You are viewing a plain text version of this content. The canonical link for it is here.
Posted to rivet-dev@tcl.apache.org by Massimo Manghi <ma...@unipr.it> on 2016/06/22 09:01:24 UTC
rivet 3.0 release roadmap
Hi guys,
I've been using the code in trunk in the last 6 months and I didn't spot
any relevant issues with it. I would like to write down a few ideas for
releasing it as rivet-3.0.0 in 3 or 4 months. What is missing to rivet 3.0?
- A 3.0 branch should be created in order to separate code additions
that could bring instability in the code. Yes, I know, I'm a paranoid,
there is no 'svn commit' storm taking place on our repository :-(
- Basically the test suite should be able to run with different MPM
and different MPM bridges (not every pair of MPM module and bridge are
possible though). Furthermore some tests are just meant to avoid
regression with commands and in principle should be bridge and MPM
independent. Anyway I will try to make the test suite run in a sort of
'for' loop for each meaningful MPM and bridge pairing
further developments to be done in 'trunk' (hence separated by the 3.0
branch)
- I'm considering to place into the repository the framework I've been
working on for quite a few years. I was afraid to propose it and it
probably needs more work, but if I don't throw it into the repository it
will have no chance to grow. Taking inspiration from a motto quoted on
@members recently the philosophy of the framework is 'less is more (at
least sometimes)'. I mean the idea of the framework was to have a very
simple pattern of collaborating components on which you may develop more
elaborate application features. The code was and it's still written
using Itcl but moving to TclOO should be straightforward. Furthermore
you need it if you want to reproduce tcl.apache.org and
tcl.apache.org/rivet/
- The framework needs a few configuration scripts to be set up
consistently. I'm wondering if we could imagine a way to create some
sort of scripted configuration, something that should enable the
programmer to set up a single configuration directive with a tcl script
that should fetch the basic environmental data and prepare the final
configuration (or blend into it somehow). An alarm bell keeps ringing in
my ears, does it open the way to subtle security holes? I can't figure
out a way to rule it out but also I can't image how this could less
secure of the current way to set up mod_rivet environments
sorry for the lengthy message.
-- Massimo
--
Dipartimento di Neuroscienze
Unit� Biofisica e Fisica Sanitaria
via Volturno 39
43125 Parma
---------------------------------------------------------------------
To unsubscribe, e-mail: rivet-dev-unsubscribe@tcl.apache.org
For additional commands, e-mail: rivet-dev-help@tcl.apache.org
Re: rivet 3.0 release roadmap
Posted by Massimo Manghi <mx...@apache.org>.
Yes, of course this is a good idea. My testing on other platforms
doesn't get any further than having rivet built correctly on the many
architectures supported, or non officially supported, by Debian, but
having it actually tested would be a step forward.
-- Massimo
P.S. I've just read the terrible news about Miguel Sofer who died
recently. I never had a chance to meet him but following the Tcl core
mailing list I remember the alacrity and competence emerging from his
messages when he was trying to improve and solve several issues with the
NRE. His writing had always an unassuming tone and I instinctively liked
him. When we fire Tcl 8.6 we use his hard work and cleverness
---------------------------------------------------------------------
To unsubscribe, e-mail: rivet-dev-unsubscribe@tcl.apache.org
For additional commands, e-mail: rivet-dev-help@tcl.apache.org
Re: rivet 3.0 release roadmap
Posted by Harald Oehlmann <ha...@elmicron.de>.
Hi guy ;-)
Happy to hear from you ! Massimos messages are always a pleasure to
read. I like the details ;-) Rivet 3.0 is great !!!!
About the installation:
- I can propose to test something in the Open Build System, where we
have many Linux Distributions in multiple versions available. Just call
if you want to test something on multiple platforms...
Enjoy,
Harald
Am 22.06.2016 um 11:01 schrieb Massimo Manghi:
> Hi guys,
>
> I've been using the code in trunk in the last 6 months and I didn't spot
> any relevant issues with it. I would like to write down a few ideas for
> releasing it as rivet-3.0.0 in 3 or 4 months. What is missing to rivet 3.0?
>
> - A 3.0 branch should be created in order to separate code additions
> that could bring instability in the code. Yes, I know, I'm a paranoid,
> there is no 'svn commit' storm taking place on our repository :-(
>
> - Basically the test suite should be able to run with different MPM and
> different MPM bridges (not every pair of MPM module and bridge are
> possible though). Furthermore some tests are just meant to avoid
> regression with commands and in principle should be bridge and MPM
> independent. Anyway I will try to make the test suite run in a sort of
> 'for' loop for each meaningful MPM and bridge pairing
>
> further developments to be done in 'trunk' (hence separated by the 3.0
> branch)
>
> - I'm considering to place into the repository the framework I've been
> working on for quite a few years. I was afraid to propose it and it
> probably needs more work, but if I don't throw it into the repository it
> will have no chance to grow. Taking inspiration from a motto quoted on
> @members recently the philosophy of the framework is 'less is more (at
> least sometimes)'. I mean the idea of the framework was to have a very
> simple pattern of collaborating components on which you may develop more
> elaborate application features. The code was and it's still written
> using Itcl but moving to TclOO should be straightforward. Furthermore
> you need it if you want to reproduce tcl.apache.org and
> tcl.apache.org/rivet/
>
> - The framework needs a few configuration scripts to be set up
> consistently. I'm wondering if we could imagine a way to create some
> sort of scripted configuration, something that should enable the
> programmer to set up a single configuration directive with a tcl script
> that should fetch the basic environmental data and prepare the final
> configuration (or blend into it somehow). An alarm bell keeps ringing in
> my ears, does it open the way to subtle security holes? I can't figure
> out a way to rule it out but also I can't image how this could less
> secure of the current way to set up mod_rivet environments
>
> sorry for the lengthy message.
>
> -- Massimo
> --
> Dipartimento di Neuroscienze
> Unità Biofisica e Fisica Sanitaria
> via Volturno 39
> 43125 Parma
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: rivet-dev-unsubscribe@tcl.apache.org
> For additional commands, e-mail: rivet-dev-help@tcl.apache.org
>
--
ELMICRON Dr. Harald Oehlmann GmbH
Koesener Str. 85
06618 Naumburg
Germany
Phone: +49 (0)3445 78112-0
Fax: +49 (0)3445 78112-19
www.Elmicron.de
German legal references:
Geschaeftsfuehrer: Dr. Harald Oehlmann, Jens Oehlmann
UST Nr. / VAT ID No.: DE206105272
HRB 212803 Stendal
---------------------------------------------------------------------
To unsubscribe, e-mail: rivet-dev-unsubscribe@tcl.apache.org
For additional commands, e-mail: rivet-dev-help@tcl.apache.org