sQLux v1.0 (The Xmas Turkey)

[ql_freak]

 
Now you're making me curious. What does PE_BGON do additionally compared to e.g. PIE_ON.

[RalfR]

Now you're making me curious. What does PE_BGON do additionally compared to e.g. PIE_ON.

That wasn't my intention. What I meant was that the frame buffers that ptr_gen uses via PE_BGON/OFF have nothing to do directly with SMSQ/E; the pointer I/F handles that alone.

And PIE/PEX/PICE hook into the pointer I/F; there's no other way.

[ql_freak]

I have now changed my sQLux setup. I'm using ptr_gen 2.01 and wman 2.01 from Dilwyns page and the latest pex version (3.4, Toolkit section: you should also download 3.3 cause this has the documentation).

It works on sQLux and cursor comes up even in none pointer programs and output to buried windows is working (tested with a SuperBASIC program). I have not yet installed PNICE (especially PICE), so you don't see the output if a window is below another. But as soon you bring it to front the actual output is visible. I will install PICE and report shortly. All tested on my GPD Pocket 2 in the Pub :-)

[RalfR]

I will install PICE and report shortly.

And please, show your BOOT file, so I do not need to reinvent the wheel .

[ql_freak]

Okay Pice must / should(?) not be used1), PEX includes the facility of PICE with its SuperBASIC command PEX_RFSH (first 2 parameters are the same as for the PICE command). My whole boot file (it are in fact 2) would be much too big. But it's very easy, here the most important:

Code: Select all

470 REMark other LRESPR which should/can be loaded before Pointer/Extended Environment
485 REMark LRESPR etc$&'ptr_gen'
486 LRESPR etc$&'ptr_gen_2v01'
490 REMark LRESPR etc$&'wman'
491 LRESPR etc$&'Wman_2v01'
493 LRESPR etc$&'pex34_byt'
495 LRESPR etc$&'hot_rext'
500 LRESPR etc$&"ptrmen_cde"
510 LRESPR etc$&"menu_rext"
520 LRESPR etc$&"qpac2"
530 REMark other LRESPR which should be loaded after Extended Environment
...
2000 :
2010 REMark now let them all go:
2020 HOT_GO
2030 REMark PE_BGON: REMark Allow output to windows (programs) while they are (partly) overlayed (in background) from other windows/programs. You can switch it off with PE_BGOFF. NOTE: REMarked cause not available with ptr_gen 2.01 instead we use:
2040 PEX_INI:PEON:PAUSE 20:PEX_RFSH 8,50
...
10000 REMark p.s.: PEX_RFSH has a bug with Pointer progs (e.g. Sysmon, it's not updated if it's e.g. buried by Channels) :‑(

Works like a charm :‑D But it seems in ptr_gen (or is it wman) 2.01 the windows do not yet have a shadow. But that's cosmetic and modern GUIs (e.g. Android, Windows) do not have shadow anymore. I'm not sure, if the PAUSE 20 is required, I think NOT, H.-P. always wrote very clean and tested code.
--
1) Maybe you should try the current PICE, it's documentation says, that it also supports the shadow of the windows. PEX_RFSH seems not to support it (if ptr_gen/wman V2.01 do have a shadow at all, dunno).

Edit: Some correction, especially PICE may eventually be used (either PICE or PEX_RFSH, not both, as both start a job). You can also try nice_obj, if the foreground job will have problems with user input (too slow).

p.s.: Cause I had to save my boot file to DOS (MicroEMACS cannot use "Copy to Scrap", at least not my installation, is there a scrap extension for uQLx/sQLux as for QPC which copies the scrap to the Windows clipboard?) I have loaded it into Notepad++ with the Syntax Hiliting (can be found somewhere on Dilwyn's page) – LOOKS GREAT! Does MicroEMACS also have a SuperBASIC mode (it has it for C).

[swensont]

> Does MicroEMACS also have a SuperBASIC mode (it has it for C).

It does for both SuperBasic and SSB. Load a file into Microemacs with either a _bas or _ssb extension. Then in the menu system click on "var./modes". Look at the bottom of the popup and click on "HILITE" mode. Then click on "Add a Mode". The display should change to show keyboards. Note that the hilite mode looks different between _bas and _ssb. Thierry added SSB support many moons ago.

Tim

[ql_freak]

Have now tried it, looks nice but IMHO Notepad++ does it better. The advantage of Emacs is, that it has a full blown programming language. I have written a very long time ago (when MicroEmacs was the only good GUI editor for Windows 3.1) block handling routines. The language is not bad, similar good as Archive for Archive.

When I find time, I will perhaps try to write a SuperBASIC edit mode, which will check erroneous lines at input as ED. As I don't want to write a SuperBASIC parser, my intention is to use a MultiBASIC job, to which any line is sent via a pipe (e.g. a named pipe, available e.g. from Hans Lub and of course in SMSQ/E), and another pipe which will respond with result or error and Emacs will refuse bad lines. This should be possible with MicroEmacs, a good shell, some ( perhaps) self written jobs and (<Minerva is out) MultiBASIC.

[ql_freak]

UQLX/sQLux has problems with the WIN disk images!

I have today lost my *.win "hard disk" for sQLux (this also has had happened with uQLx). Sometimes (not yet found the reason), the *.win file will be destroyed. Fortunately my *.win file has only a size of 132 MB, so I can make a lot of backup copies on Windows. Don't forget to make backup copies when using UQlX/sQLux(!). Sometimes QPC2 may help (you need it currently anyway, if you want to create directories on a UQLX/sQLux *.win drive) with defect *.win drives (can read it, albeit UQLX/sQLux can't).

It's shit. SBASIC is not compatible and has a lot of bugs (compared to Minerva) and UQLX/sQLux is not really stable (as QPC2 is mostly).

Edit: I think I must buy Q-emuLator

[XorA]

I did ask for testers for a version where MAKE_DIR works. I got silence.

[Derek_Stewart]

I did ask for testers for a version where MAKE_DIR works. I got silence.

I mentioned the MAKE_DIR not working on the QL Discord channel, or does this not count?