Hi Esteban,

Sure, I'm aware of those and other solutions and I've even used few of them. I'm sorry I can't explain myself adequately now, but I'm dreaming about some other kind of persistence, the one that (just like GemStone) doesn't make me to rebuild my application to migrate from image-based to a more granular solution. Or better, use the one from the very beginning if I want. Smalltalk monitors changes in source code (though it still does it in a bit coarse way� I mean storing final results instead of deltas etc.). We have great metaprogramming and reflection capabilities. We can hook quite deep into message passing. And we can patch our VM as a last resort. Why don't we still have some of the persistence features GemStone has, and have them as a loadable framework? Even for the price of lower performance?�

Perhaps, I'm a dreamer (sure I am) and don't understand some basic obstacles for that, but can someone explain what are they?

--

Best regards,


Dennis Schetinin



2013/4/8 Esteban Lorenzano <estebanlm@gmail.com>
Hi,�

for smaller projects you have:�

- the image it self :)
- ... and I'm sure there are others :)

Esteban

On Apr 6, 2013, at 4:40 PM, Dennis Schetinin <chaetal@gmail.com> wrote:

+1�

BTW, I think that's a shame for smalltalkers to have so many libs for external persistence solutions. That's not a paradox. Alternatives are good, but Smalltalk was the first environment that had built-in persistence as much as 35, if not 40 years ago. Is there much progress since that time? I wouldn't say so. GemStone is great, but it's monolithic. And pricey. And that's one more Smalltalk in our �Balkans.�

Why Pharo or Squeak don't have a framework (based on Fuel for example) that allows to control changes, execute transaction, save some parts of image instead of whole image etc.? This could be less efficient then GemStone, but loadable on demand, customizable and usable in smaller projects. Maybe there's such a project living already? Or is it too much for modern Smalltalk?�

It's not complaining. I know the answer: there's no usable persistence solution because I didn't create it :) �I just want to know if someone else thinks the same way as I do? Consider that a very first step towards starting such a project, if it is feasible at all�


--

Best regards,


Dennis Schetinin



2013/4/6 stephane ducasse <stephane.ducasse@free.fr>


Hi
2013/4/1 Victor Stan <victor.stan@gmail.com>
Hi,�

Are there any open source OODB's out there? Alternatives to GemStone?

I think there is no alternavise to GemStone.

I hope not :)
Gemstone is good but an ecosystem with only one solution would look rather limited to me.�

Because GemStome is not just data store. It is applicaton server where you work in multi user image system.
Imagine single Pharo image which can work under different parallel VM instances (maybe running on different machines) which have concurrent access to same image objects, can commit it changes (save same image).