Proposal for a Gnats 3.114 release

Proposal for a Gnats 3.114 release

[Yngve Svendsen]

Gnats 4 will take some time still, as far as I can see. Therefore, I would 
like to propose that we reease the current tip of the v3 tree as Gnats 
3.114. A "wrapup release" of Gnats 3 would complement Gnatsweb 2.8.0 
perfectly, I think.

Fixes for the following bugs have been committed after the 3.113 release:
- gnatsd used to crash when a 'classes' file contained a large number of 
classes.
- Fixed to gnats.el
- When the X-GNATS-Notify: field is present but empty, gnatsd segfaults
- Some added timezones.

And the chief reason that I want a 3.114 release: The data loss bug 
described in 
http://sources.redhat.com/ml/gnats-devel/2001-q2/msg00058.html . We have 
been running the latest Gnats 3 CVS, plus the fix detailed in the posting 
above for a week in full production, and have seen absolutely no problems.

Milan: I am ready to take on the task of committing the fix above, packing 
the new release (with Gnatsweb 2.8.0), uploading and announcing it.

What do you all think?

Yngve Svendsen
IS Engineer
Clustra AS, Trondheim, Norway
yngve.svendsen@clustra.com

Re: Proposal for a Gnats 3.114 release

[Rick Macdonald]

On Fri, 4 May 2001, Yngve Svendsen wrote:

> Gnats 4 will take some time still, as far as I can see. Therefore, I would 
> like to propose that we reease the current tip of the v3 tree as Gnats 
> 3.114. A "wrapup release" of Gnats 3 would complement Gnatsweb 2.8.0 
> perfectly, I think.

I have a dev version of TkGnats that I should have released a year ago. If
I update my web page, will you include it in the gnats 3.114 release? It
will be tkgnats v3.0.18.

Here is the changelog:

=================
Changes in 3.0.17 PRE_RELEASE
=================

27APR2000

o  The Help for Category, Submitter-Id and Responsible is now dynamic
   and descriptions are taken from the categories, submitters and
   responsibles files as configured in gnats-adm. (This info is
   retrieved via gnatsd or [n]query-pr.

09MAR2000

o  In SendPr, unqualified names in the Notify-List only get the 
   default domainname appended if the name is _not_ found in the
   responsibles list. For this to work, it is assumed that the
   first and third fields of the responsible entry are different.
   For example, this will not get the default domainname appended:
           rickm:Rick Macdonald:rickm@example.com
   but this will:
           rickm:Rick Macdonald:
   
o  A new config variable TkGnats(QueryFileMenu) adds items to the File
   menu of the main Query window. See tkgnats.config for examples.

18FEB2000

o  Fixed error when handling gnatsd access denied problems:

   Error in startup script: can't read "TkGnats(UserAccess)": no such
element
   in array while executing
   "if {$TkGnats(UserAccess) == "noaccess"} {
       # The userid/password isn't valid or otherwise returned
access=none.
       if {[info exist..."
       (procedure "get_gnats_config" line 13) invoked from within
       "get_gnats_config"
           (file "/opt/tkgnats/share/tkgnats/tkquerypr.tcl" line 64)

o  Updated delete_pr to work. It's been broken since about gnats-3.107.

16FEB2000

o  Line wrapping in the Full, Raw and Description_Summary reports
   has been fixed, again. 

o  The Full PR report now begins a new page for each PR.

04FEB2000

o  If you edit a PR where the Responsible person has been removed 
   from the responsibles file, the PR would get saved with a blank
   Responsible field. This triggers a bug in gen-index which causes
   the previous index record to be duplicated and the one with the
   blank Responsible person gets lost. To fix this, the Responsible
   field is now hard-wired as mandatory for the EditPr window, and
   the user is warned about the invalid value.

...RickM...

Re: Proposal for a Gnats 3.114 release

[Yngve Svendsen]

At 15:54 04.05.2001 -0600, Rick Macdonald wrote:
>I have a dev version of TkGnats that I should have released a year ago. If
>I update my web page, will you include it in the gnats 3.114 release? It
>will be tkgnats v3.0.18.

Absolutely. If Milan OK's me handling this, you can just send it to me.

- Yngve

Re: Proposal for a Gnats 3.114 release

[Rick Macdonald]

On Fri, 4 May 2001, Yngve Svendsen wrote:

> At 15:54 04.05.2001 -0600, Rick Macdonald wrote:
> >I have a dev version of TkGnats that I should have released a year ago. If
> >I update my web page, will you include it in the gnats 3.114 release? It
> >will be tkgnats v3.0.18.
> 
> Absolutely. If Milan OK's me handling this, you can just send it to me.

In the past, only the tarfile was put into contrib. The zip file is for
Windows (identical code but with lf/cr conversion), but don't bother
including both unless you think it's worthwhile.

http://www.cuug.ab.ca/~macdonal/tkgnats/tkgnats-3.0.18.tar.gz
http://www.cuug.ab.ca/~macdonal/tkgnats/tkgnats-3.0.18.zip

I found these additional changes and included them:

04MAY2001

o  Add an entry field to the print dialog to collect a title to put 
   at the top of a printed report.  If the title entered is blank 
   the default is used. (Thanks to jamie.briggs@computalog.com)

o  Fixes to the DeletePR function (only available when accessing
   the GNATS database directly on disk (not recommended)).
   (thanks to John Zuckerman <john@math.tamu.edu>).

...RickM...

Re: Proposal for a Gnats 3.114 release

[Milan Zamazal]

>>>>> "YS" == Yngve Svendsen <yngve.svendsen@clustra.com> writes:

    YS> Gnats 4 will take some time still, as far as I can see.

Yes. :-(

    YS> Therefore, I would like to propose that we reease the current
    YS> tip of the v3 tree as Gnats 3.114.

For the reason above, I think it's not a bad idea to make a bug fix
release of GNATS 3.113.

However, this version should be only useful to the people who already
run GNATS on their production systems and have problems with segfaults.
It shouldn't attract new GNATS users and let them to try the new version
and report bugs about it (most GNATS 3 bugs are not present in GNATS 4,
so this would only add me an extra work).  So I'd suggest to use the
version number 3.113.1 or something similar and clearly state that this
is only a bug-fix release and suggest to wait for GNATS 4 to be released
"soon" (in Gerald's terminology:-) for those who don't have problems
with their 3.113 systems.

    YS> Milan: I am ready to take on the task of committing the fix
    YS> above, packing the new release (with Gnatsweb 2.8.0), uploading
    YS> and announcing it.

OK, thanks, but please consult each step with me before performing it.
I also think I've got one uncommitted bug fix in my tree, I'll commit it
today or tomorrow.

As for TkGnats, the new version should definitely be included.

Regards,

Milan Zamazal

-- 
When you're in a fight with an idiot, it's difficult for other people to tell
which one the idiot is.                       -- Bruce Perens in debian-devel

Re: Proposal for a Gnats 3.114 release

[Yngve Svendsen]

At 17:33 04.05.2001 -0600, Rick Macdonald wrote:
>On Fri, 4 May 2001, Yngve Svendsen wrote:
>
> > At 15:54 04.05.2001 -0600, Rick Macdonald wrote:
> > >I have a dev version of TkGnats that I should have released a year ago. If
> > >I update my web page, will you include it in the gnats 3.114 release? It
> > >will be tkgnats v3.0.18.
> >
> > Absolutely. If Milan OK's me handling this, you can just send it to me.
>
>In the past, only the tarfile was put into contrib. The zip file is for
>Windows (identical code but with lf/cr conversion), but don't bother
>including both unless you think it's worthwhile.
>
> http://www.cuug.ab.ca/~macdonal/tkgnats/tkgnats-3.0.18.tar.gz
> http://www.cuug.ab.ca/~macdonal/tkgnats/tkgnats-3.0.18.zip

Thanks. I updated TkGnats to 3.0.18 to the branch that will become 3.113.1 
in a few days. I though it best not to include the Windows version, and I 
hope people who want it are able to find your website to download it.

Yngve Svendsen
IS Engineer
Clustra AS, Trondheim, Norway
yngve.svendsen@clustra.com