Persevere 1.0

[Kris Zyp]

Persevere release candidate has been out for a while now, and as far as
I can tell there are no blocker issues with a 1.0 release. Are there any
issues that anyone believes should be a 1.0 blocker? Of course there are
some issues that could be fixed in 1.x/1.0.x releases (or 2.0 for larger
enhancements).
Thanks,
Kris

[Neville Burnell]

Hi Kris,

Does 1.0 support Lucene FTI now?

I couldnt find any docs on this feature :(

Kind Regards

Neville

[Kris Zyp]

Neville,
No, Lucene fulltext is implemented in 2.0, I do not plan to backport it
to 1.0. My apologies that the 2.0 codebase isn't documented or quite yet
for really using yet, hopefully I will have it usable with some docs soon.
Thanks,
Kris

[Dean Landolt]

The only big missing piece that I'd really like to see as 1.0 is better JSGI coreApp support -- meaning, I'd like to be able to change a lot more of persevere's functionality via coreApp. As it stands much of the internals are pulling from the servlet request instead of coreApp so middleware can only really intercepted requests, they can't do anything to alter an inbound request before it hits persevere (unless you wrap the servlet request, but ugh)...

I've finally got myself set up with eclipse debugging so I'll try get some patches together to add the above feature. Unless 2.0 is close enough to ready for primetime in the near future I'll work on getting this functionality into 1.x. This isn't really a core feature and shouldn't block a 1.0 final release, so +1 for a release. Awesome job Kris...

Dean

[Paul Spaulding]

There are features I'd like to see, but I have not run into any issues
with Persevere that would prevent a 1.0 release. Persevere is
amazing.

On Nov 3, 10:01 am, Dean Landolt <d...@deanlandolt.com> wrote:
> The only big missing piece that I'd really like to see as 1.0 is better JSGI
> coreApp support -- meaning, I'd like to be able to change a lot more of
> persevere's functionality via coreApp. As it stands much of the internals
> are pulling from the servlet request instead of coreApp so middleware can
> only really intercepted requests, they can't do anything to alter an inbound
> request before it hits persevere (unless you wrap the servlet request, but
> ugh)...
>
> I've finally got myself set up with eclipse debugging so I'll try get some
> patches together to add the above feature. Unless 2.0 is close enough to
> ready for primetime in the near future I'll work on getting this
> functionality into 1.x. This isn't really a core feature and shouldn't block
> a 1.0 final release, so +1 for a release. Awesome job Kris...
>
> Dean
>

[dstarke]

We have run into a couple of bugs that are pretty serious and occur in
the release candidate:

The first is a concurrency issue that sometimes occurs when committing
a set of changes across several objects that have references to each
other. We have been able to work around this one by collapsing most
of our object model into one deep object, and forcing a number of
other activities to occur serially.

The second bug that is causing us some major problems has to do with
access control checks on objects during queries. In some situations
when we issue a query to Persevere, we are seeing access control
checks on objects that are not part of the query results. This
appears to happen on queries with result sets above a certain size and
that have a sort order specified. In our system, a user does not
necessarily have access to all objects of a given type, so the access
control check can fail for one object, causing the entire query to
fail, even though that user does have access to every object that
should have been returned by the query.

I would submit both of these issues as potential 1.0 blockers,
particularly the latter one as we have not yet been able to come up
with a workaround for it that doesn't force us to compromise our
security requirements.

-- David

[Kris Zyp]

I've uploaded the latest 1.0 branch nightly to the downloads page, so
Persevere 1.0 is now available for download.

Thanks,
Kris

[Dave Jackson]

Excellent! Thanks Kris and congratulations! We all appreciate your
hard work producing such an amazing product. It is certainly going to
revolutionize the way I prototype on my day job, and I have several
side projects that have been waiting for something like this to come
along for production deployments as well.

Keep up the great work!