[Math] Backward compatibility

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

[Math] Backward compatibility

Gilles Sadowski
Hi.

Is it OK to perform non backward-compatible changes in code that wasn't
offcially released yet (i.e. in code tagged with the "@since 2.2" javadoc
comment)?


Best regards,
Gilles

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Backward compatibility

James Carman
Is it part of a public API that went out in a release already?  What
method do you have in mind?

On Tue, Jun 29, 2010 at 7:12 PM, Gilles Sadowski
<[hidden email]> wrote:

> Hi.
>
> Is it OK to perform non backward-compatible changes in code that wasn't
> offcially released yet (i.e. in code tagged with the "@since 2.2" javadoc
> comment)?
>
>
> Best regards,
> Gilles
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Backward compatibility

Gilles Sadowski

> Is it part of a public API that went out in a release already? [...]
>
> > Is it OK to perform non backward-compatible changes in code that wasn't
> > offcially released yet (i.e. in code tagged with the "@since 2.2" javadoc
> > comment)?

Not in an official release since 2.2 isn't out yet.

> What
> method do you have in mind?

A candidate would be the constructor of
"TricubicSplineInterpolatingFunction" (in package "analysis.interpolation")
that currently throws a checked "DimensionMismatchException" but should
throw an unchecked "DimensionMismatchException" (from the new "exception"
package).


Gilles

P.S. I'd like to perform similar changes (precondition violation generates
     an unchecked exception) in the whole source code but for everything
     released in 2.1 and before, that'll have to wait for the developement
     of version 3.0. :-{

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Backward compatibility

James Carman
On Tue, Jun 29, 2010 at 7:46 PM, Gilles Sadowski
<[hidden email]> wrote:
> Not in an official release since 2.2 isn't out yet.
>

Well, since the code hasn't been released (it's a new
method/constructor I presume?), I would think that anything new that's
in SVN should be considered a work-in-progress.  So, I don't think it
should be a big deal.

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Backward compatibility

Phil Steitz
James Carman wrote:
> On Tue, Jun 29, 2010 at 7:46 PM, Gilles Sadowski
> <[hidden email]> wrote:
>> Not in an official release since 2.2 isn't out yet.
>>
>
> Well, since the code hasn't been released (it's a new
> method/constructor I presume?), I would think that anything new that's
> in SVN should be considered a work-in-progress.  So, I don't think it
> should be a big deal.

+!  Unreleased code is subject to change, including deletion or
incompatible API change.

Phil
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Backward compatibility

James Carman
On Tue, Jun 29, 2010 at 9:18 PM, Phil Steitz <[hidden email]> wrote:
> +!  Unreleased code is subject to change, including deletion or
> incompatible API change.
>

What he said!

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] Backward compatibility

Gilles Sadowski
In reply to this post by Phil Steitz
Hi.

> +!  Unreleased code is subject to change, including deletion or
> incompatible API change.

There is a "checkOrder" method in "util.MathUtils" throwing an
"IllegalArgumentException". I'll create the more specific
  NotIncreasingSequenceException
  NotStrictlyIncreasingSequenceException
that are instances of "IllegalArgumentException". Is it OK
(backward-compatible) that "checkOrder" throws these instead?


Gilles

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]