendant [object] of Texts.Reader I'm guessing that Texts.Reader PROC Pos (VAR R: Reader): LONGINT might be used ? But how ? I need to: set caret = ok init Scanner from caret position = ok Scan until 'Token' is found OR EOT = ok Know the position [: LONGINT] where the 'Token' was found = problem?! Thanks for any feedback, == Chris Glur

1 0

Vinci environments and the current transformation matrix, CTM.
by Easthope 24 Jan '07

24 Jan '07

Folk, I've been working on another Vinci drawing and thinking about a better way to handle dimensioning. A general objective is to size the font and line widths independently of the drawing scale. Why? Well, suppose you make a drawing at a 1:1 scale for A4 paper. Later you begin a new drawing at 1:100 scale for A4 paper. The original drawing might serve as a template for the new drawing but the lettering in the dimensions and the line widths of the new drawing should not be 1/100 the size of the lettering and line widths in the original. In fact, lettering and lines widths on the new drawing are best about the same as on the original. So how can this be achieved? Would be convenient to define globally a graphics state or environment for font size and line width, which can be retained independent of changes incurred as the drawing is built. After drawing a feature, one could temporarily drop into a "dimensioning mode", apply a dimension and then return to the "drawing mode". To find a way to do this I reread Vinci.Text, The Graphical Description Language Vinci. These questions remain. -- What exactly is an environment? -- Which of the statements invokes a new environment nested "within the current environment"? (Quoting the manual.) -- What is a graphics state? -- Is there always just one CTM which changes as the environment changes or does each environment have a CTM? -- What is the structure of the CTM? How is it applied to a graphical vector? Eg. [[Sx 0 Tx] [0 Sy Ty]] . [Vx Vy 1]^T =? [SxVx+Tx SyVy+Ty]^T There are two motivations for mentioning these topics. In the short term I can hope that someone else can shed some light. In the long term my questions might motivate better documentation in future work. Apart from this I need to study the vast source code. Regards, ... Peter E. Desktops.OpenDoc http://carnot.pathology.ubc.ca/

1 0

RFC: RAIF (Redundant Array of Independant Filesystems)
by Patrick.Hunziker＠unibas.ch 21 Jan '07

21 Jan '07

Hi all: To solve some problems with data backup, I propose a RAIF (analogue to RAID) which stands for 'redundant array of independant filesystems' allowing transparent handling of identical files on two different filesystems. A nice feature would be that 'compressing file systems', 'remote file systems' etc can be neatly reused. Please comment this first draft critically, or improve directly the proposed implementation (which compiles, but is untested): Some features are not yet done (finalize, copy etc). MODULE RAIF ; (** AUTHOR "Patrick Hunziker"; PURPOSE "RAIF=Redundant Array of Independent Filesystems"; **) IMPORT AosFS; TYPE RAID5FS=OBJECT (AosFS.FileSystem); (*similar to RAID5 concept offering redundant data storage*) VAR FSA-,FSB-: AosFS.FileSystem; (*may also be archive-based, remote, etc; this will need decoupling of the slow remote FS in separate process*) incremental*:BOOLEAN; (*adds new file version without deleting old one, which is renamed ('somename.N.Bak'*) PROCEDURE &Init*(FSA,FSB: AosFS.FileSystem; incremental:BOOLEAN); BEGIN SELF.FSA:=FSA; SELF.FSB:=FSB; SELF.incremental:=incremental; END Init; PROCEDURE Old*(name:ARRAY OF CHAR):AosFS.File; (*todo: if file in both file systems, rather choose the newer one ?*) VAR f:AosFS.File; name0,fullname: ARRAY AosFS.PrefixLength+AosFS.NameLength OF CHAR; prefix: ARRAY AosFS.PrefixLength OF CHAR; BEGIN AosFS.SplitName(name,prefix, name0); AosFS.JoinName(FSA.prefix,name0,fullname); f:=AosFS.Old(fullname); IF f=NIL THEN AosFS.JoinName(FSA.prefix,name0,fullname); f:=AosFS.Old(fullname); END; RETURN f END Old; PROCEDURE New*(name:ARRAY OF CHAR):AosFS.File; VAR f:AosFS.File; name0,fullname: ARRAY AosFS.PrefixLength+AosFS.NameLength OF CHAR; prefix: ARRAY AosFS.PrefixLength OF CHAR; BEGIN AosFS.SplitName(name,prefix,name0); AosFS.JoinName(FSA.prefix,name0,fullname); (*default*) f:=AosFS.New(fullname); RETURN f END New; PROCEDURE Delete*(name: ARRAY OF CHAR; res: LONGINT); VAR f:AosFS.File; name0,fullname: ARRAY AosFS.PrefixLength+AosFS.NameLength OF CHAR; prefix: ARRAY AosFS.PrefixLength OF CHAR; resA,resB:LONGINT; BEGIN AosFS.SplitName(name,prefix,name0); AosFS.JoinName(FSA.prefix,name0,fullname); AosFS.Delete(fullname,resA); AosFS.JoinName(FSA.prefix,name0,fullname); AosFS.Delete(fullname,resB); (*ToDo: what is the correct return value (or are 2 return values needed?)*) END Delete; PROCEDURE Register*(f:AosFS.File); VAR name, name1,name2: ARRAY AosFS.PrefixLength+AosFS.NameLength OF CHAR; prefix: ARRAY AosFS.PrefixLength OF CHAR; res:LONGINT; overwrite:BOOLEAN; BEGIN AosFS.Register(f); f.GetName(name); AosFS.SplitName(name, prefix, name1); IF prefix=FSA.prefix THEN AosFS.JoinName(FSB.prefix, name1, name2); ELSIF prefix=FSB.prefix THEN AosFS.JoinName(FSA.prefix, name1, name2); ELSE HALT(100); END; IF incremental THEN overwrite:=FALSE; HALT(101); (*incomplete implementation: ToDo: check for identity, dates; avoid name clashes*) ELSE overwrite:=FALSE END; AosFS.CopyFile(name,name2,overwrite, res); END Register; END RAID5FS; END RAIF. ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

2 1

FAT Filesystem on BluebottleServer allows duplicate files with same name
by Patrick.Hunziker＠unibas.ch 21 Jan '07

21 Jan '07

To my surprise, Files generated on a Bluebottle FAT partition in a folder sometimes have the same name, as visible in System.Directory, visible in the Bluebottle File Browser, and visible when accessing the folder in Windows. There seem no "nonprintable" characters in the filename (a problem sometimes occurring when using the Store Button in the Oberon Gadgets Interface). When trying, in Bluebottle, to Zipfile.Add a list of files containing that duplicate, the duplicate file problem becomes manifest. The same problem becomes manifest when booting windows and trying to copy that folder. A visualisation of the problem is on http://www.computational.ch/FATbug.jpg Greetings Pat ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

2 1

Does BB/Aos show all the non-ascii chars ?
by Chris Glur 20 Jan '07

20 Jan '07

And if so, what does Char(160) look like ? Does BB/Aos have the FontEditor or equivalent ? Newsgroups: comp.text,de.comp.text.misc Subject: What does char(160)='non-breaking space' look like? Some times I get text which has characters which aren't rendered by my file-viewer. I have facilities to build such characters and add them to the font-set; which I incrementally do some times. After determining the byte-value of the non-displaying char, I just need to know what it's supposed to look like, in order to manually build and add it to the font-set. 3 new chars which I'd like to add to the font set have byte-values of 147, 148 and 160. And from a list of all 256 chars I find:-- ç=93 == left double quotation mark, 93  “ á=94 == right double quote mark 94  ” =A0 == non-breaking space A0   -->   Of course I can guess what a 'left double quotation mark' looks like. And when I don't know what the char looks like, I embedd it in some existing text and view it eg. with mozilla, which DOES render these extra/strange chars. But currently my mozilla is broken. Q - where can I see what a "non-breaking space" looks like ? Thanks for any feedback, == Chris Glur. PS. this post [originally] contains some of these non-ascii chars, which may be filtered out in transmisssion ? çdirectorá, in relation to a company, means a director as defined in section 1 (1) of the Companies Act; 15 (1) or (2); 1 (1) Hex.Open ^ ><><><><><><><><><><><><><><><><><><

1 0

Bug in AosUnzip or AosInflate?
by Yaroslav Romanchenko 20 Jan '07

20 Jan '07

My new application loads its own data from archive. By the way there is two applications "IFS Viewer" and "Morphing IFS ScreenSaver" for Bluebottle :) Data is represented as binary (TRN) or text (XML) files. Binary files are from FDESIGN program released to the public domain by Doug Nelson. When data is stored in "tar" archive, loading is worked Ok, but relatively slow (about 15 sec. on Celeron 540 MHz). And when the same data is stored in "zip" archive, loading speed is very high, but some data corruption occurs. Last byte of data file is allways set to 0X. It's very simple to see if do replacement: Data = "Sage.IFSData.tar"; to Data = "Sage.IFSData.zip"; in Sage.IFSData.Mod Examples of fractals that affected by last byte corruption: ZIGZAG2.TRN, WEB.TRN, WINDY.TRN, TELEPHON.TRN, WARM.TRN, and etc. Applications can be downloaded by this link: http://www.box.net/public/5axkxyc3yo --- Cheers, SAGE http://sage.h15.ru/

2 1