Shop OBEX P1 Docs P2 Docs Learn Events
Mac/Linux/Windows IDE - Ver 0.19.3 "Now with new improved PropBasic" release - Page 23 — Parallax Forums

Mac/Linux/Windows IDE - Ver 0.19.3 "Now with new improved PropBasic" release

1202123252652

Comments

  • BradCBradC Posts: 2,601
    edited 2009-05-16 01:37
    W9GFO said...
    Feature request and small bug

    Os X 10.5.7 Intel
    0.18pre-8
    Parallax Font

    The bug, text gets garbled - in this example The Pub's name is Sweep and I put the cursor between the "e" and "p" and backspaced three times. It remains garbled while typing the letters back in, it will stay this way until I scroll.

    attachment.php?attachmentid=60953

    Feature request, it would be nice for me if the main window could be split in two so that each side could show a different tab. Or be able to have two instances of BST running side by side.

    2 days ago I did some major upgrades related specifically to font rendering, so I suspect this one is squashed.
    I've had a *lot* of trouble rendering the Parallax font properly and I've had to add quite a few hacks into the software to make it render acceptably on non-windows machines (nto to mention having to strip the bitmaps out of the font completely).

    The split-pane thing is on my todo list, but with the way I store configuration and project information and the file-association machinery associated with OSX and Linux makes running more than one copy quite complex. This is why I allow you to associate a propeller/port with each editor tab, so you can very easily manage multiple propellers in the one window.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • W9GFOW9GFO Posts: 4,010
    edited 2009-05-18 03:03
    Found another anomaly,

    Os X 10.5.7 Intel
    0.18pre-8

    In the menu on the left that lists the .spin files available - files ending in .SPIN (uppercase) do not show up.

    Rich H
  • BradCBradC Posts: 2,601
    edited 2009-05-18 10:18
    W9GFO said...
    Found another anomaly,

    Os X 10.5.7 Intel
    0.18pre-8

    In the menu on the left that lists the .spin files available - files ending in .SPIN (uppercase) do not show up.

    Rich H

    Well spotted. The file selector was case sensitive. I've had to write a proper case-insensitive matcher to allow it to work properly on case-sensitive filesystems.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • AleAle Posts: 2,363
    edited 2009-05-18 11:03
    Hei Brad:

    I forgot to mention that from the listing window at least in OSX you cannot copy, you can select but copy does not copy anything to the clipboard.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Visit the home of pPropQL, pPropQL020 and OMU for the pPropQL/020 at omnibus.uni-freiburg.de/~rp92
  • BradCBradC Posts: 2,601
    edited 2009-05-18 11:33
    Ale said...
    Hei Brad:

    I forgot to mention that from the listing window at least in OSX you cannot copy, you can select but copy does not copy anything to the clipboard.

    Ta, I'll get that fixed. Oh, and I'm not ignoring your questions on the @/@@/@@@ stuff, I just *really* have to be in the zone to dive back into that code and get my head around what does what, and I've just not been there lately.. I will get to it.

    I do remember that @@@ is actually valid SPIN syntax and using @@@ in a SPIN block the Parallax compiler will generate something different to what homespun and bstc does. I was always going to re-assign that operator but after much hassle getting it implemented I don't think anyone actually ever used it.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • EricREricR Posts: 35
    edited 2009-05-18 12:41
    Brad,

    (found selections hard to see with Color Syntax Highlighting)

    The background of highlighted items with 'Edit/Find' can sometimes match the background of the text. This is true when Color Syntax Highlighting is in use, and you are in a 'PUB' routine. Please see attached picture for comparison of 'find' results with Color Syntax Highlighting off and on.

    I would like to suggest that when the found area is in a PUB routine, that find highlighting be yellow, or some other contrasting color.

    (insert key not available on Macbook)
    Regarding an earlier post, when I dock my Macbook at home with an attached PC keyboard (HP), there is an insert key. This key works well with BST. (background)The Aluminum Macbooks have fewer keys than the Macbook Pros, and there seems to be no way to simulate an insert key with the Macbook's keyboard.

    Eric

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Eric

    Post Edited (EricR) : 5/18/2009 12:52:22 PM GMT
  • BradCBradC Posts: 2,601
    edited 2009-05-18 12:47
    EricR said...
    Brad,

    (found selections hard to see with Color Syntax Highlighting)

    The background of highlighted items with 'Edit/Find' can sometimes match the background of the text. This is true when Color Syntax Highlighting is in use, and you are in a 'PUB' routine. Please see attached picture for comparison of 'find' results with Color Syntax Highlighting off and on.

    I would like to suggest that when the found area is in a PUB routine, that find highlighting be yellow, or some other contrasting color.

    Eric

    Ahh, the wonders of default OSX colours. I have my machines set to the brushed metal theme for that reason [noparse];)[/noparse] The highlight colour is specified by the OS. I guess I need to add an option to override it if you are using OSX with that theme. (It's the only one that suffers from it).

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • EricREricR Posts: 35
    edited 2009-05-18 13:04
    Thanks Brad,

    Sorry I can't seem to delete the duplicate posting.... You answered really fast (wow) and what I thought was a post edit became anther post. Then it seems that if I delete the duplicate post, the original vanishes. I am going to just leave it there I think.

    Thank you for you response. If you find time to add the over ride of OSX colors that is great.

    (on insert modes in MacBook Aluminum, an earlier post)
    I found that if I attach a 'PC' keyboard with an Insert key, I can go from insert to overtype mode.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Eric
  • BradCBradC Posts: 2,601
    edited 2009-05-18 14:08
    EricR said...
    T\
    (on insert modes in MacBook Aluminum, an earlier post)
    I found that if I attach a 'PC' keyboard with an Insert key, I can go from insert to overtype mode.

    Yes, it _did_ however on OSX I've now removed that key binding and all the latest -pre releases use shift-delete. At least those keys are on all Macs [noparse]:)[/noparse]

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • T ChapT Chap Posts: 4,223
    edited 2009-05-18 16:53
    The problem in PC .17 where on boot the bst would always load up an old file stating 'we are recovering..." has been solved in the latest snapshot.
  • BradCBradC Posts: 2,601
    edited 2009-05-20 07:17
    Ale said...

    0.18pre8 Mac OSX: it hangs when the terminal is receiving and I try to compile and load again (or I use the option "disconnect" from the menu). Force Quit does not terminate the program neither does kill -9 (probably are the same). Turning it off and on again seems to solve the problem lol.gif

    I can _not_ reproduce this problem here (sigh).

    What I did to try and reproduce it was write a little prop program that hammers strings with a terminating <CR> out the serial port continuously at 115200 BAUD. While this was merrily scrolling past in the terminal I compiled about 10 different SPIN programs ranging from very large to very small and then tried doing a series of connect/disconnect cycles with the terminal.

    I forgot on OSX when you disconnect the terminal it resets the prop [noparse]:)[/noparse] I've tried, tried and tried some more to find a workaround for this, but OSX just will _not_ honour the HUP/HUPCL flags in its termios structure. As soon as you disconnect the port it changes DTR state.

    Anyway, I digress.. I then repeated the procedure while trying various loads and still can't make it fail. I'm testing on a Core Duo 1.8Ghz Mac Mini with OSX 10.5.7 and bst0.18-pre8 downloaded directly from the archive. Can you possibly give me any more information that might help me reproduce this?

    I did find one bug where we were dropping characters if bst could not keep up with the incoming data, and I've fixed that.. but I can't get it to hang/lockup. I'll try and check it on my PPC machine under 10.4 later if I can get it running

    One last thought.. was it specific to a particular SPIN file? I wonder if something in that file is tickling a bug in the compiler that is doing something funky with threading?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • AleAle Posts: 2,363
    edited 2009-05-20 08:52
    Brad: I do not think there was anything special about the compiled file. The strings that I sent from the prop where not particularly ASCII friendly... because the program was just executing badly due to some bugs in my spin code, so it was sending whatever was in memory :-(. It happened twice. I can try to strip that portion of the program and try to reproduce it. Now that I think of it, it could be related to some FT232 driver bug.. I'm just not sure which one I have. I can try to reproduce it as it is and test later with the newest incarnation (the one that was out a couple of month ago).
    I fixed my spin bugs (did I say I think it has a very strange syntax? This guy Hanno says that it looks like C, well I think he never used python).

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Visit the home of pPropQL, pPropQL020 and OMU for the pPropQL/020 at omnibus.uni-freiburg.de/~rp92
  • BradCBradC Posts: 2,601
    edited 2009-05-20 08:57
    Ale said...
    Brad: I do not think there was anything special about the compiled file. The strings that I sent from the prop where not particularly ASCII friendly... because the program was just executing badly due to some bugs in my spin code, so it was sending whatever was in memory :-(. It happened twice. I can try to strip that portion of the program and try to reproduce it. Now that I think of it, it could be related to some FT232 driver bug.. I'm just not sure which one I have. I can try to reproduce it as it is and test later with the newest incarnation (the one that was out a couple of month ago).
    I fixed my spin bugs (did I say I think it has a very strange syntax? This guy Hanno says that it looks like C, well I think he never used python).

    Ahh.. it never occurred to me to use non-printable characters. I have *no* idea what will happen in that case. I'll build some test cases and see if I can get it to explode. If it explodes under OSX with that it'll play up under Linux and Win32 also.

    I really feel like SPIN is a hybrid of Pascal and Python myself.. (both languages I love and use often!)

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • T ChapT Chap Posts: 4,223
    edited 2009-05-20 16:42
    Sorry if this is discussed already:

    18 pre8 XP While typing, sometimes the text does not show up until I drag the mouse over the area as in selecting the area, then it appears.

    All versions

    Open Recent never has the most recent file

    if x := 100 will compile.

    Thanks for the hard work Bradc

    Post Edited (TChapman) : 5/20/2009 10:28:33 PM GMT
  • SSteveSSteve Posts: 808
    edited 2009-05-20 22:32
    greitz said...
    How does the Catalina compiler compare to ImageCraft's C compiler (not necessarily the IDE) in terms of performance and features? (versus being free and running on Mac/Linux, which is awesome!)
    Wrong thread. You want this one.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    OS-X: because making Unix user-friendly was easier than debugging Windows

    links:
    My band's website
    Our album on the iTunes Music Store
  • T ChapT Chap Posts: 4,223
    edited 2009-05-22 18:09
    Would predefined hotkeys be a waste to impliment considering there are already hotkey softwares?

    For example control>w inserts waitcnt(80_000_000 + cnt) wherever the cursor is.
  • BradCBradC Posts: 2,601
    edited 2009-05-23 03:04
    TChapman said...
    Sorry if this is discussed already:

    18 pre8 XP While typing, sometimes the text does not show up until I drag the mouse over the area as in selecting the area, then it appears.

    I'll look into this next time I've got windows open. I have made some pretty significant changes to the editor since -pre8 so we'll see if that is fixed.
    TChapman said...


    Open Recent never has the most recent file


    Hrm, most peculiar. How are you opening the most recent file? (File menu, Directory selector or drag and drop)
    TChapman said...

    if x := 100 will compile.
    I'd hope so. It will always execute as the result is always true. Propeller Tool compiles it exactly the same.
    TChapman said...
    Would predefined hotkeys be a waste to impliment considering there are already hotkey softwares?

    For example control>w inserts waitcnt(80_000_000 + cnt) wherever the cursor is.

    I can add it to my todo list, but it's a _long, long, long_ way down. I've always found keyboard macros like that really untidy and inconsistent. If I'm writing stuff that requires a load of the same things I generally put it in a procedure or I do something like XxxX then do a case sensitive search/replace when I'm finished to replace all those instances with what I want to insert.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • hinvhinv Posts: 1,255
    edited 2009-05-23 05:06
    Hi Brad,

    In the file-> save as dialog box. When typing a file name it is common to "save" when I press enter after typing in the filename. Could you bind "Enter" where the filename goes to do this??

    Thanks,
    Doug
  • hinvhinv Posts: 1,255
    edited 2009-05-24 03:02
    Hi Brad,

    Another feature that would be nice would be a context sensitive file menu in the tab area. The menu should be for the tab that was under the mouse when you left click instead of the current tab no matter where you are in the tabs. You can look at firefox for an example.

    Thanks,
    Doug
  • BradCBradC Posts: 2,601
    edited 2009-05-24 06:58
    hinv said...
    Hi Brad,

    Another feature that would be nice would be a context sensitive file menu in the tab area. The menu should be for the tab that was under the mouse when you left click instead of the current tab no matter where you are in the tabs. You can look at firefox for an example.

    Thanks,
    Doug

    I tried it that way initially but the tab component I'm using reports its context inconsistently on OSX and has "issues" under GTK2. Can't do it until I can work the bugs out of the widgetset.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • BradCBradC Posts: 2,601
    edited 2009-05-24 14:34
    hinv said...
    Hi Brad,

    In the file-> save as dialog box. When typing a file name it is common to "save" when I press enter after typing in the filename. Could you bind "Enter" where the filename goes to do this??

    I just tested this here with Linux/GTK2 and it saves as soon as I press enter (Enter is bound to the SAVE button). What OS are you running?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • hinvhinv Posts: 1,255
    edited 2009-05-24 15:32
    I am using openSuSE 10.3
  • OwenSOwenS Posts: 173
    edited 2009-05-31 10:57
    BSTC mangles Unicode characters when compiling. While they may have no use on the TV display and such, since they're not in the font, in some cases you may wish to e.g. send them over a serial connection.

    Testcase:
    ' Spin Compiler Stress Test - Unicode characters
    
    CON
      _XINFREQ = 6_000_000
      _CLKMODE = XTAL1 + PLL16X
    
    OBJ
      tty: "FullDuplexSerial"
    
    PUB start
      tty.start(31, 30, 0, 9600)
      repeat
        tty.str(string("Unicode Test", 13))
        tty.str(string("#  Title              Test String", 13))
        tty.str(string("== ================== ===========", 13))
        tty.str(string("1. Hiragana           &#12362;&#12399;&#12424;&#12358;",    13))
        tty.str(string("2. Katakana           &#12467;&#12531;&#12452;&#12481;&#12527;",  13))
        tty.str(string("3. Double-Width Latin &#65320;&#65349;&#65356;&#65356;&#65359;",  13))
        waitcnt(clkfreq + cnt)
    
    



    Windows users will likely need to install east-asian language support in order to display the characters correctly. Users on Linux will probably already have DejaVu Sans or another multi language font installed. Mac OS X users, I don't know. If you have the fonts installed and still can't see them properly, you may be suffering from Mojibake which means your browser is ignoring the forum's instruction to treat it's text as UTF-8.

    (In case you're wondering, the Japanese testcases are "ohayou" and "konnichiwa" respectively)

    [noparse][[/noparse]Another bug which I can't figure out at all: On Linux AMD64, the IDE dies within about 30 seconds no matter what I do.. and I tried it on both Kubuntu 9.04 and SuSE 11.1]

    Post Edited (OwenS) : 5/31/2009 11:03:49 AM GMT
  • BradCBradC Posts: 2,601
    edited 2009-05-31 11:35
    OwenS said...
    BSTC mangles Unicode characters when compiling. While they may have no use on the TV display and such, since they're not in the font, in some cases you may wish to e.g. send them over a serial connection.

    [noparse][[/noparse]Another bug which I can't figure out at all: On Linux AMD64, the IDE dies within about 30 seconds no matter what I do.. and I tried it on both Kubuntu 9.04 and SuSE 11.1]

    Interesting. How does the Propeller Tool handle the embedded unicode? I deliberately "mangle" the unicode as it hits the compiler to remain completely compatible with the Parallax compiler, but I'm not sure how its going to work for other characters. On Windows the Unicode is encoded in UTF16, while Linux and Mac use UTF8, so what would actually get passed to the compiler? Could you compile your binary on windows with the Parallax tools and send it to me so I can try and figure out how I might make it work?

    The crash bug is an interesting one. My only X86-64 machine is a server and has no X libs installed, so testing for me is a bit awkward. Does it dump a backtrace in the terminal?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • OwenSOwenS Posts: 173
    edited 2009-05-31 14:16
    Aah, this one's been fun to test - my Windows box doesn't have East Asian support installed.

    Anyway, I just compiled the code with the Propeller Tool. It also mangled the text - in a different way, I think, though this may just be because both systems use different 8-bit character sets.

    Perhaps we need some way to differentiate Unicode and non Unicode text? Perhaps borrow from Python here - Unicode text is defined as u"Blah"? In the case of Unicode text, it would be output as UTF-8, UTF-16 or UCS-4 depending upon whether it's being stored in bytes/words/longs?

    (And the IDE is being most uncooperative about crashing itself at the minute)
  • BradCBradC Posts: 2,601
    edited 2009-05-31 15:21
    OwenS said...
    Aah, this one's been fun to test - my Windows box doesn't have East Asian support installed.

    Anyway, I just compiled the code with the Propeller Tool. It also mangled the text - in a different way, I think, though this may just be because both systems use different 8-bit character sets.

    Perhaps we need some way to differentiate Unicode and non Unicode text? Perhaps borrow from Python here - Unicode text is defined as u"Blah"? In the case of Unicode text, it would be output as UTF-8, UTF-16 or UCS-4 depending upon whether it's being stored in bytes/words/longs?

    I thought of a dirty way around it, which is to encode your strings in a file in UTF-8 and import that into the DAT section. You could also manually encode your UTF-8 characters into a string() construct, just avoid extraneous 0's.

    The thought of supporting multiple unicode encodings just makes me want to scream. It's bad enough I have to deal with UTF16 without adding UCS-4 into the mix.

    I'll have to give it some substantial thought. How it works now is all files that contain unicode are converted to UTF16 prior to being passed to the compiler. The compiler then translates the Parallax specific unicode characters into the standard Parallax 0-$FF character table and passes it to the compiler. Its done this way to retain compatibility with the Windows UTF16 files the Parallax gear uses. That poses a problem for any multi-byte encodings as the surrogate characters are all mapped to other glyphs already.

    There is no easy out for the compiler to process characters wider than 8 bits without a complete re-write of the tokeniser. It's just not geared to cope with multi-byte characters.
    The reason bst[noparse][[/noparse]c] mangles the characters differently to the Parallax compiler is : that was an option I never considered when writing the compiler, so I never tested theirs to see what it would do if you fed it something outside the standard character set.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • W9GFOW9GFO Posts: 4,010
    edited 2009-06-02 02:55
    OSX 10.5.7 intel
    BST .18-pre8

    The 'Compile Information' window is not showing the proper relationship between objects.

    The Servo32_Ramp_V1.spin object should be a child of Servo32v5. There is absolutely no reference to it in the LCD object, it is a child of the Servo32v5 object.

    Also, why does it add the number "2" after some of them? None of the object names end in a "2".

    Rich H

    attachment.php?attachmentid=61285

    Post Edited (W9GFO) : 6/2/2009 3:03:19 AM GMT
    353 x 425 - 60K
  • BradCBradC Posts: 2,601
    edited 2009-06-02 05:11
    W9GFO said...
    OSX 10.5.7 intel
    BST .18-pre8

    The 'Compile Information' window is not showing the proper relationship between objects.

    The Servo32_Ramp_V1.spin object should be a child of Servo32v5. There is absolutely no reference to it in the LCD object, it is a child of the Servo32v5 object.

    Also, why does it add the number "2" after some of them? None of the object names end in a "2".

    Rich H

    attachment.php?attachmentid=61285

    Oh YAY! I have been chasing that bug for _ages_ and have not been able to reproduce it. Could you archive that set of files for me and email it to me at proptools at fnarfbargle.com ?

    I spent a while yesterday trolling through the forum posts looking for a trigger or way to reliably reproduce it.

    I've squashed about 30 niggling little bugs this week, and its feeling a lot more solid, so there is likely to be a new revision in the not to distant future.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
  • W9GFOW9GFO Posts: 4,010
    edited 2009-06-08 02:10
    Feature request! Lets see if I can describe this well enough.

    I think it would be really super useful if there could be an option to split the editor window right down the middle so that anything typed on the left side did not affect the right, as in commenting. You would have the option of setting just which column the split took place. If you type on the right, it wouldn't affect the left side.

    I'd like to better document my code, or even document before writing it. As it stands now I spend a lot of time moving the comments around either because I hit "enter" instead of moving the cursor to the next line or I make changes in the code which moves the comments around. There's also the issue of lining up the comments nicely - that would be easy in a split window. It would be so nice if I could just type away to my heart's content without messing up the formatting.

    Rich H
  • BradCBradC Posts: 2,601
    edited 2009-06-08 03:38
    W9GFO said...

    I think it would be really super useful if there could be an option to split the editor window right down the middle so that anything typed on the left side did not affect the right, as in commenting. You would have the option of setting just which column the split took place. If you type on the right, it wouldn't affect the left side.

    Yeah, that would be neat. Unfortunately, the amount of work required to keep both editor buffers synchronised like that, or have multiple editors share the same buffer is quite extensive with the infrastructure I'm using. You can read that as "it's on my list, but it's a metric ton of work and a long way down". If you can get access to a windows box, have a look at the "Align" mode in the Propeller Tool and let me know if it does more what you are looking for. I've never used it, but I believe its supposed to be quite good for that sort of thing and if I was poked hard enough I could slide it up the todo list a bit.

    Likewise the tear off tabs, split screens and floating editor windows like Propeller Tool can do. They are all on my radar, but as I never use them myself combined with that being a whole lot of extra work and re-structuring, they are quite likely a long way off.

    At the moment I'm tackling one last niggling error that has come in when I've introduced a way to refresh the directory window, and it's a bear to nail as it requires an Intel Mac with a lot of network mapped directories on a slow network to reproduce. It's the little things that take the most time and effort unfortunately.

    On the upside, I've got all the Mac keys properly mapped and working now and I've been spending a lot more time on the Mac to ensure things are as polished as I can make them.

    @Ale, that odd bug you found with the serial terminal locking up the compiler and not being able to quit was not related to the serial terminal at all. It was a race in the way I implemented threading that would only manifest itself on the Mac.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "VOOM"?!? Mate, this bird wouldn't "voom" if you put four million volts through it! 'E's bleedin' demised!
Sign In or Register to comment.