[Opensim-dev] Warning Safari (was Re: Current Status of OpenSim Git Master)

Mon Nov 23 07:32:14 UTC 2015

Hi,

On Sat, Nov 21, 2015 at 07:39:02PM +0100, Melanie wrote:
> OpenSim has been alpha code for as long as it exists. It will
> continue to be alpha code for a good long time to come. This has to
> do with liabilities and fitness for purpose claims. Those can't
> stand against alpha code.

You don't have to worry about that at all, as the BSD license applies,
which explicitly states:

"THIS SOFTWARE IS PROVIDED BY THE DEVELOPERS ``AS IS'' AND ANY EXPRESS OR
IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE CONTRIBUTORS BE LIABLE FOR ANY
DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
ADVISED OF THE POSSIBILITY OF SUCH DAMAGE."

> Therefore, unless you now have no further interest in helping,
> please do go ahead.
> 
> I don't understand why you would be hesitant about commenting unused
> variables after I and one other core member have already said it's
> ok to do so. We're not "reverting" those patches, but rather
> developers will simply have to uncomment not only the debug output
> but also the variable that feed them. That is OK and unless there is
> a major break in the code and we need everyone to collect logs with
> the debug output in it, those will rarely be committed again.

Ok, I shall convert the patch to do that, and remove the patches that 
take out commented out code. Will have to wait for the next weekend
though.

On Sat, Nov 21, 2015 at 03:04:25PM -0500, Michael Emory Cerquoni wrote:
> I think it also needs to be pointed out that fixing warnings in windows may
> be creating new warnings for Linux users, most of our developers are using
> Linux environment and developing with Mono not microsoft .net.  While
> things may appear to be improving for some, for other its making the
> situation worse potentially.  Just a heads up.  cross platform is hard and
> stuff!

I'm on linux as well (and as I read somewhere else, mono does a more thorough
job of checking the code [static code analysis] at compile time, so more warnings
to be expected).

Anyway, good point, I guess I need to fire up some windows installaten
somewhere and see if i can get it to compile on that too and try to make
both 

Is there a 'reference target platform' for both windows and linux? For
example, I know there's a buildbot somewhere - does it have build slaves
for both windows and linux or just one of the two? 

- Tom

> 
> - Melanie
> 
> On 21/11/2015 19:22, Tom Frost wrote:
> > Hi,
> > 
> > On Sat, Nov 21, 2015 at 04:42:00PM +0100, Melanie IMAP wrote:
> >> I gave you the OK for a warning safari but I believed you understood that this is alpha code.
> > 
> > I was under the impression that the avination merge was complete (as per
> > Diva's email from 16 november), so i was under the impression the code
> > for the avination features was done. I also thought the code was in use
> > in production in avination, so I assumed it would be of production
> > quality. Sorry I misunderstood.
> > 
> >> I would like for you to continue, but please take these things to heart:
> >> 
> >> - Unused variables, including those assigned to but not used, can only be commented, not removed. In many cases you will find that the values ganerated are used in debugging statements further down that have been commented!
> > 
> > I'm a little hesistant now to even comment them - as that would only
> > lead to a patch that will be largely reverted anyway.
> > 
> >> - Commented code needs to be left where it is. Also,  a warning safari is not a mandate to beautify code. No reordering and no whitespace corrections should be done.
> >> 
> >> Limited code cleanup or whitespace correction can be done outside the warning safari and after it's complete.
> >> 
> >> The issue here is that whitespace changes break pulling from git for downstream projects needlessly. So does code reordering, etc.
> > 
> > I shall wait until the new avination code is more mature, and leave
> > things alone for now. Please close the mantis issue for now. I'll lurk
> > on this list to see when things become more mature.
> > 
> >> Regarding the team, we are a multinational team and not all of us have English as their first language or America as their cultural origin. Please do feel free to approach people about perceived slights, but let's not make a big deal out of a wording. What counts is the spirit, not the letter.
> > 
> > My first language is not English and I don't have America as my cultural origin.
> > 
> > - Tom
> > 
> >> 
> >> - Melanie
> >> 
> >> On 21 Nov 2015, at 10:14, Tom Frost <tomfrost at linkwater.org> wrote:
> >> 
> >> Hi opensim devs,
> >> 
> >> > On Fri, Nov 20, 2015 at 11:50:09PM +0100, Tom Frost wrote:
> >> >> On Fri, Nov 20, 2015 at 02:33:53PM +0100, Melanie wrote:
> >> >> a warning safari is highly appreciated. You don't need to create a
> >> >> Mantis for each warning, simply create one for the warnings and
> >> >> attach the patches to it. For pure warning removal with no
> >> >> functional changes, the patches can be larger than what we usually
> >> >> ask for with functional changes.
> >> > 
> >> > Alright, i've gone ahead and created this issue:
> >> > 
> >> > http://opensimulator.org/mantis/view.php?id=7746
> >> > 
> >> > and attached 13 patches to address a first batch of low-hanging 
> >> > fruit. I thought it would be convenient for picking which ones
> >> > to apply.. But I can resubmit it as one large path if that
> >> > is preferred. Just let me know!
> >> 
> >> ---+ Unused variables and warnings
> >> 
> >> I got this response:
> >> 
> >> 'Hi, thanks but please stop for now :( You are also removing comments, dead code,
> >> etc, and that needs to be checked case by case, even some unused variables are
> >> there waiting future use You removed for example debug messages that we put back every time we
> >> need.  I can't apply patches like this automatically. I will use them as
> >> reference.  I did remove a few of those warnings, increasing the total reported by 2
> >> ;)´
> >> 
> >> Unused variables that are there waiting for future use and that
> >> generate warnings are in violation with the coding standards as stated
> >> on http://opensimulator.org/wiki/Coding_standards:
> >> 
> >> "Warnings
> >> 
> >>    Please keep the code warning free, using #pragma only if absolutely
> >> necessary. For instance ..."
> >> 
> >> Are all the core devs aligned in the acceptance of the practice of
> >> leaving in unused variables that generate warnings during compilation
> >> because someone at some point may want to use them?
> >> 
> >> If so, please rewrite the coding standards, so that new people like me
> >> don't waste their time.
> >> 
> >> ---+ Debug statements and commented-out code
> >> 
> >> I'm ok with redoing the patches to leave in large fragments of
> >> commented-out code and stray debug log statements that have been
> >> commented out (if indeed all core devs are fine with the ugliness this
> >> introduces into the codebase).
> >> 
> >> As for debug statements, wouldn't it be better to define them as such by
> >> not commenting them out but leaving them in and make them conditional on
> >> a variable that configures the debug level of logging?
> >> 
> >> ---+ Bad attitude
> >> 
> >> As a sidenote, telling someone to 'please stop for now :(' is not very
> >> productive. If I need to do things differently just say so, instead of
> >> using discouraging wording. From the comment above it feels like my work
> >> is not appreciated and i should not attempt to help out.
> >> 
> >> 
> >> 
> >> Please let me know how to proceed.
> >> 
> >> Thanks,
> >> 
> >> Tom
> >> _______________________________________________
> >> Opensim-dev mailing list
> >> Opensim-dev at opensimulator.org
> >> http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-dev
> >> _______________________________________________
> >> Opensim-dev mailing list
> >> Opensim-dev at opensimulator.org
> >> http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-dev
> >> 
> > _______________________________________________
> > Opensim-dev mailing list
> > Opensim-dev at opensimulator.org
> > http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-dev
> > 
> _______________________________________________
> Opensim-dev mailing list
> Opensim-dev at opensimulator.org
> http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-dev
> 