Main
Date: 03 Nov 2007 14:18:18
From: Pascal
Subject: Scid 3.6.19 released (chess database and training software)
Hi,

I released Scid 3.6.19 at
http://prolinux.free.fr/scid/

(Source files for Linux and binary for Windows).

Note that all Scid Pocket code corresponding to version 0.25 is now up
to date and can be found in the pocket directory along with engines
ported to Pocket PC.

For those curious, here is the changelog:

- clean up of configuration files (thanks to cin Kasperski)
- books' path is now configurable through ./configure and Makefile.conf
(SHAREDIR) (thanks to cin Kasperski)
- some code added to support Novag's chess board. This code is not
finished (and will not be finished until I put a hand on a Novag
Citrine), so the menu entries are hidden to the user until further notice.
- Added a new menu entry : "open base as Tree window" (a kind of
shortcut that avoids switching between bases)
- Refresh of Tree window can be toggled
- A Tree window can be opened for each base through the command tools- >
open base as tree. So now the meaning of "locked" for a tree window has
changed. This used to mean display information of *this* base in the
Tree window, but now it means, if the Tree window is locked, that
closing the Tree implies closing the associated base. This seems natural
because using "open base as tree" means the user wants to get a kind of
"book/tree" and not really a base. So it makes sense to link base's fate
to tree's one. Moreover, as a Tree window, a Graph window and a Best
Games window, etc. can be opened for each base, more than 30 windows can
be opened related to Trees, so the closing of a Tree window has to clean
up things.
When several Tree windows are opened the top most is refreshed first,
then others.
- Memory leaks fixed (particularly in Game class)
- New API : sc_tree free to release memory used by Tree windows (sc_tree
search). This sould only be used for systems with low memory constraints.
- New API : sc_game truncatefree : when truncating a game, moves are
freed. This sould only be used for systems with low memory constraints.
- New API : sc_info logmem : useful with checkmemlog.tcl script to
detect memory leaks.
- New API : sc_var deletefree : when deleting a variation, moves are
freed. This sould only be used for systems with low memory constraints.
- modified scmerge so it can take an unlimited number of input files
(each separated by a carriage return) (thanks to cin Kasperski).
Example usage :
find . -name '*.si3'




 
Date: 04 Nov 2007 14:08:16
From: Andreas Schwarz
Subject: Re: Scid 3.6.19 released (chess database and training software)
On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:

> I released Scid 3.6.19 at
> http://prolinux.free.fr/scid/


This is confusing, there is no new scid version at

http://scid.sourceforge.net




--

PGP: 0x661AB571



  
Date: 04 Nov 2007 18:23:24
From: Pascal
Subject: Re: Scid 3.6.19 released (chess database and training software)
Andreas Schwarz a �crit :
> On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:
>
>> I released Scid 3.6.19 at
>> http://prolinux.free.fr/scid/
>
>
> This is confusing, there is no new scid version at
>
> http://scid.sourceforge.net
>

Yes, (nearly) everybody agrees with it. Scid has not been updated at
Sourceforge since 2004. This is sad, but pleases some people.

Pascal


   
Date: 05 Nov 2007 21:05:46
From: Andreas Schwarz
Subject: Re: Scid 3.6.19 released (chess database and training software)
On Sun, 04 Nov 2007 18:23:24 +0100, Pascal wrote:

> Andreas Schwarz a �crit :
> > On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:
> >
> >> I released Scid 3.6.19 at
> >> http://prolinux.free.fr/scid/
> >
> >
> > This is confusing, there is no new scid version at
> >
> > http://scid.sourceforge.net
> >
>
> Yes, (nearly) everybody agrees with it. Scid has not been updated at
> Sourceforge since 2004. This is sad, but pleases some people.

Ok, but if you use the same name, it's not possible to include your
version into the package managment systems (for example, the freebsd
ports structure). Your version collide with the original scid, this
causes some problems.


--

PGP: 0x661AB571



    
Date: 07 Nov 2007 05:43:03
From: Dave
Subject: Re: Scid 3.6.19 released (chess database and training software)
Andreas Schwarz wrote:

>>> This is confusing, there is no new scid version at
>>>
>>> http://scid.sourceforge.net

Yes, agreed. It is very confusing.

>>>
>> Yes, (nearly) everybody agrees with it. Scid has not been updated at
>> Sourceforge since 2004. This is sad, but pleases some people.

> Ok, but if you use the same name, it's not possible to include your
> version into the package managment systems (for example, the freebsd
> ports structure). Your version collide with the original scid, this
> causes some problems.
>
>
I'm somewhat biased in this, as I have a long running disagreement with
Pascal, but leaving that aside for the minute, this is my take on this.

To allow the real Scid from Shane to be in a package management system,
as well as a fork of it, several things should be done.

1) The fork needs renaming.

IMHO, Pascal#s original name scid-pg was fine. Scid2 is perhaps a bit
confusing, as someone said. Scidpro suggests some sort of superiority to
what Shane has produced. I would not argue he has more features than
Shane's code, but if Shane wants to come back (and if Pascal is telling
the truth, Shane has told him that less than 6 months ago), then it
would be unfair to Shane to have someone else with a 'pro' version.

Personally I think my own decision to rename to chessdb

http://chessdb.sourceforge.net/

with adequate notices that it is based on Scid, was the right decision.

But lets not argue again.

2) Some of the binaries, and executable shell scripts should be renamed.
Currently installing the real scid, and Pascals' on the same system is
not too easy, as there will be name conflicts.

A partial list of the programs that need renaming are:

scid scmerge scid.exe tkscid scid.gui scidlet
sc_addmove sc_epgn sc_remote sc_tree
sc_eco sc_import sc_spell scidpgn
spliteco

(There might be one or two missing from that list).

3) The real scid looks for the name 'scid' in the binary database files.
To maintain backward compatibility, that needs to be kept, so a global
renaming of scid to scidxyx/scid-pg/scidfoobaz is not desirable.


4) There are some sound files (one of which he lifted from chessdb!)
which are used. As long as they are stored in a directory specific to
the original scid or a fork, there is no need to rename them.

5) The files ratings.ssp, ratings.ssp and scid.eco don't need renaming,
as long as they are stored in their own directory.

If ALL the above are done (plus any I have overlooked) it will allow:

1) Shane's original Scid
2) Pascal's fork
3) My own fork (chessdb), http://chessdb.sourceforge.net/

to ALL be in the same package management system with no name conflicts.
It will allow a user to install ALL versions, safe in the knowledge that

a) When installing the original, or any fork, no files will be
overwritten. That will mean zero confusing warnings from package
management systems.

b) The programs will not conflict if two or more are used.

c) The programs can all share the database files.

To me, that has got to be sensible.

The current release of chessdb has some name conflicts with the real
scid and Pascal's fork. I intend totally removing those conflicts. My
original reason for renaming binaries was to avoid conflicts, but I did
not take that far enough. But I intend using 'ch' rather than 'sc' in
the binaries/executable scripts that need it.

Has anyone thought why the GNU C compiler is called gcc? I guess it was
to allow the original C compilers on UNIX systems (all called cc) to
coexist happily with the GNU compiler. Same with the vim, the editor
based on vi.

IMHO, if you create a fork, you should take care to avoid conflicts like
this.




     
Date: 07 Nov 2007 10:39:12
From: Pascal
Subject: Re: Scid 3.6.19 released (chess database and training software)
Dave a �crit :
> Andreas Schwarz wrote:
>
>>>> This is confusing, there is no new scid version at
>>>> http://scid.sourceforge.net
>
> Yes, agreed. It is very confusing.

It is confusing or you don't like people working on Scid ?

> Personally I think my own decision to rename to chessdb
> with adequate notices that it is based on Scid, was the right decision.

Renaming Scid to chessDB 1600+ times, putting your "(c) D Kirkby" where
you put no code is the right decision ?!
I personaly will never do such thing. And you're talking about confusion !

> But lets not argue again.

Yes. You're right :

> 4) There are some sound files (one of which he lifted from chessdb!)
> which are used.
Sorry, I was not aware *you* produced a sound file one day. Can you tell
me which one in order to give credit to you ?
You also kept claiming I took some code *you* wrote : can you tell me
the number of lines of *original* code *you* wrote, so I can acknowledge
you ?
No you can't. Why ?
Hum, yes. This is confusing.


    
Date: 06 Nov 2007 12:26:02
From: Pascal
Subject: Re: Scid 3.6.19 released (chess database and training software)
Andreas Schwarz a �crit :
> On Sun, 04 Nov 2007 18:23:24 +0100, Pascal wrote:
>
>> Andreas Schwarz a �crit :
>>> On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:
>>>
>>>> I released Scid 3.6.19 at
>>>> http://prolinux.free.fr/scid/
>>>
>>> This is confusing, there is no new scid version at
>>>
>>> http://scid.sourceforge.net
>>>
>> Yes, (nearly) everybody agrees with it. Scid has not been updated at
>> Sourceforge since 2004. This is sad, but pleases some people.
>
> Ok, but if you use the same name, it's not possible to include your
> version into the package managment systems (for example, the freebsd
> ports structure). Your version collide with the original scid, this
> causes some problems.

That is right. Maybe renaming to Scid2 would ease things ?

Anyway is there any interest in having anywhere the two versions at the
same time ? I do because I need to make some regression tests, but I
doubt any user would need both.

Pascal


     
Date: 06 Nov 2007 22:09:09
From: Andreas Schwarz
Subject: Re: Scid 3.6.19 released (chess database and training software)
On Tue, 06 Nov 2007 12:26:02 +0100, Pascal wrote:

> Andreas Schwarz a �crit :
> > On Sun, 04 Nov 2007 18:23:24 +0100, Pascal wrote:
> >
> >> Andreas Schwarz a �crit :
> >>> On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:
> >>>
> >>>> I released Scid 3.6.19 at
> >>>> http://prolinux.free.fr/scid/
> >>>
> >>> This is confusing, there is no new scid version at
> >>>
> >>> http://scid.sourceforge.net
> >>>
> >> Yes, (nearly) everybody agrees with it. Scid has not been updated at
> >> Sourceforge since 2004. This is sad, but pleases some people.
> >
> > Ok, but if you use the same name, it's not possible to include your
> > version into the package managment systems (for example, the freebsd
> > ports structure). Your version collide with the original scid, this
> > causes some problems.
>
> That is right. Maybe renaming to Scid2 would ease things ?

Maybe this will help, but it's also confusing (it's to close
to the original name).


> Anyway is there any interest in having anywhere the two versions at the
> same time ? I do because I need to make some regression tests, but I
> doubt any user would need both.

This is not the question. A package maintainer can not add your version
without removing the original scid.



--

PGP: 0x661AB571



      
Date: 07 Nov 2007 00:11:19
From: Pascal
Subject: Re: Scid 3.6.19 released (chess database and training software)
Andreas Schwarz a �crit :
> On Tue, 06 Nov 2007 12:26:02 +0100, Pascal wrote:
>
>> Andreas Schwarz a �crit :
>>> On Sun, 04 Nov 2007 18:23:24 +0100, Pascal wrote:
>>>
>>>> Andreas Schwarz a �crit :
>>>>> On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:
>>>>>
>>>>>> I released Scid 3.6.19 at
>>>>>> http://prolinux.free.fr/scid/
>>>>> This is confusing, there is no new scid version at
>>>>>
>>>>> http://scid.sourceforge.net
>>>>>
>>>> Yes, (nearly) everybody agrees with it. Scid has not been updated at
>>>> Sourceforge since 2004. This is sad, but pleases some people.
>>> Ok, but if you use the same name, it's not possible to include your
>>> version into the package managment systems (for example, the freebsd
>>> ports structure). Your version collide with the original scid, this
>>> causes some problems.
>> That is right. Maybe renaming to Scid2 would ease things ?
>
> Maybe this will help, but it's also confusing (it's to close
> to the original name).

It is hard to use a really different name without making silly things
(massive renaming of "Scid" tokens to XYZ in all files, source files,
help files, etc.), with stricly no added value. This would also add
confusion, if someone wants to continue development of original Scid one
day.

>> Anyway is there any interest in having anywhere the two versions at the
>> same time ? I do because I need to make some regression tests, but I
>> doubt any user would need both.
>
> This is not the question. A package maintainer can not add your version
> without removing the original scid.

So why not renaming it to Scid2 ? This would allow the use of both
binaries, but then there will be a problem with user's configuration
files.... One could think that if an engine is configured in Scid, it
should also appear in Scid2 (this is just one of the problem that arises).
That's not so easy. And from my own point of view (an user's point of
view) this is a work that is a bit useless. But if someone wants to take
the code of "Scid2" and tweak it to make it compatible with the original
Scid, he can of course do that.

Pascal


       
Date: 07 Nov 2007 00:37:59
From: Andreas Schwarz
Subject: Re: Scid 3.6.19 released (chess database and training software)
On Wed, 07 Nov 2007 00:11:19 +0100, Pascal wrote:

> Andreas Schwarz a �crit :
> > On Tue, 06 Nov 2007 12:26:02 +0100, Pascal wrote:
> >
> >> Andreas Schwarz a �crit :
> >>> On Sun, 04 Nov 2007 18:23:24 +0100, Pascal wrote:
> >>>
> >>>> Andreas Schwarz a �crit :
> >>>>> On Sat, 03 Nov 2007 14:18:18 +0100, Pascal wrote:
> >>>>>
> >>>>>> I released Scid 3.6.19 at
> >>>>>> http://prolinux.free.fr/scid/
> >>>>> This is confusing, there is no new scid version at
> >>>>>
> >>>>> http://scid.sourceforge.net
> >>>>>
> >>>> Yes, (nearly) everybody agrees with it. Scid has not been updated at
> >>>> Sourceforge since 2004. This is sad, but pleases some people.
> >>> Ok, but if you use the same name, it's not possible to include your
> >>> version into the package managment systems (for example, the freebsd
> >>> ports structure). Your version collide with the original scid, this
> >>> causes some problems.
> >> That is right. Maybe renaming to Scid2 would ease things ?
> >
> > Maybe this will help, but it's also confusing (it's to close
> > to the original name).
>
> It is hard to use a really different name without making silly things
> (massive renaming of "Scid" tokens to XYZ in all files, source files,
> help files, etc.), with stricly no added value. This would also add
> confusion, if someone wants to continue development of original Scid one
> day.

You can use names like "ScidPro", "ScidPlus", etc., this is more
uniq than Scid2.


> >> Anyway is there any interest in having anywhere the two versions at the
> >> same time ? I do because I need to make some regression tests, but I
> >> doubt any user would need both.
> >
> > This is not the question. A package maintainer can not add your version
> > without removing the original scid.
>
> So why not renaming it to Scid2 ? This would allow the use of both
> binaries, but then there will be a problem with user's configuration
> files.... One could think that if an engine is configured in Scid, it
> should also appear in Scid2 (this is just one of the problem that arises).

You can scan for a scid config dir and offer some kind of import
function
to the user. An own config dir is also necessary when you add new
features
which are incompatible to scid.


> That's not so easy. And from my own point of view (an user's point of
> view) this is a work that is a bit useless. But if someone wants to take
> the code of "Scid2" and tweak it to make it compatible with the original
> Scid, he can of course do that.

It's your choice.



--

PGP: 0x661AB571



     
Date: 06 Nov 2007 18:10:12
From: Morten Skarstad
Subject: Re: Scid 3.6.19 released (chess database and training software)
Pascal skrev:
>>>>> I released Scid 3.6.19 at
>>>>> http://prolinux.free.fr/scid/
>>>>
>>>> This is confusing, there is no new scid version at
>>>> http://scid.sourceforge.net
>>>>
>>> Yes, (nearly) everybody agrees with it. Scid has not been updated at
>>> Sourceforge since 2004. This is sad, but pleases some people.
>>
>> Ok, but if you use the same name, it's not possible to include your
>> version into the package managment systems (for example, the freebsd
>> ports structure). Your version collide with the original scid, this
>> causes some problems.
>
> Anyway is there any interest in having anywhere the two versions at the
> same time ? I do because I need to make some regression tests, but I
> doubt any user would need both.

What if Shane starts releasing updates to the original Scid again? Or
for that matter, anybody else picks up the code and says "Hey, I just
released Scid 4.0! Get it at
http://geocities.yahoo.com/therealscid"


      
Date: 06 Nov 2007 19:53:00
From: Pascal
Subject: Re: Scid 3.6.19 released (chess database and training software)
Morten Skarstad a �crit :
> Pascal skrev:
>>>>>> I released Scid 3.6.19 at
>>>>>> http://prolinux.free.fr/scid/
>>>>>
>>>>> This is confusing, there is no new scid version at
>>>>> http://scid.sourceforge.net
>>>>>
>>>> Yes, (nearly) everybody agrees with it. Scid has not been updated at
>>>> Sourceforge since 2004. This is sad, but pleases some people.
>>>
>>> Ok, but if you use the same name, it's not possible to include your
>>> version into the package managment systems (for example, the freebsd
>>> ports structure). Your version collide with the original scid, this
>>> causes some problems.
>>
>> Anyway is there any interest in having anywhere the two versions at
>> the same time ? I do because I need to make some regression tests, but
>> I doubt any user would need both.
>
> What if Shane starts releasing updates to the original Scid again?

This would be great news.

Or
> for that matter, anybody else picks up the code and says "Hey, I just
> released Scid 4.0! Get it at
> http://geocities.yahoo.com/therealscid"

Unfortunately there is nothing behind your link. Do you have some that
works ?


       
Date: 06 Nov 2007 19:03:31
From: Thomas T. Veldhouse
Subject: Re: Scid 3.6.19 released (chess database and training software)
Pascal <[email protected] > wrote:
>
> Unfortunately there is nothing behind your link. Do you have some that
> works ?

Uhm ... Pascal .... he made it up! It was for the purposes of illustrating
his point.

--
Thomas T. Veldhouse

When the ax entered the forest, the trees said, "The handle is one of us!"
-- Turkish proverb