openfoam-extend-general

[Openfoam-extend-general] Development process of OpenFOAM-1.5-dev

[Rostislav S. <ros...@gm...>]

Hi together

first of all I'd like to ask you if this mailing list is the right way
to communicate with you? Seems like noone uses it.

Now about the development process - I just downloaded r1388 read
    Core/OpenFOAM-1.5-dev/README
    Core/OpenFOAM-1.5-dev/buildInstructions.txt
and tried
$ Core/OpenFOAM-1.5-dev/Allwmake

the result is >100 errors. (the log is not attached it has > 330kB)
Seems like at first you just squash all the changes for a new version
to the repo and then some magic happens and... ok, whatever...

Here some questions and points
- which is the last compilable OpenFOAM-1.5-dev revision?
- Core/OpenFOAM-1.5-dev/buildInstructions.txt obviously does not
belong to Core/OpenFOAM-1.5-dev. It's taken from OpenFOAM-1.4.1-dev.
Wrong build instructions might sound better than no instructions for
small projects but not for anything size of the openfoam - 800
Millions LOC
- keeping README and README.html containing the same is either just a
redundant effort or yet another source of errors - don't you think so?
(the same applies to the ReleaseNotes*)

Bost

Re: [Openfoam-extend-general] Development process of OpenFOAM-1.5-dev

[Håkan N. <ha...@ch...>]

Dear Rostislav,

Yes, that file seems to be related to the 1.4.1-dev version. It is up to
Hrv to keep it, delete it, or update it. Most slightly experienced
OpenFOAM users realize this and simply compile the software as it is it
described at openfoam.org, i.e. 'the usual way'.

It is in fact quite easy to install the dev-version. You just get the
revision you want from openfoam-extend and type ./Allwmake in the root of
that version. This is similar to what you do if you compile all of the
vanilla version. Note that it requires that you have the third-party
products correctly set up first - you do that as described at
openfoam.org. In your case I think that you did not follow the
instructions at openfoam.org for setting up the environment and installing
the third-party products. Since you didn't provide any error messages it
is difficult to say more than that.

Note that you are dealing with a development line, meaning that some
revisions might give you error messages while compling. As long as you
will not use the libraries those error messages are related to, you don't
have to care about this.

Please don't be so upset about some minor mistakes, and instead see how
much you get for free from lots of people who didn't get anything back
from you yet. I'm sure that you can set up a support contract with some
developer to keep all the README files typo-error-free, but for most
developers it is in fact the code itself that has the highest priority.

Best regards,
Håkan.

> Hi together
>
> first of all I'd like to ask you if this mailing list is the right way
> to communicate with you? Seems like noone uses it.
>
> Now about the development process - I just downloaded r1388 read
>     Core/OpenFOAM-1.5-dev/README
>     Core/OpenFOAM-1.5-dev/buildInstructions.txt
> and tried
> $ Core/OpenFOAM-1.5-dev/Allwmake
>
> the result is >100 errors. (the log is not attached it has > 330kB)
> Seems like at first you just squash all the changes for a new version
> to the repo and then some magic happens and... ok, whatever...
>
> Here some questions and points
> - which is the last compilable OpenFOAM-1.5-dev revision?
> - Core/OpenFOAM-1.5-dev/buildInstructions.txt obviously does not
> belong to Core/OpenFOAM-1.5-dev. It's taken from OpenFOAM-1.4.1-dev.
> Wrong build instructions might sound better than no instructions for
> small projects but not for anything size of the openfoam - 800
> Millions LOC
> - keeping README and README.html containing the same is either just a
> redundant effort or yet another source of errors - don't you think so?
> (the same applies to the ReleaseNotes*)
>
> Bost
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> 30-Day
> trial. Simplify your report design, integration and deployment - and focus
> on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now.  http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Openfoam-extend-general mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openfoam-extend-general
>

Re: [Openfoam-extend-general] Development process of OpenFOAM-1.5-dev

[Rostislav S. <ros...@gm...>]

Hi together

2009/8/8 Håkan Nilsson <ha...@ch...>:
> It is in fact quite easy to install the dev-version. You just
[..]
> Most slightly experienced OpenFOAM users realize this and [that]
[..]
> don't be so upset about some minor mistakes, and instead see how
> much you get for free from lots of people who didn't get anything back
> from you yet.

I have a friend of mine he is a brilliant mechanical engineer with
unfortunately no idea what is a 'swap partition', 'code repository' or
'preprocessor directive'. Your doctor does not blame you for knowing
how to say knee in Latin. So don't blame a mechanical engineer for a
lack of software engineering skills. Frankly, if you go to doctor you
get a service worth of thousands of Euro sometimes, although you payed
just a fraction of it. And if the doc makes a mistake you're about to
... well not a word about being thankful for how much you got for
free.

I can tell you, that friend of mine has spent last 6 month on learning
'linux and all the stuff around', on learning something he is not
interested in! I repeat: He is a mechanical not software engineer. And
if I track back his problems: Most of the time waste is caused by -
from your point of view - minor mistakes.

Would _you_ be thankful for that?

> Note that you are dealing with a development line, meaning that some
> revisions might give you error messages while compling.

??? Do you realize implications of your statement? If anyone breaks it
no one else can compile it! Isn't it _the_ emergency situation to be
avoided at any cost?!?

> I'm sure that you can set up a support contract with some
> developer to keep all the README files typo-error-free, but for most
> developers it is in fact the code itself that has the highest priority.

Hey, hey, hold on! What do you think how many people out-there in the
wild have enough software engineering know-how and linux skills for
this job. Besides, linux is still too exotic and difficult even for
more than a 90% of IT guys (based on ongoing everyday experience) And
what do you think how much do you have to pay to a such guy? And what
do you think who would be interested in such a work if there is plenty
of more interesting jobs at hand?

[..]
> see how much you get for free from lots of people who didn't get anything back
> from you yet.

How can you claim that? Aren't you using *that* linux where everyone
works for free? And what would you say about people which are never
gonna contribute to your project? People working on KDE, Gnome,
Kernel, GCC, Firefox, Ubuntu, Debian, Amarok, Vim, Emacs etc.? And
what about all the people busy with introducing linux and especially
the OpenFOAM to the others? You think it's a 2 minutes work with no
impact on you?

Ok, next time I'll be a bit more constructive & productive

Bost

Re: [Openfoam-extend-general] Development process of OpenFOAM-1.5-dev

[Bernhard G. <Ber...@ic...>]

Hi all!

Let's please not escalate this any further. I feel a flame-war coming
on. 

At first to the initial mail. The problem with compiling is, that most
of the time if one thing goes wrong it results in a lot of error. You
not posting the whole log here is greatly appreciated. But it would be
of great help if you gave an indication where the first error occured
(plus the text of the first error).

To my knowledge the current version in the repository of 1.5-dev
compiles alright.

I guess the buildInstructions were not changed because the basic
process did not change apart from the version numbers

README in two formats is something that originates in the
upstream-project. 

>>>>> On Sun, 9 Aug 2009 12:46:24 +0200
>>>>> "RS" == Rostislav Svoboda <ros...@gm...> wrote:

    RS> Hi together 2009/8/8 Håkan Nilsson <ha...@ch...>:
    >> It is in fact quite easy to install the dev-version. You just
    RS> [..]
    >> Most slightly experienced OpenFOAM users realize this and
    >> [that]
    RS> [..]
    >> don't be so upset about some minor mistakes, and instead see
    >> how much you get for free from lots of people who didn't get
    >> anything back from you yet.

    RS> I have a friend of mine he is a brilliant mechanical engineer
    RS> with unfortunately no idea what is a 'swap partition', 'code
    RS> repository' or 'preprocessor directive'. Your doctor does not
    RS> blame you for knowing how to say knee in Latin. So don't blame
    RS> a mechanical engineer for a lack of software engineering
    RS> skills. Frankly, if you go to doctor you get a service worth
    RS> of thousands of Euro sometimes, although you payed just a
    RS> fraction of it. And if the doc makes a mistake you're about to
    RS> ... well not a word about being thankful for how much you got
    RS> for free.

More on the doctor below.

    RS> I can tell you, that friend of mine has spent last 6 month on
    RS> learning 'linux and all the stuff around', on learning
    RS> something he is not interested in! I repeat: He is a
    RS> mechanical not software engineer. And if I track back his
    RS> problems: Most of the time waste is caused by - from your
    RS> point of view - minor mistakes.

    RS> Would _you_ be thankful for that?

I see your point that nobody can be a specialist for everything, but
the fact currently is that with OF it is quite good if you have a
doctor in the house. But at least you have to know how to work with a
first aid kit.

    >> Note that you are dealing with a development line, meaning that
    >> some revisions might give you error messages while compling.

    RS> ??? Do you realize implications of your statement? If anyone
    RS> breaks it no one else can compile it! Isn't it _the_ emergency
    RS> situation to be avoided at any cost?!?

Hakan was talking about more subtle problems: solvers behaving
slightly different etc. Usually the dev compiles.

And using a version from a SVN/GIT/CVS usually is risky. If you don't
need the special stuff there it is better to stick to the releases

    >> I'm sure that you can set up a support contract with some
    >> developer to keep all the README files typo-error-free, but for
    >> most developers it is in fact the code itself that has the
    >> highest priority.

    RS> Hey, hey, hold on! What do you think how many people out-there
    RS> in the wild have enough software engineering know-how and
    RS> linux skills for this job. Besides, linux is still too exotic
    RS> and difficult even for more than a 90% of IT guys (based on
    RS> ongoing everyday experience) And what do you think how much do
    RS> you have to pay to a such guy? And what do you think who would
    RS> be interested in such a work if there is plenty of more
    RS> interesting jobs at hand?

    RS> [..]
    >> see how much you get for free from lots of people who didn't
    >> get anything back from you yet.

    RS> How can you claim that? Aren't you using *that* linux where
    RS> everyone works for free? And what would you say about people
    RS> which are never gonna contribute to your project? People
    RS> working on KDE, Gnome, Kernel, GCC, Firefox, Ubuntu, Debian,
    RS> Amarok, Vim, Emacs etc.? And what about all the people busy
    RS> with introducing linux and especially the OpenFOAM to the
    RS> others? You think it's a 2 minutes work with no impact on you?

"THAT Linux where everyone works for free" Is that meant ironic? There
is a lot of work done by people "for free" and they're doing valuable
stuff, but the reality is that most of the projects you mentioned have
a core of programmers THAT get paid for doing that stuff by some
company because they have strategic interest (IBM and others pay
people to contribute to Linux, Google founded Mozilla, just some
examples) or direct interest (Redhat ...) 

Nobody expects you to contribute to OF as much as you do to other
projects, but you've got to be aware that most of the people who
provide help on the forum have day jobs where they have to answer to
the people who foot their bills (their customers, their students)
about giving away support for free while they are paying for it (don't
get me wrong: these people are remarkably cool about it, but they want
to see results for their projects too). Or to get back to your "pay a
few bucks to the doctor for services worth several thousands": I guess
you were never in the unlucky situation to live in a country without
proper health insurance OR needing services that were not covered by
your health insurance. And the only reason why you have to pay your
doctor so little is because there is some kind of system in place that
make sure that this highly qualified professional can earn a
living. 

    RS> Ok, next time I'll be a bit more constructive & productive

    RS> Bost

Bernhard

    RS> ------------------------------------------------------------------------------
    RS> Let Crystal Reports handle the reporting - Free Crystal
    RS> Reports 2008 30-Day trial. Simplify your report design,
    RS> integration and deployment - and focus on what you do best,
    RS> core application coding. Discover what's new with Crystal
    RS> Reports now.  http://p.sf.net/sfu/bobj-july
    RS> _______________________________________________
    RS> Openfoam-extend-general mailing list
    RS> Ope...@li...
    RS> https://lists.sourceforge.net/lists/listinfo/openfoam-extend-general


-- 
---------------------------------------------------------------------------
DI Bernhard F.W. Gschaider
---------------------------------------------------------------------------
EMail: 	Ber...@ic...
WWW  : www.ice-sf.at
Jabber : bgs...@ja...
Tel:	+43(3842)98282-42	Fax:	+43(3842)98282-02
---------------------------------------------------------------------------

Re: [Openfoam-extend-general] Development process of OpenFOAM-1.5-dev

[Rostislav S. <ros...@gm...>]

On Mon, Aug 10, 2009 at 20:36, Bernhard
Gschaider<Ber...@ic...> wrote:
> Let's please not escalate this any further.

Yea, good idea. I promissed you to be constructive this time. Ok, let's go:

> At first to the initial mail. The problem with compiling is,

I managed to compile everything on a 32-bit Ubuntu 9.04. The key were
missing ThirdParty packages and flex.
However the ThirdParty packages were not recompiled. I took them
directly from OF 1.5.

Now we're trying to compile everything from scratch including
ThirdParty packages from OF 1.5. under 64-bit Ubuntu 8.04. It's gonna
take some time...

Anyway I squashed everything including the ThirdParty packages from OF
1.5 to my personal git repo. IMO ThirdParty packages belong to the
repo, you cannot compile the OF without them. I intend to publish my
repo on github if you don't give me any better location or other
instructions what to do with it.

> README in two formats is something that originates in the
> upstream-project.

I took a look at the README and buildInstructions.txt - there's a lot
of information in these files. I think it makes sence to mark them
clearly as "IN PROGRESS" and correct them later on step by step.

Bost

Re: [Openfoam-extend-general] Development process of OpenFOAM-1.5-dev

[Bernhard G. <Ber...@ic...>]

>>>>> On Mon, 10 Aug 2009 23:13:23 +0200
>>>>> "RS" == Rostislav Svoboda <ros...@gm...> wrote:

    RS> On Mon, Aug 10, 2009 at 20:36, Bernhard
    RS> Gschaider<Ber...@ic...> wrote:
    >> Let's please not escalate this any further.

    RS> Yea, good idea. I promissed you to be constructive this
    RS> time. Ok, let's go:

    >> At first to the initial mail. The problem with compiling is,

    RS> I managed to compile everything on a 32-bit Ubuntu 9.04. The
    RS> key were missing ThirdParty packages and flex.  However the
    RS> ThirdParty packages were not recompiled. I took them directly
    RS> from OF 1.5.

    RS> Now we're trying to compile everything from scratch including
    RS> ThirdParty packages from OF 1.5. under 64-bit Ubuntu
    RS> 8.04. It's gonna take some time...

    RS> Anyway I squashed everything including the ThirdParty packages
    RS> from OF 1.5 to my personal git repo. IMO ThirdParty packages
    RS> belong to the repo, you cannot compile the OF without them. I
    RS> intend to publish my repo on github if you don't give me any
    RS> better location or other instructions what to do with it.

Actually there is such a thing on the SVN:

https://openfoam-extend.svn.sourceforge.net/svnroot/openfoam-extend/tags/OpenCFD_Release/ThirdParty-1.5/

It looks like an empty directory, but if uses external-references and
pulls in a complete ThirdParty-directory (without paraview) if you do
a "svn checkout" on it. The problem is that the "git svn"-commands
don't support externals

    >> README in two formats is something that originates in the
    >> upstream-project.

    RS> I took a look at the README and buildInstructions.txt -
    RS> there's a lot of information in these files. I think it makes
    RS> sence to mark them clearly as "IN PROGRESS" and correct them
    RS> later on step by step.

Can't comment on those as it is a long time that I read them. The
build process is always similar with OpenFOAM. I know that this isn't
a satisfactory answer for those who do it the first time. 

Bernhard

-- 
---------------------------------------------------------------------------
DI Bernhard F.W. Gschaider
---------------------------------------------------------------------------
EMail: 	Ber...@ic...
WWW  : www.ice-sf.at
Jabber : bgs...@ja...
Tel:	+43(3842)98282-42	Fax:	+43(3842)98282-02
---------------------------------------------------------------------------