AMOS Pro V2.10 alpha release

[Umpal]

Sounds like a compiler bug, not a problem with AmosPro itself.

You're right. Any chance of fixing it? I figured out some workaround but to be honest it requires a lot of work in my couple of thousand lines of code and after all it'd be code over code - something that should work without such a contrive.

By the way. I need someone to enlighten me. Is there an exe AMOSPro v 2.10 or just those fixed libraries available in download section that require only swap with original 2.0? I already did that but was wondering if there is a fixed main exe file.

I didn't say that but: Great work guys, whoever is/was involved in fixes. I hope the work is continued. Thanks from my part.

[bruceuncle]

Yep, it's a bug in the compiler.  It's related to a bug in the ValRout() function in the AMOS core library.  That function evaluates a numeric string and returns an integer, single or double.  It returns a zero and a fail flag if anything goes wrong (overflow or 'bad' number) or if the string is not a number. 

The function is used for two tasks:

1: By the tokeniser in the editor to convert what's been typed in into a number token.

2: By the AMOS Basic Val() function.

The ValRout() function has minor bugs in a branch to the wrong label and inconsistent return values.  Unfortunately, the function and parts of it were copied and pasted (complete with bug!) to other parts of the sources, including the compiler.  Tracking down what calls the function was not straightforward as at least one reference is via a JSR(An).

Anyway, the function also takes a param in a data register (typing on my phone, so sorry about lack of specifics) which indicates whether to include the sign or not.  Reason being that the tokeniser will already have parsed and created a token for any sign, so all it wants is the digits.  Whereas the Val() function will need the full signed result.

So it looks like you've found a new bug where the compiler isn't setting the 'sign required' param.

V2.10 is still being worked on.  I had a major, major back operation last year which I'm still recovering from.  So it's been a bit slow  going...

[Umpal]

Bruceuncle thanks a lot for you reply. It's very complex and explains to the root why the compiler has driven me crazy. I'm glad I didn't lost my mind 

I'm sorry to hear about your back issue. I can say I totally understand what it means as I'm experimenting similar thing (but I'm still looking for a non surgical solution). I wish you very fast and efficient recovery. And of course I hope for a fix for Val() bug.

Cheers,
Peter

PS Have you seen my feature suggestion?  I understand it's not easy to apply but it'd be fantastic to have such thing in AMOS Editor

[yurif74]

V2.10 is still being worked on.  I had a major, major back operation last year which I'm still recovering from.  So it's been a bit slow  going...

Hello Bruceuncle,
I would wish you a very fast recovery too
thanks for your great work, hope to see an update in the future 
Juri

[morpa]

This 2.10 sounds exciting. Anyone heard from bruceuncle in a while?

[SamuraiCrow]

We haven't heard from bruceuncle in months.

[morpa]

I hope he is all well and fine, considering his last posts.

In the facebook group for AMOS F. Lionet has posted about wished fixes in AMOS, and this would be a perfect timing for bruceuncle join the discussion there, considering his hard work with AMOS 2.10

[Volvo_0ne]

I hope he is all well and fine, considering his last posts.

In the facebook group for AMOS F. Lionet has posted about wished fixes in AMOS, and this would be a perfect timing for bruceuncle join the discussion there, considering his hard work with AMOS 2.10

SYNCHRO [n] or [n TO n]

would be handy for only running amal channels when required

[DarkN3mo]

what a pity that this dear AMOS PRO is no longer updated (I'm talking about the real version, not AOZ)

[DarkN3mo]

Is AMOS 2.10 still relevant?

[FOL]

Is AMOS 2.10 still relevant?

Sorry, there appears to be no one visiting any more.
I have plans to reinvent site and start to get it back to its millions of views
.

[DarkN3mo]

yes, this site seems to be dead... it's a shame.

[FOL]

yes, this site seems to be dead... it's a shame.

Blame Facebook. I believe they all use that instead. I hate Facebook.

[DarkN3mo]

yes, it's a shame, a website is much more practical than facebook.