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

RE: Re (2): [Oberon] OFS directory-format documented ?
by Treutwein Bernhard 30 Nov '07

30 Nov '07

Hi, not very long ago I bought two twin CF-Card to IDE adapters each for about 10 EUR including shipping & handling from Honkong to Germany via Ebay. Although they say that they ship to the European Community, but I guess that they may ship also to South-Africa, so you could try it. This is a 5-Pack of Dual CF-IDE Adapters: http://cgi.ebay.de/5X-Doppelseitig-CF-Compact-Flash-zu-IDE-ATA-Konverter_W0… itemZ230196815086QQihZ013QQcategoryZ131824QQssPageNameZWDVWQQrdZ1QQcmdZViewIt em Interestingly the offer is identical in UK (just for fun I changed .de -> co.uk): http://cgi.ebay.co.uk/5X-Doppelseitig-CF-Compact-Flash-zu-IDE-ATA-Konverter… 0QQitemZ230196815086QQihZ013QQcategoryZ131824QQssPageNameZWDVWQQrdZ1QQcmdZVie wItem regards -- Bernhard Treutwein, IuK, Ref. III A 3 Bernhard.Treutwein(at)verwaltung uni-muenchen de > -----Original Message----- > From: PETER EASTHOPE [mailto:peasthope@shaw.ca] > Sent: Thursday, November 29, 2007 6:33 PM > To: oberon(a)lists.inf.ethz.ch > Subject: Re (2): [Oberon] OFS directory-format documented ? > > > Chris, > > cg> So flash cards are much better the fd0, > > Infinitely better in the sense that a CF can > replace a fixed disk as primary working storage. > > cg> ... but still not as reliable as IDE ? > > I've had two failures in about two years of use. > In both cases the CF could be reformatted; > then continued to operate with no problem. > > cg> I couldn't buy IDE-to-flash-card-adaptors here, > > You don't need to buy the adapter and make the > parts to mount it as I did. At least one company > sells adapters ready to install. You can have > them sent by mail? > > Of course, most view this arrangement as obsolete. > With Aos you can connect USB flash storage. Still, > CF is faster according to my test. > Desktops.OpenDoc http://carnot.yi.org/ExampleSystems.html > > cg> I still use fd0 to move stuff between boxes 3m apart. > > What happened to Ethernet and FTP? > > Regards, ... Peter E. > > > http://carnot.yi.org/ > > > >

1 0

Re (2): [Oberon] OFS directory-format documented ?
by PETER EASTHOPE 29 Nov '07

29 Nov '07

Chris, cg> So flash cards are much better the fd0, Infinitely better in the sense that a CF can replace a fixed disk as primary working storage. cg> ... but still not as reliable as IDE ? I've had two failures in about two years of use. In both cases the CF could be reformatted; then continued to operate with no problem. cg> I couldn't buy IDE-to-flash-card-adaptors here, You don't need to buy the adapter and make the parts to mount it as I did. At least one company sells adapters ready to install. You can have them sent by mail? Of course, most view this arrangement as obsolete. With Aos you can connect USB flash storage. Still, CF is faster according to my test. Desktops.OpenDoc http://carnot.yi.org/ExampleSystems.html cg> I still use fd0 to move stuff between boxes 3m apart. What happened to Ethernet and FTP? Regards, ... Peter E. http://carnot.yi.org/

1 0

OFS directory-format documented ?
by easlab＠absamail.co.za 29 Nov '07

29 Nov '07

Is there any documentation/description available for the N-O file system [OFS] structure, without reading the source code files ? Using Partitions.ShowBlocks I can see that: - the directory starts at block 1032, - each entry is 40 bytes wide, - they are [re]arranged alphbetically by fileID, - the 'table' starts: 00: <4 bytes> <number of dir-entries> 10:<32 chars for 1st file ID> <8 bytes= 2 words> seemingly only the 1st word after the ID-field is used? Simplistically this would be an offset ? But when I re-formatted [which I later discovered does NOT delete/reset previous bytes] and entered 4 'new' files: one, two, three, four, of sizes as below, I see: FuleID arg-word File-size Dif-words =============== use const width font to view=== one 5829 75 29 two 5858 75 29 three 5887 75 29 four 5916 115 29 So the [apparently redundant] word-val 29 is independant of the file size ? In fact for a big-directory all the 'arg-words' are multiples of 29; and the 1st chronologically entered File's arg-word: 5829; equals 201*29. Thanks for any feedback, == Chris Glur. PS. yes you guessed that I want to recover some lost valuable files.

2 1

Development tools.
by Chris Glur 29 Nov '07

29 Nov '07

IMO EditKeys.Tool which apparently dates back to 1993 is inappropriately ignored. Perhaps because it's mostly not understood. Finally this seems to be a tool which is practical for entering source code by picking-structures instead of chars. I think it can insert the structures already pretty-printed too. The default file containing the defined key-macros is: EditKeys.Text One of the behaviour which I don't understand, and hopefully we can collaboratively understand is that when I select my chosen filename & activate EditKeys.Definitions ^ , the viewer: EditKeys.Definitions is loaded, instead of <my chosen filename> as expected. Never mind, <my chosen filename> also becomes 'active for macros' indeed so do all other 'text gadgets'. Let's immediately see if/how to create the nested structure:-- while if B then Repeat Stmt Until else if B2 then B2ifStmt else B2elseStm end-while The existing macro definition for 'WHILE' is near the bottom of EditKeys.Text . Type 'WHILE'<Insert>. That's the 'insert-key' immedidte after the string "WHILE". I don't like the result, so I've made an extra macro called 'w' :- w("WHILE " write caret " DO" write cr tab cr "END;" write) As described this can be tested manually, but initially just include the new definition[s] in EditKeys.Text and do EditKeys.Read EditKeys.Text ~ to reload the extended macro set. Now test it with: w<Insert> Wow, it even copies the colours/font of the macro definition. But it doesn't indent properly eg. if the initial token "WHILE" is not at the line extreme left. Let's see if we can add an IF-THEN-ELSE macro which DOES indent appropriately ? if("IF " write "Condt" caret " THEN Stmt" write " END" write cr indent "END;" write) This tests OK [in blue for me ;-]. But let's see how/if it indents inside of: -While <3spaces.if -endWhile ?? -While IF Condt THEN Stmt END ELSE Stmt END; -endWhile Yes! it indents OK. So now building towards our final structure, let's try: while if B then Stmts end else if B2 then B2ifStmt else B2elseStm end end-while These things need colours to help manage? So far the 'w' and 'if' macros give me:- WHILE bwhile DO manually entered stmt1; IF Condt THEN Stmt END ELSE IF Condt THEN Stmt END ELSE Stmt END; END; END; Ok, the 2nd nested If-Then-Else indents only to the previous ELSE. Perhaps if the macro's 'ELSE line' started with 'tab' ? Offiicially C has at least 2 recommended indentation styles for If-Then-Else. Perhaps someone can juggle this to get good auto-pretty-printing ? The macros can be tested maually, without resaving the EditKeys.Text file by cycling thru: 1. clear the previous version: EditKeys.Reset 2. relaod the default macros: EditKeys.Read EditKeys.Text ~ 3. write the new macro at the end of a Document-frame 4. select it 5. read & enable the new macro for testing: EditKeys.Read @ 6. test the new macro Q - what is the meaning in EditKeys.Tool , of; "Text stretches are pushed on a parameter stack when processed." ? 'processed' by what ? Q - if "^0" inside a macro definition means; pop the parameter & insert the corresponding text at the macro's current writing position, then how is the parameter initially 'pushed' ? Thanks for any feedback, == Chris Glur. PS. I'm trying to put this to: http://www.edgar-schwarz.de/mm/ProblemsPendingSolutions if/when I can rediscover how to drive the wiki. It mentions login -> userName, ID ..? That's why we need more menu-driven systems instead of having to remember ?

1 0

Re: [Oberon] OFS directory-format documented ?
by Chris Glur 29 Nov '07

29 Nov '07

PETER EASTHOPE <peasthope(a)shaw.ca> wrote:- > Chris, > > Your approach is interesting; sorry, I can't > help with documentation for the file system. > > cg> ... recover some lost valuable files. > > I recovered some data from a flash card by > copying successive chunks of the partition > to a file. Each chunk was searched with > ET until the lost text was found. Not a > comprehensive method but it recovered a few > days worth of my journal. > > Partitions.PartitionToFile IDE1#03 Blocks <n x 50000> 50000 ~ > Start with n=0 and work until you find what you want. > > Now I back the journal several times a day. > Nice idea, I'll try that too. So flash cards are much better the fd0, but still not as reliable as IDE ? I couldn't buy IDE-to-flash-card-adaptors here, so I still use fd0 to move stuff between boxes 3m apart. >From a previous mail suggesting a solution for [your?] PS2-mouse, I got the ideas to fix when N-O backup hangs: just see the 2 modules listed in the Trap-listing and System.Free <them>. This allows you to continue without booting and losing days of 'setup'. Now I've got a similar linux fd0 'false read-only setting' when the fd0 fails once, which also can't [with my present knowledge] be fixed without rebooting and losing days of 'setup'. I use the one box with linux-oberon for inet fetching and I prefer the other 2001:N-O box for processing the files. Thanks, == Chris Glur.

1 0

Re: [Oberon] OFS directory-format documented ?
by PETER EASTHOPE 28 Nov '07

28 Nov '07

Chris, Your approach is interesting; sorry, I can't help with documentation for the file system. cg> ... recover some lost valuable files. I recovered some data from a flash card by copying successive chunks of the partition to a file. Each chunk was searched with ET until the lost text was found. Not a comprehensive method but it recovered a few days worth of my journal. Partitions.PartitionToFile IDE1#03 Blocks <n x 50000> 50000 ~ Start with n=0 and work until you find what you want. Now I back the journal several times a day. Regards, ... Peter http://carnot.yi.org/

1 0

OFS directory-format documented ?
by Chris Glur 28 Nov '07

28 Nov '07

My previous mail lack a proper header [N-O authenticating mailer not yet stable] and may have looked like spam. ---------------------- Is there any documentation/description available for the N-O file system [OFS] structure, without reading the source code files ? Using Partitions.ShowBlocks I can see that: - the directory starts at block 1032, - each entry is 40 bytes wide, - they are [re]arranged alphbetically by fileID, - the 'table' starts: 00: <4 bytes> <number of dir-entries> 10:<32 chars for 1st file ID> <8 bytes= 2 words> seemingly only the 1st word after the ID-field is used? Simplistically this would be an offset ? But when I re-formatted [which I later discovered does NOT delete/reset previous bytes] and entered 4 'new' files: one, two, three, four, of sizes as below, I see: FuleID arg-word File-size Dif-words =============== use const width font to view=== one 5829 75 29 two 5858 75 29 three 5887 75 29 four 5916 115 29 So the [apparently redundant] word-val 29 is independant of the file size ? In fact for a big-directory all the 'arg-words' are multiples of 29; and the 1st chronologically entered File's arg-word: 5829; equals 201*29. Thanks for any feedback, == Chris Glur. PS. yes you guessed that I want to recover some lost valuable files.

1 0

ANN: UnixAos rel. 0.6
by Guenter Feldmann 22 Nov '07

22 Nov '07

Hi all Oberon and Bluebottle fans The last alpha release (0.6) of UnixAos is now available at "ftp://ftp.informatik.uni-bremen.de/home/fld/oberon/UnixAos" It is usable on the x86 variants of Solaris, Linux and MacOSX (Darwin). Starting with this release UnixAos comes up with the BlueBottle window- manager. The Oberon desktop has been integrated and can be started from within the windowmanager. Current state of the system: - The SAFE attribute of active objects has no effect. As a result the system gets frozen (or partly frozen) if a message sequenzer (WMMessages.MsgSequencer) traps! Especially in the Darwin port this happens often as a result of an unidentified X11-IO-error. - There is no sound output. - The system contains no hardware drivers. If you want to read from or write to hardware interfaces you must access these as ordinary Unix device files. - The network stuff (Unix/AosUDP, Unix/AosTCP) needs more testing. Is seems to be very slow when using the BB browser. The next release will have support for 'safe' active objects and hopefully be in beta state. -- Guenter

1 0

Re: Glur re. google-maps ?
by Chris Glur 10 Nov '07

10 Nov '07

> I have had a look but can't recognise the airfield, do you have > the lat/lon? The river is quite full and clearly visible although > the picture is not as high resolution (1 metre) as some maps in GE. > I believe the airfield is lat=27 30.05, lon ~ 26 53' Apparently the road to JHB is at ~50 degrees - leaving Parys and a rail-line leaves at 100 degrees, later turning to ~135 degrees; and another rail-line leaves at 180 degrees, later turning to ~255 degrees. Apparently the airfield is just north of the JHB road, perhaps on a small by-road ? > > I'm sure we can do something. GE will print a map or I can do a screen > save in *.jpg Thanks, what ever's easiest to email. == Chris Glur.

1 0

ETH-Oberon's accessable common platform
by Chris Glur 09 Nov '07

09 Nov '07

Aubrey McIntosh wrote: >From time to time, I think it would be good if the Oberon community >had some common platform that most of us have access to. I'm not sure what you mean, but http://www.edgar-schwarz.de/mm/SwOberon works well. It's got some sub-headings: http://www.edgar-schwarz.de/mm/ArmObero http://www.edgar-schwarz.de/mm/OberonContributions http://www.edgar-schwarz.de/mm/BugFixes etc... == Chris Glur.

1 0