Compo returns

Second attempt…

Implosion has not been removed from SparkFS. The Compo ReadMe for example in C120demo/zip extracts with no problem on my working copy of the open source version, but not on 1.56. It seems the problem is the Zip module supplied with 1.56.

The last version of SparkFS I produced is available from my website (for free), just for this sort of eventuality, but it would be better if everyone went with the latest open version, in the long run it will be better and have more features. ROOL have put a lot of effort into it.

When working on the open source version of SparkFS, I used Spark and thought what a nice little program it is – one thing it can do is show the compression format of files in archives at a glance.

When working on the open source version of SparkFS, I used Spark and thought what a nice little program it is – one thing it can do is show the compression format of files in archives at a glance.

I’ve sometimes thought it’d be handy if the Filer had an extra column when in ‘Full info’ display mode, allowing a string of meta data of some sort (this would need the underlying OS_GBPB to request it of the filing system, and perhaps offer it around to other interested modules). In SparkFS’ case that would allow the compression type to be shown, and things like the dimensions of JPEGs and videos and similar.

Windows has this already – the columns available seem to depend on the context, so when inside a ZIP you get more options than for a normal display.

Here I go, throwing rocks, the implode/explode problem appears to be compiler related.

2002 compiler OK, modern ones not.

it’d be handy if the Filer had an extra column when in ‘Full info’ display mode, allowing a string of meta data of some sort

Yes could be a some fun there for not much extra complexity.

Going back it was said that it should be possible to use the RISC OS filer to display your own data, so that writing something like Spark one would not have to clone the filer. I think that is what can be done on Windows – “shell extension” (?) But it is not as easy to write file systems on Windows.

But it is not as easy to write file systems on Windows.

Not that I could write a filer*, but my experience is that it’s not as easy to write anything on Windows (or Mac for that matter).

* – Well, I suppose I probably could, but I’d have to learn a lot of stuff first…

Back on Compo… I’ve now succeeded in the attempts to get my brain to tell my fingers to write a composcript that can plot a series of different antenna patterns onto the compo canvas so the usr can see the results of different antenna arrays. Quite nice to see antenna field patterns pop up one after another. :-) At some point I’ll see if I can add the far field patterns that are related and indicate the ‘gain’, etc. Potentially a handy toy for RF use.

OK, I’ve now written a version of the Compo ‘antenna patterns’ script, etc, that seems OK here. Useful as a test/demo. All being well, I plan to write about this in detail for ‘Archive’ magazine as a first show of what using scripting can do! However since I need to provide the magazine with the files needed I decided to put them up at

http://jcgl.orpheusweb.co.uk/Compo/CompoRevived/Plotblank.zip

now so that people can have a look and:

1) Check I’ve not fscked it up. :-)

2) Gives some idea of the scope/potential.

To generate, just DND the ‘call_’ file on the Compo canvas.. As normal for !Compo the helper apps need to have been seen by the filer. The far-field antenna plot is a drawfile you then add to the canvas.

When considering (2) take into account the possibility that the field patterns could be animations, showing the field propagating outwards! Realy nice as a visual for explaining the mysteries of how those funny rooftop antennas behave…

I’d probably add labels, but you can work out which near field pattern is for which antenna arrangement by comparing it with the DrawFile’s graph of the far-field pattern.

As usual, the code by me is rubbish, it just staggers to doing what I wanted. I then quit whilst ahead. :-)

It’s good to see this demonstration on wave propagation and the potential of !Compo.

I looked at your code, but the equations are not explicit enough at my level. I will try to draw a radiation pattern with Pari/GP
There is a !Compo example/script showing superimposing a drawing on an image.

For your code, it must remain readable which is the case, perhaps a little more documentation on the subject to be illustrated…

It would be easier to explain in terms of complex numbers/values. The ‘C’ I wrote just tries to hack/duck past that. What I’ve done thus far is really aimed at a demo to show what !compo can do given the ability to ‘plot onto an object on the canvas! I’m doing an “Archive” article that also does this. Probably need another diagram for it. (The article also shows some example antennas which should help.) But I’m also wanting to avoid too much ‘antenna theory’ for Archive! May also do a webpage that deals with that and uses the images as illustrations. But one step at a time…

I have trouble reading my own code. I just shuffle the statements until it works. The only lessons I ever had on programming were for an ICL1900 back in the 1970s. FORTRAN, punched cards and tapes… Different world.

I’ve been experimenting with using Compo / CompoScript and quite impressed with what it can do for a task that is otherwise quite difficult. This is to generate illustrative ‘snapshot’ plots of the near-field patterns around array antennas. In my case so I can explain ‘how they work’ etc on webpages on the basis that a diagram can do more than lots of words or eqiations. Compo can now plot out examples like this

http://jcgl.orpheusweb.co.uk/Compo/CompoRevived/DRDpattern.jpeg

The small ‘object’ in the middle of the field pattern represents the antenna. In this case a simple 3-element ‘Yagi’ array of the kind used to get good VHF/FM radio reception. Fuller explanations would be in the text. But this shows neatly some aspects of their behavior which even antenna engineers tend to overlook!

More detailed explanation will be on the webpage once I’ve managed to write it.

@Jim

I’d certainly welcome a copy of the latter parts on scripting.

(Oh dear, a whole month has passed). I have scanned 20% of the manual, OCR’d the scanned pages, saved them as RTF, displayed them in TechWriter, copied the text into Ovation Pro (its TransRTF filter didn’t work for me) , applied styles and printed to PDF.

Here is a sample, being Appendices 4 (Introduction to CompoScript to acknowledge Jim’s request) and 5 (Keyboard shortcuts).

Now that I have a satisfactory workflow, I will continue with the project, with thanks to John McC and David G for their kind offers to share the load. It doesn’t seem so daunting now.

Ooh, it’s nice to see Compo being mentioned again. It’s an excellent application.

NetSurf’s old throbber animation was made with CompoScript.

Thanks, Bernard!

My printed copy says it is “Version 1” and I wrote “Sept 95” inside the front cover. I think it is pre CompoScript! Once your scans get to other parts I’ll compare with my copy to see if it might have anything useful to add to your later version. But I suspect your covers it.

Now that I have a satisfactory workflow, I will continue with the project, with thanks to John McC and David G for their kind offers to share the load. It doesn’t seem so daunting now.

Is the reduction in your level of “daunt” because your satisfactory workflow reduces it to a one-man job or because David G and I are here to help?

Whichever it is, it’s fine by me.

Ah, I wasn’t clear. The scanning, which is the quickest bit, is done. The OCR and transfer into OPro is straightforward, if clumsy. But then all the images have to be separately snapped from a running copy of Compo, and that takes a little more time. It’s hard to share that task without the original printed manual in front of you, and in any case the results go straight into the draft OPro document. My thanks were for the offers which I now think I don’t need to take up.

I note that my copy of the manual is dated 2005 and came with version 1.23c on CD-ROM. I have already spotted one difference from the newly redistributed 1.23a of 2004, and there my be more. The manual also has a 22-page supplement on ‘PCA Applets’ which I am of course including.

Then there is the one HTML page ‘CompoScript 1.22f Reference Manual’ (Lenny, 21-Oct-2003) which might be the basis for a script reference chapter in the new document. What do people think?

My thanks were for the offers which I now think I don’t need to take up.

Thanks received with pleasure.

My view is that all info should be included. Just how that is done and related into the main info, depends on what seems best I guess.

Okay, I will add a chapter called ‘CompoScript Language Reference’ based on anything I can find by either Rob Davison or ‘Lenny’ (who he?). This will complement the existing Appendix 4 ‘Introduction to CompoScript’.

And perhaps, Jim, you could help me to add a para or two about the input into Compo via CompoScript of data generated by program? There’s no mention of that possibility in what I have so far beyond a definition of the command ‘star’ showing only that it can interface via the CLI with Filer and WimpTask (for InterGif animation).

I’ll write something about it. At present I’m relying on the example’s source code to show it. In essence the script selects an ‘object’ on canvas and that then has the ‘focus’. In general, items are held as sprites in memory. So the code can then find the data and treat this as peeking and poking the data in a sprite file.

IIUC the script can also allow a series of objects on the canvas to be selected. This then is a series of blocks of memory with data in sprite format, and the program can select one or another as it chooses.

I think the code I wrote is in fact bodged into a different example that Rob wrote ages ago. The ‘clue’ there is that any code that looks tidy and rational was written by Rob or someone else who knows how to program properly. I just hack about until the bell rings!

I can’t recall who Lenny or Swampy were. But there did used to be what was essentially an informal user-group who exchanged emails. I’ll see if I can find some info on that by looking back at my ancient emails. Some of which are from the days when I used a different email prog which could compact data in a way I now can’t expand.

Occurs to me that it would make sense if I added a webpage that explains how to do what the new page does. Once I’ve done that it may serve as an illustration for what to add to the new manual.

Just to udate: Currently writing another item for Archive/Web that details how to use Compo to plot graphs like polar plots and/or colourfields. Whilst doing so another ‘penny dropped’…

Because it give access to the image data, Compo scripting could be used to read data from a graph. i.e. within the resolution of the graphic it has the poential to generate a series of data values for examination or analysis. That’s potentially quite handy. May give it a go at some point.

But at present focussing on demos of how to create various sorts of data graphics. And writing this up for people to see “how to”.

David, I think I bought SparkFS a long time ago. Does it read Rar files these days?

I think I bought SparkFS a long time ago. Does it read Rar files these days?

Had you thought about loading the latest version and testing what it does?

David, I think I bought SparkFS a long time ago. Does it read Rar files these days?

No it does not do RAR.

People have asked recently, but the only source code I can find is for a very comprehensive RAR implementation, splitting archives across discs and so on. In other words a lot of work is involved.

Perhaps fortunate that something like Zip was implemented when Zip files were a lot less complex than today.

SparkFS has of course been turned into an open source project/application by ROOL, so you can get the latest version for free – still won’t do RAR.

Does it read Rar files these days?

Does anyone use RAR these days?