1. Phix 0.7.8 released

Usual place: http://phix.x10.mx/download.php

Version 0.7.8 
============= 
10/01/2018: BUGFIX: when the file/directory is not found, apply get_proper_path()  
            recursively to the parent directory. 
21/01/2018: BUGFIX: rand() effectively preserved the high bit for ranges over 
            #7FFFFFFF. Changed a jge (signed jump) to jae (unsigned).  
            Also changed the 64-bit store to deal with 64-bit results above 
            #7FFF_FFFF_FFFF_FFFF correctly. 
03/02/2018: Corrected definition of xGetExitCodeThread. 
15/02/2018: New LiteZip wrapper and documentation added. 
15/02/2018: BUGFIX: Slice replacement was out-by-one, eg 
                s = { 1,2,3,4,5,6 }  
                s[1..1] = 9 
            was incorrectly {9,1,2,3,4,5}, now yields {9,2,3,4,5,6}. 
            Thanks to Tom for finding this.  
17/02/2018: Bugfix: parse_json() did not cope with negative numbers. 
            Thanks to Chris for finding this.  
17/02/2018: Bugfix: setup.ew crashed for registry keys of length 0. 
            Thanks to Kat for finding this.  
20/02/2018: Removed spurious sanity check in change_timezone(). 
            Thanks to Chris for finding this.  
24/02/2018: =$ can now be used anywhere except the first in an enum. 
            Previously =$ only worked for delta==+1, but DoEnum()  
            now has a prev var, so that any "by delta" now works. 
28/02/2018: Added peek_wstring() and poke_wstring(). 
04/03/2018: adjust_timedate() no longer clobbers DT_MSEC aka DT_DOW. 
05/03/2018: timedate.e: added "ms" to extract/print milliseconds. 
            (not thoroughly tested, possible ambiguity issue noticed) 
06/03/2018: constant integer {a,b} = <expr> style syntax now supported. 
06/03/2018: date(DT_GMT) now returns the GMT (==UTC) time, irrespective  
            of location, with milliseconds.  
14/03/2018: Fixed potential thread safety issue in sprint(). 
14/03/2018: Upgraded to IUP 3.24. Three routines have been removed: 
            IupColorbar, IupColorBrowser, and IupDial. 

Enjoy!

new topic     » topic index » view message » categorize

2. Re: Phix 0.7.8 released

thank you

new topic     » goto parent     » topic index » view message » categorize

3. Re: Phix 0.7.8 released

could it be that matchrepl.e is missing in the setup?

new topic     » goto parent     » topic index » view message » categorize

4. Re: Phix 0.7.8 released

Hey, Pete,

The first link for a Windows download still shows phix.0.7.7.setup.exe, but it does load the 0.7.8 installer.

new topic     » goto parent     » topic index » view message » categorize

5. Re: Phix 0.7.8 released

seems like findrepl.e is missing too.

new topic     » goto parent     » topic index » view message » categorize

6. Re: Phix 0.7.8 released

euphoric said...

Hey, Pete,

The first link for a Windows download still shows phix.0.7.7.setup.exe, but it does load the 0.7.8 installer.

Oops. I was wrong. Apparently it installed 0.7.7.

Uh. Maybe it's just that the executable is reporting 0.7.7. Did you change the revision date in the source? grin

How would I know I'm running 0.7.8?

new topic     » goto parent     » topic index » view message » categorize

7. Re: Phix 0.7.8 released

begin said...

could it be that matchrepl.e is missing in the setup?
seems like findrepl.e is missing too.

They sure are. Added to the build script, thanks. Trust you're ok to grab them from the repository?

euphoric said...
euphoric said...

Hey, Pete,

The first link for a Windows download still shows phix.0.7.7.setup.exe, but it does load the 0.7.8 installer.

Oops. I was wrong. Apparently it installed 0.7.7.

Uh. Maybe it's just that the executable is reporting 0.7.7. Did you change the revision date in the source? grin

How would I know I'm running 0.7.8?

My bad, I somehow forgot to run p -cp! If you do that you should be fine.
(Amazingly, I never once thought of adding that to the build script, it's certainly there now!)
That typo on the download page also fixed, thanks.

Pete

new topic     » goto parent     » topic index » view message » categorize

8. Re: Phix 0.7.8 released

petelomax said...

My bad, I somehow forgot to run p -cp! If you do that you should be fine.
(Amazingly, I never once thought of adding that to the build script, it's certainly there now!)
That typo on the download page also fixed, thanks.

OK, got 0.7.8 installed.

Microsoft Windows Defender is NOT liking the pw.exe file. Keeps trying to quarantine it. getlost

new topic     » goto parent     » topic index » view message » categorize

9. Re: Phix 0.7.8 released

running on eset smart security i have no problems

new topic     » goto parent     » topic index » view message » categorize

10. Re: Phix 0.7.8 released

I did have a Trojan:Win32/Fuerboos.A!cl last month (16/2, and of course that was a somewhat different file), but Windows Defender scan is clean with the latest updates (1.263.562.0).

new topic     » goto parent     » topic index » view message » categorize

11. Re: Phix 0.7.8 released

Downloaded the latest version today, and I get a crash when running the tests on Linux Mint 64-bit:

Testing[56]: /home/irv/phix/test/t60td.exw 
 
/home/irv/phix/test/t60td.exw:232 
attempt to divide by 0 
 

   s = adjust_timedate(s,timedelta(hours:=7.5))                         
   if s!={2000,1,5, 3, 0,0,4,5,0} then ?9/0 end if 

Phix hybrid interpreter/compiler.

Version 0.7.8 (64 bit Linux) Copyright Pete Lomax 2006..2016

new topic     » goto parent     » topic index » view message » categorize

12. Re: Phix 0.7.8 released

irv said...

Downloaded the latest version today, and I get a crash when running the tests on Linux Mint 64-bit:

Testing[56]: /home/irv/phix/test/t60td.exw 
 
/home/irv/phix/test/t60td.exw:232 
attempt to divide by 0 
 

   s = adjust_timedate(s,timedelta(hours:=7.5))                         
   if s!={2000,1,5, 3, 0,0,4,5,0} then ?9/0 end if 

Thanks for finding that. I can confirm it is not a significant issue, caused by "adjust_timedate() no longer clobbers DT_MSEC aka DT_DOW", and you can safely remove or replace that test as below.

I have changed it to:

--22/3/18 
--s = adjust_timedate(s,timedelta(hours:=7.5))                      if s!={2000,1,5, 3, 0,0,4,5,0} then ?9/0 end if 
s = adjust_timedate(s,timedelta(hours:=7.5))        s[DT_MSEC] = 0  if s!={2000,1,5, 3, 0,0,0,5,0} then ?9/0 end if 

and likewise replaced the 7th entry in the literal constants with zero in the tests all the way down to line 275 (was 4,4,5,5,5,6,0,7,3,4, ie the next 10 tests)

Aside: these routines do not really need to be quite so accurate as those tests - just so long as all the text/format_timedate() tests pass, which they do.

Pete

PS This happened on windows as well - I have been a bit lazy and not fired up a 64 bit linux vm to test it.

new topic     » goto parent     » topic index » view message » categorize

Search



Quick Links

User menu

Not signed in.

Misc Menu