Empathy List Archives

LH

Les Hall

Sun, Dec 21, 2014 11:48 AM

That statement is going on your permanent record, young man!

Of course software crashes. This is free software. You got it for zero dollars, so don’t complain when it crashes.

I have known people with this attitude. They are not the doers of the world, they are the skeptics, the nay-sayers who mock and taunt and jeer and criticize every little imperfection in the work of others when they have not ever accomplished anything themselves. Don’t go down that path, go the other way - be someone who strives to achieve great things, because in the end you’ll look back on your life and feel much better about it.

Les

On Dec 21, 2014, at 5:39 AM, Bananapeel lunatica.xiaoyu@gmail.com wrote:

Software should never crash, regardless of misuse.

--
View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10679.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

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

That statement is going on your permanent record, young man! Of course software crashes. This is free software. You got it for zero dollars, so don’t complain when it crashes. I have known people with this attitude. They are not the doers of the world, they are the skeptics, the nay-sayers who mock and taunt and jeer and criticize every little imperfection in the work of others when they have not ever accomplished anything themselves. Don’t go down that path, go the other way - be someone who strives to achieve great things, because in the end you’ll look back on your life and feel much better about it. Les > On Dec 21, 2014, at 5:39 AM, Bananapeel <lunatica.xiaoyu@gmail.com> wrote: > > Software should never crash, regardless of misuse. > > > > -- > View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10679.html > Sent from the OpenSCAD mailing list archive at Nabble.com. > > _______________________________________________ > OpenSCAD mailing list > Discuss@lists.openscad.org > http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

AC

Alan Cox

Sun, Dec 21, 2014 12:13 PM

On Sun, 21 Dec 2014 05:48:00 -0600
Les Hall thingsofrings@icloud.com wrote:

On Dec 21, 2014, at 5:39 AM, Bananapeel lunatica.xiaoyu@gmail.com wrote:

Software should never crash, regardless of misuse.
That statement is going on your permanent record, young man!

Of course software crashes.

<rant>It shouldn't, and until the world stops saying "it always breaks" it
won't get fixed 8)</rant>

In most cases with OpenSCAD the problem is not OpenSCAD but the libraries
it uses. They have some very tight input requirements (or they break),
some significant performance problems (mathematicians write code for
correctness not performance), and very demanding rules about numeric
behaviour which mix in with this.

If you want OpenSCAD to fly and be solid in all cases write an
engineering not maths CSG library. One that tolerates small input
errors, deals with mathematical inaccuracy by whatever armwaving fuzzing
or grid snapping is needed and is prepared to be fast rather than
perfectly accurate. We can't print atoms (yet) so why worry about that
level of accuracy.

It's free software, you have all the pieces, send fixes!

Alan

On Sun, 21 Dec 2014 05:48:00 -0600 Les Hall <thingsofrings@icloud.com> wrote: > On Dec 21, 2014, at 5:39 AM, Bananapeel <lunatica.xiaoyu@gmail.com> wrote: > > Software should never crash, regardless of misuse. > That statement is going on your permanent record, young man! > > Of course software crashes. <rant>It shouldn't, and until the world stops saying "it always breaks" it won't get fixed 8)</rant> In most cases with OpenSCAD the problem is not OpenSCAD but the libraries it uses. They have some very tight input requirements (or they break), some significant performance problems (mathematicians write code for correctness not performance), and very demanding rules about numeric behaviour which mix in with this. If you want OpenSCAD to fly and be solid in all cases write an engineering not maths CSG library. One that tolerates small input errors, deals with mathematical inaccuracy by whatever armwaving fuzzing or grid snapping is needed and is prepared to be fast rather than perfectly accurate. We can't print atoms (yet) so why worry about that level of accuracy. It's free software, you have all the pieces, send fixes! Alan

S

Scalpel78

Sun, Dec 21, 2014 12:24 PM

Les Hall wrote

I’ve noticed there are two kinds of computer users: those who try to make
software work and those who try to break it and then complain that it
broke. Which one are you?

(...)

Of course software crashes. This is free software. You got it for zero
dollars, so don’t complain when it crashes.

Les

Don't get me wrong, I'm not complaining. Not at all. I'm just giving
positive feedback and engaging in a discussion. Being a software developer I
know that bugs are to be expected. But if no one says anything when they
happen then chances are they are not going to be fixed either.

Your comments could be interpreted like this; "OpenSCAD will only crash if
you do stupid things". If you look at the issue-database in GitHub I'm sure
you'll find more than enough examples where the application crashes even
though the user has written reasonable code.

You asked what kind of user I am; I am the one that wants OpenSCAD to
succeed. I want it to become the defacto standard CAD software for
programmers. For that to happen there needs to be a low threshold for
providing feedback in the community. If we set a standard where we're saying
that only experienced users can partake then we're headed in the wrong
direction. Let's not be a community where we scoff at "stupid" questions.

I appreciate the dicussion, Les :)

--
View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10686.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Les Hall wrote > I’ve noticed there are two kinds of computer users: those who try to make > software work and those who try to break it and then complain that it > broke. Which one are you? > > (...) > > Of course software crashes. This is free software. You got it for zero > dollars, so don’t complain when it crashes. > > Les Don't get me wrong, I'm not complaining. Not at all. I'm just giving positive feedback and engaging in a discussion. Being a software developer I know that bugs are to be expected. But if no one says anything when they happen then chances are they are not going to be fixed either. Your comments could be interpreted like this; "OpenSCAD will only crash if you do stupid things". If you look at the issue-database in GitHub I'm sure you'll find more than enough examples where the application crashes even though the user has written reasonable code. You asked what kind of user I am; I am the one that wants OpenSCAD to succeed. I want it to become the defacto standard CAD software for programmers. For that to happen there needs to be a low threshold for providing feedback in the community. If we set a standard where we're saying that only experienced users can partake then we're headed in the wrong direction. Let's not be a community where we scoff at "stupid" questions. I appreciate the dicussion, Les :) -- View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10686.html Sent from the OpenSCAD mailing list archive at Nabble.com.

B

Bananapeel

Sun, Dec 21, 2014 12:59 PM

Reporting a crash in a software isn't "mocking and taunting," it's helping to
make the software better.

--
View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10689.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

Reporting a crash in a software isn't "mocking and taunting," it's helping to make the software better. -- View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10689.html Sent from the OpenSCAD mailing list archive at Nabble.com.

DE

David Eccles (gringer)

Sun, Dec 21, 2014 7:34 PM

A crash due to a segmentation fault can occasionally be deliberately and
maliciously manipulated to not crash and excecute arbitrary code as the
user. If it's possible to produce an error prior to a crash (and, ideally,
avoid the crash entirely), then it is a good idea to do that.

Yes, bugs happen, all the time. One of my common phrases in emails when I
write code is something along the lines of "I have changed this code, so
expect odd behaviour and bugs, and please let me know when that happens."
That doesn't mean that known bugs (especially ones that halt a program)
should be ignored.

It doesn't look like Scalpel78 was particularly annoyed by this bug, and was
in an excellent position to report this bug because he found a [somewhat]
repeatable situation that caused OpenSCAD to crash. Bug reporting,
particularly with included examples (as is easy with OpenSCAD), is a good
idea, and I don't like discouraging people from doing that.

--
View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10691.html
Sent from the OpenSCAD mailing list archive at Nabble.com.

A crash due to a segmentation fault can occasionally be deliberately and maliciously manipulated to not crash and excecute arbitrary code as the user. If it's possible to produce an error prior to a crash (and, ideally, avoid the crash entirely), then it is a good idea to do that. Yes, bugs happen, all the time. One of my common phrases in emails when I write code is something along the lines of "I have changed this code, so expect odd behaviour and bugs, and please let me know when that happens." That doesn't mean that known bugs (especially ones that halt a program) should be ignored. It doesn't look like Scalpel78 was particularly annoyed by this bug, and was in an excellent position to report this bug because he found a [somewhat] repeatable situation that caused OpenSCAD to crash. Bug reporting, particularly with included examples (as is easy with OpenSCAD), is a good idea, and I don't like discouraging people from doing that. -- View this message in context: http://forum.openscad.org/Crash-when-using-minkowski-on-my-object-tp10668p10691.html Sent from the OpenSCAD mailing list archive at Nabble.com.

MK

Marius Kintel

Sun, Dec 21, 2014 7:39 PM

On Dec 20, 2014, at 15:38 PM, Scalpel78 frode@lillerud.no wrote:

Hi, I made an object, and then wanted to see what happened if I did a
Minkowski sum of the object with a small sphere.

I can reproduce this on Mac - something goes wrong deep inside CGAL it seems. Usually, CGAL will throw an exception which can catch, and sometimes CGAL crashes hard. I’ll try to look into it and open a github issue.

-Marius

On Dec 20, 2014, at 15:38 PM, Scalpel78 <frode@lillerud.no> wrote: > Hi, I made an object, and then wanted to see what happened if I did a > Minkowski sum of the object with a small sphere. I can reproduce this on Mac - something goes wrong deep inside CGAL it seems. Usually, CGAL will throw an exception which can catch, and sometimes CGAL crashes hard. I’ll try to look into it and open a github issue. -Marius

MK

Marius Kintel

Sun, Dec 21, 2014 7:52 PM

See https://github.com/openscad/openscad/issues/1097

-Marius

See https://github.com/openscad/openscad/issues/1097 -Marius

discuss@lists.openscad.org

Crash when using minkowski on my object