1. temporary eubins

I noticed that jeremy's eubins appear to have temporarily stopped working. (After 1121 they are merely just empty tarballs.)

So, I have put up linux eubins for revision 1136 at http://malcom.unkmar.com/1136.tar.gz to help anyone (including jeremy) get rebootstrapped.

I am working on w98 at the moment and I post the eubins when I have them ready.

new topic     » topic index » view message » categorize

2. Re: temporary eubins

w98 eubins are available at http://malcom.unkmar.com/1136.zip

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

3. Re: temporary eubins

jimcbrown said...

w98 eubins are available at http://malcom.unkmar.com/1136.zip

Thank you. I can update my CHM help file to this version.

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

4. Re: temporary eubins

Eubins are functioning again, sorry about the down time...

http://jeremy.cowgar.com/eubins/

Jeremy

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

5. Re: temporary eubins

Jeremy, thanks a lot for all the work with timings of Euphoria components, I saw this for the first time today. I don't want to appear ungrateful, but a teensy request please? Most of the timings have a value of zero cos Euphoria is so fast ;) Can the timing scale be changed to something meaningful?

Regards Alan

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

6. Re: temporary eubins

Hi Jeremey, I tried a ec -wat using your eu4.0 v1159 against my source, it created a emake.bat ok. Source name or includes longer than 6.3 format do not have enough remaining places to append the "_1.C", and the linking then fails. I assume this is a known issue..? No search is available in this forum yet so can you say what the workaround is please? The win32lib includes will be harder to rename than my source!

Regards Alan

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

7. Re: temporary eubins

AlanO said...

Hi Jeremey, I tried a ec -wat

Try ecw rather than ec

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

8. Re: temporary eubins

Thanks Derek, the ecw -wat seems to support 8+.3 filenames fine. But now I am getting init-.c(4404): Error! E1175: Too many initializers Although the bound or translated version of my app (under eu v2.5) works fine. I'll see if my Eu v2.5 translator does the same thing, if not then rtfm Watcom manuals!

Cheers Alan

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

9. Re: temporary eubins

AlanO said...

But now I am getting init-.c(4404): Error! E1175: Too many initializers

That message usually means that the euphoria.h file in your Euphoria include directory is not the correct one for the translator. You might be pointing to an old include directory. I found that I had this message when the euphoria.h had the wonrg definition for this struct...

struct routine_list { 
	char *name; 
	int (*addr)(); 
	int seq_num; 
	int file_num; 
	short int num_args; 
	short int convention; 
	char scope; 
}; 

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

10. Re: temporary eubins

AlanO said...

Jeremy, thanks a lot for all the work with timings of Euphoria components, I saw this for the first time today. I don't want to appear ungrateful, but a teensy request please? Most of the timings have a value of zero cos Euphoria is so fast ;) Can the timing scale be changed to something meaningful?

Regards Alan

Basically you cannot, because this is provided by the OS. And a resolution of 1ms is definitely too coarse.

However, there are foure ways out:

  1. Count how many times a task executes in a iven time. Tasks may be a good tool for this;
  2. Compute how much time many, many executions of the task take;
  3. Under Windows, there is a high resolution timer interface, and you can for instance <plug> use the hirestime.e library </plug>.
  4. Directly ask the CPU its interal clock tick counter value. I think that's what T. škod's CPU_time library does.

The latter twwo approaches will fail on non Pentium machines. I don't know about an analogue of 3/ for Unix systems. Under DOS, using 4/ would be easier than 3/ anyway.

CChris

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

Search



Quick Links

User menu

Not signed in.

Misc Menu