On 3 April 2012 00:38, Karl Wright <daddy...@gmail.com> wrote: > "Whether it is necessary is another matter, and is not easy to answer > in general as "it depends"." > > I'm going to treat this as "unnecessary", because we were extremely > careful to maintain backwards compatibility when writing the changes.
The API may be backwards compatible, but the behaviour presumably isn't, otherwise there would be no need to apply the patch. [The physical API of an AA battery is identical for Zinc-carbon, NiCd and NiMh, and they are often interchangeable, but their behaviour is not identical.] Can you be certain that the non-standard jars will not interfere with the standard jars? > > Karl > > On Mon, Apr 2, 2012 at 6:59 PM, sebb <seb...@gmail.com> wrote: >> On 2 April 2012 22:45, Karl Wright <daddy...@gmail.com> wrote: >>> I think the bigger picture here is that resolving differences of this >>> kind requires time, and during the interim we need to be able to >>> release software anyway. I have no wish to maintain a forked copy of >>> either xerces or httpclient indefinitely, but right at the moment we >>> are not prepared or able to use the released versions of these >>> libraries. I hope that we can all agree that pragmatism has a value; >>> we're certainly not trying to step on people's toes, we're simply >>> trying to solve a problem. >> >> One sure way to solve the problem is to use your own package namespace >> for any such imported source. >> >> That is definitely sufficient. >> >> Whether it is necessary is another matter, and is not easy to answer >> in general as "it depends". >> >>> Thanks, >>> Karl >>> >>> >>> On Mon, Apr 2, 2012 at 5:28 PM, Karl Wright <daddy...@gmail.com> wrote: >>>> "what's the use-case for non-well-formed >>>> XML?' This thread is probably not the best place to delve further in >>>> that direction." >>>> >>>> Simple use case: parsing malformed RSS feeds. I agree, though, we're >>>> getting into the weeds here; I'd love to have this discussion >>>> elsewhere, but the point does stand that none of these patch decisions >>>> was done lightly. Hopefully the community can accept that. >>>> >>>> Karl >>>> >>>> On Mon, Apr 2, 2012 at 4:54 PM, Benson Margulies <bimargul...@gmail.com> >>>> wrote: >>>>> Karl, >>>>> >>>>> I'm exceedingly sorry here that the IPMC as a whole let you down by >>>>> not turning into these issues and dealing with them at the outset. >>>>> There's been a lot of sensitivity expressed lately to Apache projects >>>>> stepping on each other's toes. >>>>> >>>>> Personally, I have no objection to including mutant jars in a -deps >>>>> binary with a clear explanation of what they are, but I would like to >>>>> see some support for that view, because I'd could imagine some >>>>> objections based on recent email. >>>>> >>>>> In the longer term, if ManifoldCF really wants to include an "XML >>>>> parser with a difference", then it's certainly possible for you to >>>>> maintain and release a fork of Xerces under your own package names. I >>>>> agree with Sebb that you would be well advised to find some other way >>>>> around it. My personal reaction, in complete isolation from the >>>>> problem at hand, was 'really? what's the use-case for non-well-formed >>>>> XML?' This thread is probably not the best place to delve further in >>>>> that direction. >>>>> >>>>> --benson --------------------------------------------------------------------- To unsubscribe, e-mail: general-unsubscr...@incubator.apache.org For additional commands, e-mail: general-h...@incubator.apache.org
