Re: [Freedos-devel] Tree v3.7.3 Was: Re: Xcopy 1.5

2024-02-18 Thread Jerome Shidel via Freedos-devel
Hi Jeremy, 

> On Feb 17, 2024, at 9:58 PM, perditionc--- via Freedos-devel 
>  wrote:
> 
> On Sat, Feb 17, 2024 at 9:09 AM Wilhelm Spiegl via Freedos-devel
>  wrote:
>> 
>> Hi Jim,
>> thanks for your response. Xcopy is not the only thing, maybe you remember 
>> tree 3.72 in this forum?
>> 
> ...
>> 
>> tree: do not know the actual situation, I got a promise that someone will 
>> check whats the matter.
>> 
> ...
> 
> Good evening,
> 
> Please see http://server.fdos.org/fdos/tree/TREE373.zip for an update
> to the tree program that should fix the issues (sorry for the long
> delay).
> 
> Changes from tree v3.7.1/3.7.2
>  - reverted argument processing back to code from v3.7 [.0], the
> changes to use getopt were really broken
>  - minor tweaks to included cats to default NLSPATH to ".;.\NLS", a
> future version will update tree to kitten
>  - added extra check so if no serial# returned, assumes call failed
> and doesn't print invalid serial# (note, the way it calls int21h is a
> bit funky, in pdTree the carry flag is forced set before the call as I
> recall some kernels don't set it on error for the called function, but
> this version just clears struct and assumes if still clear then call
> failed as I'm not sure how to set the flags register prior to the
> int86 call (it's not the standard version, no regs struct passed) and
> don't have all the compiler variants easily available anymore to test
> changes to it.
>  - included in the zip is all the current? message catalogs, note:
> these are straight from the FreeDOS nls archive on gitlab, hopefully
> didn't miss/mess any up.
>  - some documentation updates to reflect that I no longer have
> @computer.org email address.
> 
> A version 4 will be released eventually, but I'm not making any
> commitments for when.  In the meantime, if anyone has any issues with
> this version, please let me know and I'll try to push out an update a
> little quicker. :-)

Great news. :-)

I created an “unstable” branch in the FreeDOS GitLab Archive. That way the new 
version will be in the next Interim Test Build for every one to try out. 

I did make a few minor adjustments to the LSM file. Mostly for compatible with 
the current automated maintenance utilities. For example, most of those tools 
can properly deal with multiple line entries. However, I can be lazy at times 
and a few other less important tools ignore text that is not preceded by a 
field identifier. 

Also, I noticed that you did not included the help translations from FD-NLS in 
the zip. So, I just included them. Although, I forget if TREE displays those 
versions of help when using a different language. 

I’ll definitely have to give it a try.

> 
> (ps a new test kernel release is pending, but I'm trying to figure
> out/fix a driver hang during load issue first).

Looking forward to that as well. :-)

> 
> Thanks,
> Jeremy

Thanks,

Jerome

___
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel


[Freedos-devel] Tree v3.7.3 Was: Re: Xcopy 1.5

2024-02-17 Thread perditionc--- via Freedos-devel
On Sat, Feb 17, 2024 at 9:09 AM Wilhelm Spiegl via Freedos-devel
 wrote:
>
> Hi Jim,
> thanks for your response. Xcopy is not the only thing, maybe you remember 
> tree 3.72 in this forum?
>
...
>
> tree: do not know the actual situation, I got a promise that someone will 
> check whats the matter.
>
...

Good evening,

Please see http://server.fdos.org/fdos/tree/TREE373.zip for an update
to the tree program that should fix the issues (sorry for the long
delay).

Changes from tree v3.7.1/3.7.2
  - reverted argument processing back to code from v3.7 [.0], the
changes to use getopt were really broken
  - minor tweaks to included cats to default NLSPATH to ".;.\NLS", a
future version will update tree to kitten
  - added extra check so if no serial# returned, assumes call failed
and doesn't print invalid serial# (note, the way it calls int21h is a
bit funky, in pdTree the carry flag is forced set before the call as I
recall some kernels don't set it on error for the called function, but
this version just clears struct and assumes if still clear then call
failed as I'm not sure how to set the flags register prior to the
int86 call (it's not the standard version, no regs struct passed) and
don't have all the compiler variants easily available anymore to test
changes to it.
  - included in the zip is all the current? message catalogs, note:
these are straight from the FreeDOS nls archive on gitlab, hopefully
didn't miss/mess any up.
  - some documentation updates to reflect that I no longer have
@computer.org email address.

A version 4 will be released eventually, but I'm not making any
commitments for when.  In the meantime, if anyone has any issues with
this version, please let me know and I'll try to push out an update a
little quicker. :-)


(ps a new test kernel release is pending, but I'm trying to figure
out/fix a driver hang during load issue first).

Thanks,
Jeremy


___
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel