tcl-core — Tcl Core Team discussions about Tcl development

Re: [TCLCORE] CFV warning: TIP #615: string is index

[Jan N. <jan...@gm...>]

Op ma 17 nov 2025 om 18:44 schreef Jan Nijtmans <jan...@gm...>:
> > % lsearch -index [list {}] {{a 1} {b 0} {c 2}} 0
> > Empty string cannot be compiled as index
>
> That's clearly a bug. I'll have a look at it. Thanks!

    $ tclsh9.1
    % lsearch -index [list {}] {{a 1} {b 0} {c 2}} 0
    index "" out of range

Bug fixed now. Same behavior as "-1" in 9.0.

This error-message "Empty string cannot be compiled as
index" was meant to make sure the byte-code engine to
take shortcuts means for index -1. It's never supposed
to be visible from scripts. Should be fixed now.

Byte-code improvements will be implemented later,
after this TIP is accepted. They won't change any
behavior, just speedup.

Hope this helps,
     Jan Nijtmans

Re: [TCLCORE] CFV warning: TIP #615: string is index

[Jan N. <jan...@gm...>]

Op ma 17 nov 2025 om 18:17 schreef Ashok:
> Further, it also leads to incompatibilities with Tcl 9.
>
> Tcl 9:
>
> % lindex {a b c} {}
> a b c
>
> TIP 615:
> % lindex {a b c} {}
> (empty string)


Hm. Did you actually try that, or did you derive it from the TIP?
I'm getting this with the TIP #615 branch:
    $ tclsh9.1
    % lindex {a b c} {}
    a b c
So there is no incompatibility at all.

Explanation: 'lindex' doesn't have an index as the first argument,
its argument is a list of indexes.   I'l clarify this in the TIP.

This also means, the remainder of your argumentation on
consistency doesn't make much sense to me.

> % lsearch -index [list {}] {{a 1} {b 0} {c 2}} 0
> Empty string cannot be compiled as index

That's clearly a bug. I'll have a look at it. Thanks!

       Jan Nijtmans

Re: [TCLCORE] CFV warning: TIP #615: string is index

[<apn...@ya...>]

TIP 615 is really two parts.

The first part is introduction of the [string is index] command. This seems
straightforward and may be useful for validation purposes. However, I would
like clarity what it will return for index forms that can be used with
commands like lindex. For example,

% string is index {0 0 0}
0

Given {0 0 0} is a valid index to lindex (and other commands) is this
intentional or a bug? If this is by design, it means the command cannot be
used to check if an index is valid to be passed to lindex, lsearch etc. If
this is a bug, and the command should return 1, then the command cannot be
used to check if an index is valid for lreplace and similar. This very much
limits the usefulness of the command and some clarity on the use case would
be beneficial.

The second part of the TIP proposes the use of empty strings as indices.
This gives me some heartburn from both conceptual and practical
perspectives.

Leaving aside my personal dislike of attaching semantics to the empty
string, special "tokens" should at least have the same semantics wherever
they are used. (As an aside, Tcl already violates this in its use of "end"
where it means the position of the last unit in some commands and the
position *after* the last unit in others.) The empty string semantics in TIP
615 is even worse in that semantics are dependent not only on the command
but on the context as well, in particular both the position of the index
argument AND the value of *other* arguments. This leads to behavior that is
inconsistent and non-intuitive.

Consider the commands

set s abcd
string range $s $index $index
string replace $s $index $index X

The natural expectation is that at most one character will be returned or
replaced. Instead, we have

% set index {}
% string range $s $index $index
abcd
% string replace $s $index $index X
X

The problem is that the empty string in the first position means something
different from the empty string in the second position. What exacerbates
this is the interpretation of the empty string in the second position
depends on the value in the first position. Further examples of this -

% string range $s 0 end
abcd
% string range $s -1 end
abcd
% string range $s 0 $index
a
% string range $s -1 $index
abcd

Note how the interpretation of "end" is the same in the first two. $index
(empty string) is on the other hand interpreted differently depending on the
value of a different argument being 0 or -1.

This is not only unintuitive, it also makes it harder to reason about
programs. Even the simplest of statements "string range $s $index $index" is
no longer simple to reason about in terms of effects.

Further, it also leads to incompatibilities with Tcl 9.

Tcl 9:

% lindex {a b c} {}
a b c

TIP 615:
% lindex {a b c} {}
(empty string)

Left unfixed, this means 9.1 is incompatible with 9.0 in this (and probably
other) cases. On the other hand, if you fix it, it means the lindex command
treats the empty string index differently from lrange et al which would be
horrible.

I also expect the syntax to be consistently usable in all index contexts.
Permitting it to be used for some indices and not others creates complexity
for users. Contrasting use of the empty string with the "end" index in TIP
615

% lsearch -index [list end] {{a 1} {b 0} {c 2}} 0
1
% lsearch -index [list {}] {{a 1} {b 0} {c 2}} 0
Empty string cannot be compiled as index

Perhaps the above is a fixable bug, not a conceptual failing. But otherwise,
consider the difficulty in using constructs like [list $index] as an
argument to lsearch where it is not known whether $index can be empty.

I have not looked at the effects on the byte code compiler if any. One thing
that would need to be looked at is the encoding of the empty string within
bytecode. As it currently stands, the encoding of indices already covers the
entire domain of from -INT_MAX:INT_MAX with negative numbers indicating
end-N index value range. There is no room for another index representing the
empty string. And there is certainly no current byte code mechanism to
encode based on the value of another argument. I haven't looked at the
implementation though because of the above objections. Maybe commands with
empty string indices are not being compiled.

Finally, with respect to benefits related to Tk compatibility, feels like
the tail wagging the dog. Missteps in index semantics in Tk (my *uninformed*
opinion because I do not understand the need there either) need not carry
back to Tcl.

/Ashok

Re: [TCLCORE] CFV: TIP 736

[Kevin W. <kw...@co...>]

TIP 736: YES

> On Nov 17, 2025, at 9:51 AM, Csaba Nemethi <csa...@t-...> wrote:
> 
> Attn. TCT members,
> 
> This is a CFV for TIP 736: Scroll entry with mouse wheel (formerly: Add mouse wheel and <TouchpadScroll> event bindings for entry widgets).
> 
>    https://core.tcl-lang.org/tips/doc/trunk/tip/736.md
> 
> The TIP is kindly sponsored by Marc Culler.  Many thanks, Marc!
> 
> Please send your votes to this list until next Monday, 2025-11-24 20:00 UTC.
> 
> Best regards,
> 
> Csaba
> 
> --
> Csaba Nemethi  https://www.nemethi.de  mailto:csa...@t-...
> 
> 
> 
> _______________________________________________
> Tcl-Core mailing list
> Tcl...@li...
> https://lists.sourceforge.net/lists/listinfo/tcl-core

Re: [TCLCORE] CFV: TIP 736

[Jan N. <jan...@gm...>]

Op ma 17 nov 2025 om 15:51 schreef Csaba Nemethi <csa...@t-...>:
> This is a CFV for TIP 736: Scroll entry with mouse wheel (formerly: Add
> mouse wheel and <TouchpadScroll> event bindings for entry widgets).

TIP #736:   YES

Regards,
     Jan Nijtmans

Re: [TCLCORE] CFV: TIP 736

[Harald O. <har...@el...>]

Am 17.11.2025 um 15:50 schrieb Csaba Nemethi:
> Attn. TCT members,
> 
> This is a CFV for TIP 736: Scroll entry with mouse wheel (formerly: Add 
> mouse wheel and <TouchpadScroll> event bindings for entry widgets).
> 
>      https://core.tcl-lang.org/tips/doc/trunk/tip/736.md
> 
> The TIP is kindly sponsored by Marc Culler.  Many thanks, Marc!
> 
> Please send your votes to this list until next Monday, 2025-11-24 20:00 
> UTC.
> 
> Best regards,
> 
> Csaba
> 

+1
great work, Csaba.

As chatted at the start of the telco:
The spinbox big button is a style modification only.
It may be added to the core the same way as "Toolbutton" style option, 
but for the Spinbox.

Thanks for all,
Harald

[TCLCORE] CFV: TIP 736

[Csaba N. <csa...@t-...>]

Attn. TCT members,

This is a CFV for TIP 736: Scroll entry with mouse wheel (formerly: Add 
mouse wheel and <TouchpadScroll> event bindings for entry widgets).

     https://core.tcl-lang.org/tips/doc/trunk/tip/736.md

The TIP is kindly sponsored by Marc Culler.  Many thanks, Marc!

Please send your votes to this list until next Monday, 2025-11-24 20:00 UTC.

Best regards,

Csaba

-- 
Csaba Nemethi  https://www.nemethi.de  mailto:csa...@t-...

[TCLCORE] Biweekly telco on 17th of November at 12:00 UTC (TIP 561 added)

[Harald O. <har...@el...>]

Dear Tcl/Tk team,

please feel invited to the next Tcl/Tk beweekly telco:
17th of November at 12:00 UTC
At: https://meet.jit.si/TclMonthlyMeetup

Agenda proposal:

Top 1) Release calender (TIP 713)
- 9.0.3: RELEASED
Congratulations!
https://core.tcl-lang.org/tk/info/cf296ac849
ttk::notebook on the MAC
* Fixed by Csaba
* Still overlapping tabs, Nicolas does not like that, Marcs comment 
requested.
To many 9.0.x releases scheduled in 2026. One every 4 month to 1 each 6 
month would be sufficient. Maybe, the next may be skipped.

- 9.1a1: November (2 weeks left)
May be late, Don not in telco but probably back in service.

Top 2) Test reform by Eric Leunissen
Big appreciation! More welcomed!

Top 3) TIP 672: $(1+1)
Looks like the implementation moves to {=}(1+1)
More welcome.
It was thought, that $((1+1)) was the way to go.
And this is JIM compliant.
The implementation may require a new tolken type.

Top 4) TIP 723: timer in/at
"clock clicks" is also a candidate for monotonic clock.
Really important to have monotonic clock.
It is seen as a bugfix.
after should be based on monotonic time.
clock command should be wall clock.
Current API Tcl_GetTime may use monotonic clock.
New API Tcl_GetWallclockTime may use wall clock.
Strong support!
Need help for Linux

Top 5) TIP 733: accessability
Accepted, merged. Tests pass now.
Ready for alpha release for additional testing.
Need people to try it out.

On Linux, would be great to decide on runtime, if support libraries are 
present, not at compile time.
(like with icu library in tk)

Top 6) TIP 732: TCL library path
Good work!
Initialization is a big swamp in Tcl.
Mac_OS makes most troubble.

Top 7) TIP 715: supported build systems
Should be voted next year.
Issue is C11, what is a fact, as main branch requires C11.
C11 is 14 years old.

Top 8) TIP 615 string is index
Title of TIP: better say "syntax on string index".
Supported!

Top 9) TIP 736: Mouswheel for entry
Scroll entry with mousewheel.

Implementation is complex, as there are many cases.
Entry cursor may get out of the view. That is intended.

Top 10) TIP 737: Encoding API
C API for current functionality.
Output more than INT_MAX characters.
Much more complex than planned.
Much overlap with other functions, but no other function handles big data.
Internal heritage of other functions.
Not much usage for end users

Top 11) Assembler for tk.dll finalise -> Structured Error Handler
Strange, that Tk uses TkFinalize which does not throw exceptions.
The reason for this is unknown.
Would be better to know exceptions.
It is now turned on in debug mode.
Support to not tolerant against bugs.
Try in alpha to find out why this is done.

Top 12) AOB
None

Top 13) Next meeting:
         1st of December 12:00 UTC.
         (Harald is not available from 9-13 UTC, so may only do it one 
hour later.)


Thank you for all,
Harald

[TCLCORE] CFV warning: TIP #615: string is index

[Jan N. <jan...@gm...>]

This is a CFV warning for TIP #615. for Tcl 9.1+:
    string is index
    <https://core.tcl-lang.org/tips/doc/trunk/tip/615.md>

If you think this is a bad idea, speak up now. If not,
I'll start the vote in a few days.

Regards,
         Jan Nijtmans

Re: [TCLCORE] Biweekly telco on 17th of November at 12:00 UTC (TIP 561 added)

[Colin M. <col...@ya...>]

Hello All,

I see the expr alternative issue is on the agenda again. Unfortunately I 
am not able to join this call to discuss it.  I have drafted an updated 
TIP for my proposal and copied it to tcl-core.  I have started looking 
at implementing this in C but that will take me some time.

Colin.

On 14/11/2025 14:40, Harald Oehlmann wrote:
> Dear Tcl/Tk team,
>
> please feel invited to the next Tcl/Tk beweekly telco:
> 17th of November at 12:00 UTC
> At: https://meet.jit.si/TclMonthlyMeetup
>
> Agenda proposal:
>
> Top 1) Release calender (TIP 713)
> - 9.0.3: RELEASED
> - 9.1a1: November (2 weeks left)
> Top 2) Test reform by Eric Leunissen
> Top 3) TIP 672: $(1+1)
> Top 4) TIP 723: timer in/at
> Top 5) TIP 733: accessability
> Top 6) TIP 732: TCL library path
> Top 7) TIP 715: supported build systems
> Top 8) TIP 615 string is index
> Top 9) TIP 736: Mouswheel for entry
> Top 10) TIP 737: Encoding API
> Top 11) Assembler for tk.dll finalise -> Structured Error Handler
> Top 12) AOB
> Top 13) Next meeting:
>        1st of December 12:00 UTC.
>        (Harald is not available)
>
>
> Thank you for all,
> Harald
>
>
> _______________________________________________
> Tcl-Core mailing list
> Tcl...@li...
> https://lists.sourceforge.net/lists/listinfo/tcl-core

[TCLCORE] Fwd: New TIP for expr alternative `=` ?

[Colin M. <col...@ya...>]

Copying this to tcl-core:


-------- Forwarded Message --------
Subject: 	New TIP for expr alternative `=` ?
Date: 	Thu, 13 Nov 2025 18:29:39 +0000
From: 	Colin Macleod <col...@ya...>
To: 	Donal Fellows <don...@ma...>



Dear Donal,

I'm emailing you as TIP editor.  I have put together a TIP for my 
updated proposal for a concise alternative to `expr` which I will 
attach.  I am unsure whether to make this an update of my previous TIP 
676 or to withdraw that and create a new one.  It addresses the same 
issue as TIP 676 but in a significantly different way. Can you advise?

Thanks very much,
Colin.

Re: [TCLCORE] Merge intention for Tk project skip_entire_test_files

[<apn...@ya...>]

+1

-----Original Message-----
From: Harald Oehlmann <har...@el...> 

Great work, Erik.
Even if there is seldomely any feedback, we all appreciate your 
contribution to make Tk better.

Thanks for all,
Harald

[TCLCORE] Biweekly telco on 17th of November at 12:00 UTC (TIP 561 added)

[Harald O. <har...@el...>]

Dear Tcl/Tk team,

please feel invited to the next Tcl/Tk beweekly telco:
17th of November at 12:00 UTC
At: https://meet.jit.si/TclMonthlyMeetup

Agenda proposal:

Top 1) Release calender (TIP 713)
- 9.0.3: RELEASED
- 9.1a1: November (2 weeks left)
Top 2) Test reform by Eric Leunissen
Top 3) TIP 672: $(1+1)
Top 4) TIP 723: timer in/at
Top 5) TIP 733: accessability
Top 6) TIP 732: TCL library path
Top 7) TIP 715: supported build systems
Top 8) TIP 615 string is index
Top 9) TIP 736: Mouswheel for entry
Top 10) TIP 737: Encoding API
Top 11) Assembler for tk.dll finalise -> Structured Error Handler
Top 12) AOB
Top 13) Next meeting:
        1st of December 12:00 UTC.
        (Harald is not available)


Thank you for all,
Harald

Re: [TCLCORE] Tcl/Tk 9.0.3 Release Candidate

[Marc C. <cul...@gm...>]

In fact, it is not uncommon for test failures to be related to race
conditions and timings are often different for the development build versus
the deployment build.

- Marc

On Thu, Nov 13, 2025, 3:00 PM Torsten Berg <be...@ty...> wrote:

> Ah, that's interesting. Yes, this sounds like a valid explanation for
> running tests twice. Thanks for sharing!
>
> > Am 13.11.2025 um 17:21 schrieb Erik Leunissen <el...@xs...>:
> >
> > On 11/13/25 01:17, Torsten Berg wrote:
> >> I just wonder why the tests run twice (and for me it also seems that
> during the build things are done multiple teimes) but that may just be my
> lack of wisdom regarding the build process.
> >
> > When I invoke "make test" on macOS from the source code dir, tests are
> run twice:
> > A. optimized (no symbols), indicated by "Deployment"
> > B. unoptimized with debugging symbols, indicated by "Development"
> >
> > I am able to run just one of them when invoking "make test" from the
> respective build directories (siblings of the source directory) [*]:
> > build/Tcl/Deployment
> > build/Tcl/Development
> >
> > That's just what I found out about how things are organized on my
> system. I'm too
> > unfamiliar with macOS to be able to say whether that is how it's
> supposed to be,
> > or whether other setups/invocations are advised.
> >
> > Regards,
> > Erik Leunissen.
> > --
> > [*] the setup for Tk is analoguous
> >
>
>
>
> _______________________________________________
> Tcl-Core mailing list
> Tcl...@li...
> https://lists.sourceforge.net/lists/listinfo/tcl-core
>

[TCLCORE] Announce: Tclkits and Zipkits for Tcl/Tk 9.0.3

[Paul O. <pa...@po...>]

Tclkits and Zipkits using Tcl/Tk 9.0.3 for several platforms are available at https://www.bawt.tcl3d.org/apps.html

Regards,
Paul

Re: [TCLCORE] Tcl/Tk 9.0.3 Release Candidate

[Torsten B. <be...@ty...>]

Ah, that's interesting. Yes, this sounds like a valid explanation for running tests twice. Thanks for sharing!

> Am 13.11.2025 um 17:21 schrieb Erik Leunissen <el...@xs...>:
> 
> On 11/13/25 01:17, Torsten Berg wrote:
>> I just wonder why the tests run twice (and for me it also seems that during the build things are done multiple teimes) but that may just be my lack of wisdom regarding the build process.
> 
> When I invoke "make test" on macOS from the source code dir, tests are run twice:
> A. optimized (no symbols), indicated by "Deployment"
> B. unoptimized with debugging symbols, indicated by "Development"
> 
> I am able to run just one of them when invoking "make test" from the respective build directories (siblings of the source directory) [*]:
> build/Tcl/Deployment
> build/Tcl/Development
> 
> That's just what I found out about how things are organized on my system. I'm too
> unfamiliar with macOS to be able to say whether that is how it's supposed to be,
> or whether other setups/invocations are advised.
> 
> Regards,
> Erik Leunissen.
> --
> [*] the setup for Tk is analoguous
> 

Re: [TCLCORE] Merge intention for Tk project skip_entire_test_files

[Erik L. <el...@xs...>]

On 11/13/25 11:22, Harald Oehlmann wrote:

> Even if there is seldomely any feedback, we all appreciate your contribution to make Tk better.

That's good to know.


In that case :-) ...

              I'll continue with the next job regarding the test suite
(starting in a few days).

Erik Leunissen.
--

Re: [TCLCORE] Tcl 9.0.3 RELEASED

[Harald O. <har...@el...>]

Am 13.11.2025 um 10:06 schrieb Jan Nijtmans:
> Tcl 9.0.3 Release Announcement
> ==============================
> 
> Nov 13, 2025
> 
> The Tcl Core Team is pleased to announce the release of Tcl 9.0.3.
> This is the second patch of Tcl 9.0.
Great work, I appreciate!
Thanks to Jan as interim release manager!
Just in time, Don Porter is back to work due to the end of the work ban 
in the US. So, welcome back!

Take care,
Harald

Re: [TCLCORE] Merge intention for Tk project skip_entire_test_files

[Harald O. <har...@el...>]

Am 13.11.2025 um 11:02 schrieb Erik Leunissen via Tcl-Core:
> On 11/10/25 20:48, Erik Leunissen via Tcl-Core wrote:
>>
>> Unless someone raises founded objections in the meantime, I will carry 
>> out the
>> intended merge on next Wednesday (or the first opportunity for me 
>> thereafter),
>> but no earlier than that trunk is in good order, i.e. the Tk test 
>> suite passes
>> for trunk on Github CI for all platforms.
>>
> 
> Now done.
> 
> Erik Leunissen.
Great work, Erik.
Even if there is seldomely any feedback, we all appreciate your 
contribution to make Tk better.

Thanks for all,
Harald

Re: [TCLCORE] Merge intention for Tk project skip_entire_test_files

[Erik L. <el...@xs...>]

On 11/10/25 20:48, Erik Leunissen via Tcl-Core wrote:
> 
> Unless someone raises founded objections in the meantime, I will carry out the
> intended merge on next Wednesday (or the first opportunity for me thereafter),
> but no earlier than that trunk is in good order, i.e. the Tk test suite passes
> for trunk on Github CI for all platforms.
> 

Now done.

Erik Leunissen.
--

[TCLCORE] Tk 9.0.3 RELEASED

[Jan N. <jan...@gm...>]

Tk 9.0.3 Release Announcement
==============================

Nov 13, 2025

The Tcl Core Team is pleased to announce the release of Tk 9.0.3.
This is the third patch release of Tk 9.0.

The Tk Toolkit is an extension to Tcl, providing commands and support
for the creation of graphical user interfaces.  Tk originates with John
Ousterhout and his team at U.C. Berkeley in the late 1980s.  Its development
is continued by the efforts of a global network of volunteers guided by
the Tcl Core Team.

We would like to express our gratitude to all those who submit bug
reports and patches.  This information is invaluable in enabling us
to identify and eliminate problems. Such reports can be submitted here.

 > [Tk Ticket Tracker](https://core.tcl-lang.org/tk/ticket)

We ask that you log in (anonymous if you wish) to create tickets.
This deters abuse of the ticketing system:

 > [Tk Contributor Login](https://core.tcl-lang.org/tk/login)

Where to get the new releases
=============================

Tk 9.0.3 sources are freely available as open source from the Tcl
SourceForge project's file distribution area:

 > [Tcl Source Distribution](https://sourceforge.net/projects/tcl/files/)

The Tk 9.0.3 distribution is source code only.  We keep links to some
third parties offering pre-built binaries for various systems here:

 > [Tcl Binary Distribution](https://www.tcl-lang.org/software/tcltk/bindist.html)

Tcl Improvement Proposals (TIPs)
================================

Each new user-visible feature in Tk should find its origins in a Tcl
Improvement Proposal (TIP).  TIPs are published, edited, considered and
voted in public, and should contain valuable information about how a
feature came to be the way it is.  See the full collection here:

 > [TIP Index](https://tip.tcl-lang.org/)

Tk 9.0.3 Changes Summary
========================

(from changes.md in the source code distribution)

The source code for Tk is managed by fossil.  Tk developers coordinate all
changes to the Tk source code at

 > [Tk Source Code](https://core.tcl-lang.org/tk/)

Release Tk 9.0.3 arises from the check-in with tag `core-9-0-3`.

Tk 9.0.3 continues the Tk 9.0 series of releases.  The Tk 9.0 series
does not support Tcl 8.6.  The Tk 9.0 series extends the Tcl 9.0 series.
To make use of Tk 9.0.3, first a Tcl 9.0 release must be present.
As new Tk features are developed, expect them to appear in Tk 9, but not
necessarily in Tk 8.

Tk patch releases have the primary purpose of delivering bug fixes
to the userbase.

# Bug fixes
 - [Pressing Alt-<key> triggers an error
popup](https://core.tcl-lang.org/tk/tktview/a30d3e)
 - [error popup on ttk::scale with invisible
trough](https://core.tcl-lang.org/tk/tktview/d25b72)
 - [loading windows dll's in Cygwin doesn't work with
high-entropy-va](https://core.tcl-lang.org/tk/tktview/06892c)
 - [TIP 604 widget demo: canvas text fill color not
specified](https://core.tcl-lang.org/tk/tktview/6cc48b)
 - [Undefined behavior while touchpad
scrolling](https://core.tcl-lang.org/tk/tktview/92e408)
 - [leak in XCreateBitmapFromData() in
ImgGetPhoto()](https://core.tcl-lang.org/tk/tktview/6c4795)
 - [Add support for Copy/Cut/Paste keys in
X11](https://core.tcl-lang.org/tk/tktview/04e173)
 - [Aqua: compiler warning in
tkMacOSXImage.c](https://core.tcl-lang.org/tk/tktview/7f3751)
 - [tkpWinRopModes[GXnoop] is R2_NOT, should be
R2_NOP](https://core.tcl-lang.org/tk/tktview/95da0f)
 - [image svg: support paint order and fix possible nan race
condition](https://core.tcl-lang.org/tk/tktview/b43dbc)
 - [Windows: install man pages](https://core.tcl-lang.org/tk/tktview/816739)
 - [Windows: install pkg-config
file](https://core.tcl-lang.org/tk/tktview/2c240b)
 - [Windows: Missing border pixels in alt
theme](https://core.tcl-lang.org/tk/tktview/8b725d)
 - [nanosvg upstream changes: #273, #275: add paint order
parsing](https://core.tcl-lang.org/tk/tktview/b43dbc0061)
 - [Artifacts in a few themed widgets on x11 and
aqua](https://core.tcl-lang.org/tk/tktview/265ff2)
 - [Core dump Tk_GetFontFromObj](https://core.tcl-lang.org/tk/tktview/a80e5f)
 - [MS-Win canvas arcs with small extent are drawn 360
degrees](https://core.tcl-lang.org/tk/info/6051a9)
 - [pixel values can report false list
lengths](https://core.tcl-lang.org/tk/info/2a0a83)
 - [Xft text is unusable for 32-bit visual when default visual is
24-bit](https://core.tcl-lang.org/tk/info/c23f79)
 - [tk print: several issues printing i18n
text](https://core.tcl-lang.org/tk/info/4e7a78)
 - [tk print, canvas: raw smooth not
implemented](https://core.tcl-lang.org/tk/info/e10819)
 - [Aqua: avoid use-after-free during
RefocusGrabWindow()](https://core.tcl-lang.org/tk/info/6da885)
 - [Aqua: ttk::notebook tabs](https://core.tcl-lang.org/tk/info/cf296a)
 - [Fix crash on exit due to faulty asm code in
DllMain](https://core.tcl-lang.org/tk/info/44b34c)

## Known bugs
 - [Inconsistent reporting of child geometry changes to grid
container](https://core.tcl-lang.org/tk/tktview/beaa8e)
 - [Inconsistency in whether widgets allow negative
borderwidths](https://core.tcl-lang.org/tk/tktview/5f739d)
 - [slow widget creation if default font is not
used](https://core.tcl-lang.org/tk/tktview/8da7af)
 - [The wm manage command does not work on current macOS
versions](https://core.tcl-lang.org/tk/tktview/8a6012)
 - [Slow processing irregular
transparencies](https://core.tcl-lang.org/tk/tktview/919066)
 - [text's cursor width on 0th
column](https://core.tcl-lang.org/tk/tktview/47fbfc)
 - [text widget breaks graphemes with combining diacritical
marks](https://core.tcl-lang.org/tk/tktview/442208)

For additional information:
===========================

Please visit the Tcl Developer Xchange web site:

 > [Tcl Developer Xchange](https://www.tcl-lang.org/)

This site contains a variety of information about Tcl/Tk in general, the
core Tcl and Tk distributions, Tcl development tools, and much more.

--
Tcl Core Team and Maintainers
Jan Nijtmans, (replacement) Tcl Core Release Manager

[TCLCORE] Tcl 9.0.3 RELEASED

[Jan N. <jan...@gm...>]

Tcl 9.0.3 Release Announcement
==============================

Nov 13, 2025

The Tcl Core Team is pleased to announce the release of Tcl 9.0.3.
This is the second patch of Tcl 9.0.

Tcl is the Tool Command Language originated by John Ousterhout and his
team at U.C. Berkeley in the late 1980s.  Its development is continued
by the efforts of a global network of volunteers guided by the Tcl Core Team.

We would like to express our gratitude to all those who submit bug
reports and patches.  This information is invaluable in enabling us
to identify and eliminate problems. Such reports can be submitted here.

 > [Tcl Ticket Tracker](https://core.tcl-lang.org/tcl/ticket)

We ask that you log in (anonymous if you wish) to create tickets.
This deters abuse of the ticketing system:

 > [Tcl Contributor Login](https://core.tcl-lang.org/tcl/login)

Where to get the new releases
=============================

Tcl 9.0.3 sources are freely available as open source from the Tcl
SourceForge project's file distribution area:

 > [Tcl Source Distribution](https://sourceforge.net/projects/tcl/files/)

The Tcl 9.0.3 distribution is source code only.  We keep links to some
third parties offering pre-built binaries for various systems here:

 > [Tcl Binary Distribution](https://www.tcl-lang.org/software/tcltk/bindist.html)

Tcl Summary
===========

The Tcl distribution delivers C source code that builds into a C library
providing interpreters and related support to execute programs written
in the Tcl programming language.  Source code for the application program
`tclsh` is also included.  `tclsh` provides a shell for either interactive
execution of Tcl commands, or execution of files containing Tcl programs.

Tcl is an extensible language, and the Tcl C library provides interfaces
for the creation of extension libraries adding new commands and features
to the core Tcl command set.  Tcl 9 debuts the full feature set needed to
package an application written in C and Tcl into a single file executable
exploiting virtual filesystem archives.

Tcl Improvement Proposals (TIPs)
================================

Each new user-visible feature in Tcl should find its origins in a Tcl
Improvement Proposal (TIP).  TIPs are published, edited, considered and
voted in public, and should contain valuable information about how a
feature came to be the way it is.  See the full collection here:

 > [TIP Index](https://tip.tcl-lang.org/)

Tcl 9.0.3 Changes Summary
=========================

(from changes.md in the source code distribution)

The source code for Tcl is managed by fossil.  Tcl developers coordinate all
changes to the Tcl source code at

 > [Tcl Source Code](https://core.tcl-lang.org/tcl/timeline)

Release Tcl 9.0.3 arises from the check-in with tag `core-9-0-3`.

Tcl patch releases have the primary purpose of delivering bug fixes
to the userbase.

# Bug fixes
 - [On Unix, IsTimeNative() always defined but not always
used](https://core.tcl-lang.org/tcl/tktview/6b8e39)
 - [Tweak install permissions](https://core.tcl-lang.org/tcl/tktview/31d4fa)
 - [interp creation resets encoding directory search
path](https://core.tcl-lang.org/tcl/tktview/87b697)
 - [Pointer arithmetic with NULL in
buildInfoObjCmd()](https://core.tcl-lang.org/tcl/tktview/85fc8b)
 - [TclPushVarName(): pointer
overflow](https://core.tcl-lang.org/tcl/tktview/77059c)
 - [Add IWYU export pragma
annotations](https://core.tcl-lang.org/tcl/tktview/c7dc59)
 - [Windows: Install man pages](https://core.tcl-lang.org/tcl/tktview/3161b7)
 - [Windows: Install pkgconfig](https://core.tcl-lang.org/tcl/tktview/1cf49a)
 - [Non-existent variables are ignored if re is
{}](https://core.tcl-lang.org/tcl/tktview/cb03e5)
 - [bug in single-argument 'max' with
bignums](https://core.tcl-lang.org/tcl/tktview/8dd280)

# Updated bundled packages, libraries, standards, data
 - Itcl 4.3.5
 - http 2.10.1
 - opt 0.4.10
 - platform 1.1.0
 - sqlite3 3.51.0
 - tcltest 2.5.10
 - Thread 3.0.4
 - TDBC\* 1.1.13
 - dde 1.4.6
 - Unicode 17.0.0

# Known bugs
 - [changed behaviour wrt command names, namespaces and
resolution](https://core.tcl-lang.org/tcl/tktview/f14b33)
 - [windows dos device paths inconsistencies and missing
functionality](https://core.tcl-lang.org/tcl/tktview/d8f121)
 - [load library (dll) from zipfs-library causes a leak in temporary
folder](https://core.tcl-lang.org/tcl/tktview/a8e4f7)
 - [lsearch -sorted -inline -subindices incorrect
result](https://core.tcl-lang.org/tcl/tktview/bc4ac0)
 - ["No error" when load fails due to a missing secondary
DLL](https://core.tcl-lang.org/tcl/tktview/66aa81c90a)

For additional information:
===========================

Please visit the Tcl Developer Xchange web site:

 > [Tcl Developer Xchange](https://www.tcl-lang.org/)

This site contains a variety of information about Tcl/Tk in general, the
core Tcl and Tk distributions, Tcl development tools, and much more.

--
Tcl Core Team and Maintainers
Jan Nijtmans, (replacement) Tcl Core Release Manager

Re: [TCLCORE] Tcl/Tk 9.0.3 Release Candidate

[Erik L. <el...@xs...>]

On 11/13/25 01:17, Torsten Berg wrote:
> 
> I just wonder why the tests run twice (and for me it also seems that during the build things are 
> done multiple teimes) but that may just be my lack of wisdom regarding the build process.
> 

When I invoke "make test" on macOS from the source code dir, tests are run twice:
A. optimized (no symbols), indicated by "Deployment"
B. unoptimized with debugging symbols, indicated by "Development"

I am able to run just one of them when invoking "make test" from the respective build directories 
(siblings of the source directory) [*]:
build/Tcl/Deployment
build/Tcl/Development

That's just what I found out about how things are organized on my system. I'm too
unfamiliar with macOS to be able to say whether that is how it's supposed to be,
or whether other setups/invocations are advised.

Regards,
Erik Leunissen.
--
[*] the setup for Tk is analoguous

Re: [TCLCORE] Tcl/Tk 9.0.3 Release Candidate

[Torsten B. <be...@ty...>]

OK, Tcl 9.0.3 now builds fine using this:

    cd $HOME/Tcl/distrib/
    # there is a folder tcl9.0.3-arm64 there with the sources
    export CFLAGS='-mmacosx-version-min=11'
    make -j8 -C tcl9.0.3-arm64/macosx
    sudo make -C tcl9.0.3-arm64/macosx install

I did not use USE_TCL_STUBS=1 and sqlite3.51.0 still built and runs fine now with the change you made.

However, I noted that the message about "Tcl 8" is still there, not only for sqlite3.51.1 but also for itcl4.3.5, tdbcl1.1.13 and others:

    Building package 'itcl4.3.5' for Tcl 8
    ...
    Building package 'sqlite3.51.0' for Tcl 8
    ...
    Building package 'tdbc1.1.13' for Tcl 8
    ...



I then ran

    make -C tcl9.0.3-arm64/macosx test

and had no errors. It only complained about some missing dylib of tdbc drivers for some databases but I guess that is normal when I do not have these databases installed.

I just wonder why the tests run twice (and for me it also seems that during the build things are done multiple teimes) but that may just be my lack of wisdom regarding the build process.

Now, on to Tk building and testing.

Torsten


> Am 12.11.2025 um 00:10 schrieb Torsten Berg <be...@ty...>:
> 
> Great, will test this tomorrow!
> 
> Regards, Torsten
> 
> 
>> Am 11.11.2025 um 22:38 schrieb Jan Nijtmans <jan...@gm...>:
>> 
>> Op di 11 nov 2025 om 12:37 schreef Torsten Berg <be...@ty...>:
>>> I keep getting this error on macOS (Ventrua 13.7.8)
>> ....
>>> ld: Undefined symbols:
>>> 
>>> _Tcl_GetBoolFromObj, referenced from:
>>> 
>>>     _DbMain in tclsqlite3.o
>> 
>> I managed to reproduce this. The problem was in tclDecls.h. The macro
>> deciding between Tcl_GetBooleanFromObj and Tcl_GetBoolFromObj
>> only works as intended in optimized builds: The linker will discover
>> that Tcl_GetBoolFromObj is never used in a Tcl8 build, and simply
>> removes the call. In non-optimized builds, that doesn't happen.
>> 
>> Fix committed. This means there will be a rc2
>>   <https://core.tcl-lang.org/tcl/info/44c0681b3b238ae7>
>> 
>>> I had trouble compiling SQLite 3.51.0 from its sources too, until a user fixed a problem:
>>> https://sqlite.org/forum/forumpost/389916bad5
>> 
>> This is a different thing. My recommendation is always to build
>> with stubs (USE_TCL_STUBS=1) and link with the stubs library.
>> The default SQLite build doesn't do that, it just keeps the
>> symbols undefined, hoping they can be found at runtime.
>> Most platforms can do that, but at least Windows and AIX cannot.
>> 
>> Thanks!
>>       Jan Nijtmans
> 
> 
> 
> _______________________________________________
> Tcl-Core mailing list
> Tcl...@li...
> https://lists.sourceforge.net/lists/listinfo/tcl-core

[TCLCORE] SafeTCL: have there been any escapes

[Phil M. <phi...@gm...>]

G'day,

SafeTCL is an amazing thing of course and we've been using it for the last
15 years to do various control/analysis things. As part of the analysis of
a
new tool it'd be nice to see if there are any error/escapes from SafeTCL
(beyond
the starvation one).

1. Have there ever been any escapes from SafeTCL excluding
misconfigurations?
2. I've not found any whence this email?

As penance I'll write this up and link it to the appropriate pages.

-- 
Phil Maker .
web: https://eds.power.on.net/EDS or https://eds.power.net/laot-pub
emails: <phi...@gm...>,       <ph...@ci...>,
<ph...@ci...>
 <pj...@gn...>
phone: +61 (0) 439 223 469