[OpenSCAD] Why is the 2015.03 kept as the "official" release?

[Antonio Bueno]

Hello everyone.

I use the latest snapshot regularly so, when I pass an .scad file to
someone else, I've had to warn (either pre- or post-sending the file) to
use a development snapshop instead of the 2015.03 "normal" release quite a
few times.

I've been asked about it, and I don't know the answer, so here I am passing
the question forward :-)

Why is the 2015.03 kept as the "official" release?

[MichaelPFrey]

My guess is that it's just that the package maintainers haven't bothered
to
update it.

This is factually wrong.
The issue is, that there is no newer release.
The release planning is somewhat managed via
https://github.com/openscad/openscad/projects/1
<https://github.com/openscad/openscad/projects/1> and
https://github.com/openscad/openscad/milestone/4
<https://github.com/openscad/openscad/milestone/4> .

The issue is somewhat general development resources and project management.
But even simple documentation can hold back a feature from becoming
considered stable:
https://github.com/openscad/openscad/issues?q=is%3Aopen+is%3Aissue+label%3ADocumentation
<https://github.com/openscad/openscad/issues?q=is%3Aopen+is%3Aissue+label%3ADocumentation>
Do my knowledge, all developers here work on OpenSCAD as just a hobby.
As a developer my self, I have to say: We tend to fix issues that are most
relevant to us personally.
Having a bug fixed or a feature implemented and merged with "master", is
enough to solve the issue for us personally.

Then there is also the code/functional quality and stability of interfaces,
which is higher in priority then cutting edge features.

I my self wonder, why Inputdriver5 (gamepad and 3D mice support) is (still)
a development branch and not a regular experimental feature. (I understand
the historical reason(s), but not the current ones) But add the same time: I
can simply merge it locally, install it and be happy with an OpenSCAD build
with recent Customizer and Inputdriver. Not a big deal for a developer - but
something normal user are unable to do, but at the same time not my issue.

An other issue we have is Windows and MacOS X. Cross-building from Linux to
Windows is "regularly" broken, native building under Windows is a pain (not
helped by the fact, that you need experience in the Linux build chain to
debug the building process under windows). OS X also lacks developers and
has build/test issues - but I do not know the specifics. (which is not a
good thing - any code can cause issue on a different platform)

Most developers are using Linux/Ubuntu. There, you can simply build
OpenSCAD. As many devs do just that, we also maintain the scripts for
various Ubuntu versions (the ones happening to be used by developers) and
can guess when there is an issue is with a newer or older version, what
could be the cause. The more you leave current or LTS Ubuntu standard, the
more issues a user and/or developer will face. Either-way: The typical
OpenSCAD developer uses mainly Ubuntu, but a significant number of users
only use Windows.

A release would mean to seriously go trough compatibility issue of various
operating systems.
Guess how hard and painful that work is.



--
Sent from: http://forum.openscad.org/

[Troberg]

I get what you say, and I don't disagree with it, as such.

However, the official release is 3 years old. Many important improvements
(for me, customization and several bug fixes) has been made. At some point,
it makes sense to "wrap up the package".

There also is one thing I've learned as a developer: Doing a release will
expose many errors. Regular releases helps keeping things tidy. Exaclty what
constitutes regular, one might discuss, but I think one is due now.

I haven't been using the official version since I found customizations. Many
others also use non-official versions. The once who do use the official
version frequently gets told to update to other versions when things don't
work. Linux packet managers have way outdated versions. All these are also
signs that it's time to "wrap up the package".

I get your reasons. I really do. But, sometimes, one still needs to do a
release.



--
Sent from: http://forum.openscad.org/

[Torsten Paul]

I'm pretty sure everyone agrees that a new release is overdue.
That does not change the fact that there is obviously nobody
at this point who has the time to drive the efforts to do it.
What I sometimes try telling people in commercial projects for
slightly different reasons "Releases don't happen by wishing"
applies here too ;-).

I'm happy that there is still progress regardless of that,
maybe things are a bit slow lately, but OpenSCAD is far from
dead as can be seen in the github activities thanks to quite
a number of people.

Many others also use non-official versions.

However, I do want to point out that the snapshot versions are
very much official versions too, just not official releases.
Having those ready for easy installation on as much platforms
and systems as possible with limited resources is IMHO one of
the plus points of OpenSCAD. We try to not just do the "oh,
there's the source code, you should compile it yourself".

Linux packet managers have way outdated versions.

True, and I don't like that at all, but that is nothing we can
control. They have the option to use snapshots too, but I do
understand the reasons distros do that only in rare cases.
Latest Ubuntu does not ship with OpenSCAD at all, which is even
worse. We provide 3 ways to get the snapshots running easily
on Ubuntu (well 2 until the AppImages are fixed).

ciao,
Torsten.

[nop head]

I would wager the latest snapshot has more features and less bugs that the
last release, so why not make it a release?

Post by Torsten Paul
I'm pretty sure everyone agrees that a new release is overdue.
That does not change the fact that there is obviously nobody
at this point who has the time to drive the efforts to do it.
What I sometimes try telling people in commercial projects for
slightly different reasons "Releases don't happen by wishing"
applies here too ;-).
I'm happy that there is still progress regardless of that,
maybe things are a bit slow lately, but OpenSCAD is far from
dead as can be seen in the github activities thanks to quite
a number of people.

Many others also use non-official versions.

However, I do want to point out that the snapshot versions are
very much official versions too, just not official releases.
Having those ready for easy installation on as much platforms
and systems as possible with limited resources is IMHO one of
the plus points of OpenSCAD. We try to not just do the "oh,
there's the source code, you should compile it yourself".
Linux packet managers have way outdated versions.
True, and I don't like that at all, but that is nothing we can
control. They have the option to use snapshots too, but I do
understand the reasons distros do that only in rare cases.
Latest Ubuntu does not ship with OpenSCAD at all, which is even
worse. We provide 3 ways to get the snapshots running easily
on Ubuntu (well 2 until the AppImages are fixed).
ciao,
Torsten.
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[Troberg]

Post by nop head
I would wager the latest snapshot has more features and less bugs that the
last release, so why not make it a release?

Well, just like when working with a car, you need to make sure you have all
the parts in and all the nuts and bolts tightened before you let it out on
the road.

Same thing here, a lot of work being done, and it must be verified that
everything is in place and works together as intended.



--
Sent from: http://forum.openscad.org/

[nop head]

Ideally yes, but when the last release was three years ago and everybody is
using snapshots it seems we are waiting for this release to be perfect when
it is already substantially better than the last one.

Post by Troberg

Post by nop head
I would wager the latest snapshot has more features and less bugs that

the

Post by nop head
last release, so why not make it a release?

Well, just like when working with a car, you need to make sure you have all
the parts in and all the nuts and bolts tightened before you let it out on
the road.
Same thing here, a lot of work being done, and it must be verified that
everything is in place and works together as intended.
--
Sent from: http://forum.openscad.org/
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[adam purdie]

It would be very cool to have a new version, i'd be curious if the
community could help the maintainers release it?

Post by nop head
Ideally yes, but when the last release was three years ago and
everybody is using snapshots it seems we are waiting for this
release to be perfect when it is already substantially better than
the last one.>

Post by Troberg

Post by nop head
I would wager the latest snapshot has more features and less bugs
that the>> > last release, so why not make it a release?

Well, just like when working with a car, you need to make sure you
have all>> the parts in and all the nuts and bolts tightened before you let it
out on>> the road.
Same thing here, a lot of work being done, and it must be
verified that>> everything is in place and works together as intended.
--
Sent from: http://forum.openscad.org/
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org> _________________________________________________

OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[Clint Goss]

I will put in a vote, a nudge, a sentiment on behalf of the (probably)
large number of users who, like myself, make heavy use of OpenSCAD but are
not in the developer community. We download it, thank our stars that it
exists, scratch our head a bit about the 2015.03 stamp, and put it to use.

Somewhere down the road, after lots of code (11,000 lines, in my case) and
results (the open-source www.BreathFlute.com, in my case), we begin to
wonder what we might be missing. Are there useful features out there? Are
there lurking bugs that have been fixed? Is there a reasonably accessible
list of these items since 2015.03??

Yes, I do realize how daunting an official release can be, especially if it
has not been done in 3+ years.

However, I'm guessing that the developers may not always be aware of the
*value* of OpenSCAD to a wide and largely un-heard user community out there
...

-- Clint Goss, Ph.D.

Goss.com <http://www.goss.com/> ... index of all our web sites

Post by nop head
Ideally yes, but when the last release was three years ago and everybody
is using snapshots it seems we are waiting for this release to be perfect
when it is already substantially better than the last one.

Post by Troberg

Post by nop head
I would wager the latest snapshot has more features and less bugs that

the

Post by nop head
last release, so why not make it a release?

Well, just like when working with a car, you need to make sure you have all
the parts in and all the nuts and bolts tightened before you let it out on
the road.
Same thing here, a lot of work being done, and it must be verified that
everything is in place and works together as intended.
--
Sent from: http://forum.openscad.org/
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[Troberg]

Post by Clint Goss
However, I'm guessing that the developers may not always be aware of the
*value* of OpenSCAD to a wide and largely un-heard user community out there

It's extremely useful to me, and will be even more useful in the near
future, as I'm moving to a new hous and will need to do extensive
remodelling and build some new furniture. Ordinary CAD doesn't work for my
programmer brain. I have a huge gratitude debt to the devs.

Sadly, my experience in programming lies elsewhere (large administrative and
technical database systems), so I can't help much, except sharing a few
useful OpenSCAD code snippets. I also have quite a few designs which I can
use as a test bench for new versions.

I'm a hobby user, but if I were a professional user (which I am with several
other software products), I would find a 3 years old version worrying. For a
professional user, investing time (and thus money) into a product is a risk,
and you must consider stuff like "What if it stops working with an operating
system upgrade and no one fixes it? How will I deliver on my contracts?".
I'd also suspect that it'll affect sponsoring, such as Google Summer of
Code.

I think there are many benefits to "tying up loose strings" and do a
release, and I'd do what I can to help, even if it's just testing. I think
others are prepared to help as well.



--
Sent from: http://forum.openscad.org/

[Alex Gibson]

Also very willing to help as a tester.

I'm actively using OpenSCAD, but I'm stuck on the last stable release
because I fear a crash or bug taking out my fairly complex designs more than
I fear missing out on the new features... but some of them sound good!

Is there any element here of the 'best' being the enemy of the 'good'?
There's been so much great work done in 3 years it must be daunting to
consider regression testing all of it on all 3 platforms...
I wonder if there is any consensus among the devs for a smaller number of
new/fixed features that they agree are a) an improvement, and b) stable?
Then, it might be worth doing a 'beta release' that the community can then
help to bug-test?

Big thanks to all the awesome work being done by the devs.

Cheers,

Alex Gibson

admg consulting

edumaker limited

. Project management
. Operations & Process improvement
. 3D Printing

-----Original Message-----
From: Discuss [mailto:discuss-***@lists.openscad.org] On Behalf Of
Troberg
Sent: 21 August 2018 09:25
To: ***@lists.openscad.org
Subject: Re: [OpenSCAD] Why is the 2015.03 kept as the "official" release?

Post by Clint Goss
However, I'm guessing that the developers may not always be aware of the
*value* of OpenSCAD to a wide and largely un-heard user community out there

It's extremely useful to me, and will be even more useful in the near
future, as I'm moving to a new hous and will need to do extensive
remodelling and build some new furniture. Ordinary CAD doesn't work for my
programmer brain. I have a huge gratitude debt to the devs.

Sadly, my experience in programming lies elsewhere (large administrative and
technical database systems), so I can't help much, except sharing a few
useful OpenSCAD code snippets. I also have quite a few designs which I can
use as a test bench for new versions.

I'm a hobby user, but if I were a professional user (which I am with several
other software products), I would find a 3 years old version worrying. For a
professional user, investing time (and thus money) into a product is a risk,
and you must consider stuff like "What if it stops working with an operating
system upgrade and no one fixes it? How will I deliver on my contracts?".
I'd also suspect that it'll affect sponsoring, such as Google Summer of
Code.

I think there are many benefits to "tying up loose strings" and do a
release, and I'd do what I can to help, even if it's just testing. I think
others are prepared to help as well.



--
Sent from: http://forum.openscad.org/

[Troberg]

Post by Alex Gibson
I'm actively using OpenSCAD, but I'm stuck on the last stable release
because I fear a crash or bug taking out my fairly complex designs more than
I fear missing out on the new features... but some of them sound good!

Don't worry about that. Just backup your scripts, and you can always go back
to them.

Post by Alex Gibson
Is there any element here of the 'best' being the enemy of the 'good'?
There's been so much great work done in 3 years it must be daunting to
consider regression testing all of it on all 3 platforms...
I wonder if there is any consensus among the devs for a smaller number of
new/fixed features that they agree are a) an improvement, and b) stable?
Then, it might be worth doing a 'beta release' that the community can then
help to bug-test?

Well, there might be a workable route. Simply take the current version and
promote it to release candidate. Let people try it for a while, and if
nothing drastic comes up in a few months, promote it to a proper version.

I've been working with the customizations version for a long time now, and
I've yet to find a single way in which it is inferior to the 2015.03
version. In other words, I have good confidence that you could do as
suggested above and get something that's better than the current official
version.



--
Sent from: http://forum.openscad.org/

[Clint Goss]

Simply take the current version and promote it to release candidate.

That's got my vote!

Uh ... might there be a list someplace that summarizes the updates from
2015.03 to the new RC? (has a nice right to it ... "the new RC").

-- Clint Goss, Ph.D.

Goss.com <http://www.goss.com/> ... index of all our web sites

Post by Alex Gibson
I'm actively using OpenSCAD, but I'm stuck on the last stable release
because I fear a crash or bug taking out my fairly complex designs more than
I fear missing out on the new features... but some of them sound good!

Don't worry about that. Just backup your scripts, and you can always go back
to them.

Post by Alex Gibson
Is there any element here of the 'best' being the enemy of the 'good'?
There's been so much great work done in 3 years it must be daunting to
consider regression testing all of it on all 3 platforms...
I wonder if there is any consensus among the devs for a smaller number

of

Post by Alex Gibson
new/fixed features that they agree are a) an improvement, and b) stable?
Then, it might be worth doing a 'beta release' that the community can

then

Post by Alex Gibson
help to bug-test?

Well, there might be a workable route. Simply take the current version and
promote it to release candidate. Let people try it for a while, and if
nothing drastic comes up in a few months, promote it to a proper version.
I've been working with the customizations version for a long time now, and
I've yet to find a single way in which it is inferior to the 2015.03
version. In other words, I have good confidence that you could do as
suggested above and get something that's better than the current official
version.
--
Sent from: http://forum.openscad.org/
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[adam purdie]

Simply take the current version and promote it to release candidate.

Might be an over-simplification, i don't know what the release process
entails and there may be specific things that the mods/owners/etc need
to happen before cutting an RC.
Looking at the backlog of issues (500+) and PR's (70+) i'm thinking that
i'd be running screaming!

Simply take the current version and promote it to release candidate.> That's got my vote!

Uh ... might there be a list someplace that summarizes the updates
from 2015.03 to the new RC? (has a nice right to it ... "the new RC").>
-- Clint Goss, Ph.D.
Goss.com[1] ... index of all our web sites
On Tue, Aug 21, 2018 at 8:31 AM, Troberg

Post by Alex Gibson
I'm actively using OpenSCAD, but I'm stuck on the last stable release
because I fear a crash or bug taking out my fairly complex designs
more than I fear missing out on the new features... but some of
them sound good!

Don't worry about that. Just backup your scripts, and you can always
go back>> to them.

Post by Alex Gibson
Is there any element here of the 'best' being the enemy of the 'good'?
There's been so much great work done in 3 years it must be
daunting to consider regression testing all of it on all 3
platforms... I wonder if there is any consensus among the devs for
a smaller number of new/fixed features that they agree are a) an
improvement, and b) stable? Then, it might be worth doing a 'beta
release' that the community can then help to bug-test?

Well, there might be a workable route. Simply take the current
version and>> promote it to release candidate. Let people try it for a while,
and if>> nothing drastic comes up in a few months, promote it to a proper
version.>>
I've been working with the customizations version for a long time
now, and>> I've yet to find a single way in which it is inferior to the 2015.03>> version. In other words, I have good confidence that you could do as>> suggested above and get something that's better than the current
official>> version.
--
Sent from: http://forum.openscad.org/
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org>

_________________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

Links:

1. http://www.goss.com/

[Marius Kintel]

I believe all non-experimental features are stable enough for a release.
It would be really nice to get some of the experimental ones ready though, but those things take time and we’s (as always) short on developers, especially the kind who’re willing to do the boring work related to testing, documentation and release work :/

Cutting an intermediate release could buy us some time though.
Are there any experimental features that people feel absolutely need to make it into such a release?

-Marius

[Troberg]

Post by Marius Kintel
Cutting an intermediate release could buy us some time though.

It's also not good to dump too many new things on the users at once, at
least not if they mean changes to how things works.

Post by Marius Kintel
Are there any experimental features that people feel absolutely need to
make it into such a release?

I don't know if it's experimental or not, but customization is a big one for
me. I use it in every single design I do. I live and die by parametric
design.



--
Sent from: http://forum.openscad.org/

[nop head]

I live and die by parametric design.

So do I but I have never tried the customiser. I prefer all my parameters
to be set in the source code under source control, not a GUI. OpenSCAD is
not a GUI based CAD system.

I don't see why assert, echo, lc-each, etc, are still experimental.

Post by Marius Kintel
Cutting an intermediate release could buy us some time though.

It's also not good to dump too many new things on the users at once, at
least not if they mean changes to how things works.

Post by Marius Kintel
Are there any experimental features that people feel absolutely need to
make it into such a release?

I don't know if it's experimental or not, but customization is a big one for
me. I use it in every single design I do. I live and die by parametric
design.
--
Sent from: http://forum.openscad.org/
_______________________________________________
OpenSCAD mailing list
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[Marius Kintel]

Post by nop head
I don't see why assert, echo, lc-each, etc, are still experimental.

There are tickets open for all these; I believe it’s mostly about missing documentation, examples etc.

-Marius

[Troberg]

Post by nop head
So do I but I have never tried the customiser. I prefer all my parameters
to be set in the source code under source control, not a GUI. OpenSCAD is
not a GUI based CAD system.

Well, my designs are often based on body measurements. I don't want to make
another source version just to set parameters for another person.

It's one thing if you just want to tweak around to get good values, but if
you actually want to make something customizable, it's useful.

Likewise, I have designs that bend or move, and I don't want to have to
change the source to go through the full range of motion to check that there
is no collisions. Just moving a slider is much more practical.

But, that's a bit off topic. I'm happy to discuss it, but it should probably
be done in a new thread.



--
Sent from: http://forum.openscad.org/