@Gavin,

Yes, OpenERP was only installable in prod by serious serious geeks so far (not 
only computer geeks, but also ERP specialists, good consultants, so aliens in a 
word), that's why there are so few I think. Oh yes, there are plenty of those 
guys say in the aerospace/banking fields. But the point is that since it 
targets the mid markets, integrators can hardly attract them. 
But overall it already makes an ever growing few success stories and a ever 
growing skilled community.

Imagine guys skilled like this one http://www.chricar.at/ChriCar/index.html 
working at night and week ends to have its integration business sustainable. It 
was not easy, that's sure. People have to realize how difficult it is, even 
that's true, done with the right guys it can be several times cheaper than high 
end proprietary ERP's.

I've been partner of Openbravo one year when I was taking care of the tech 
survey at Smile, and in my opinion, yes, it's just worst. I should say I've 
seen quite a few guys trying both and sticking to OpenERP but never the 
reverse. But those are few stories because ERP's are so complex than learning 
two is so hard. I've also know some Openbravo partners that invested all into 
it and just died without even be able to try OpenERP, end of the story (one 
example, second French partner; dead //forge.openbravo.com/users/palmtree-it )

IMHO, Openbravo is much more professional with all the marketing, the release 
cycle, the work planing, the partnership and many things. Thanks to their fund 
raise, they hired experienced guys Tiny did not have. At some point they are 
organized just like a non open source company (still they release their work 
publicly).

BUT, the flip side, is that it costs them money, so very easily you would have 
to start paying way more to get any support.
All those non coding guys live with your money: 
http://www.openbravo.com/about-us/board-directors/
(and no, they haven't a such large real customer basis that it's negligible)

My opinion is that by 2008 they were never competitive against SAP BO/ Sage 
X3/MS Navision/Divalto in France in the 2 years term (and decision makers 
hardly think about later). Luckily for them, it might change soon, I think they 
improved a bit (they actually listen to the suggestion I gave them with making 
a REST bridge, invested on minimal modularity) and also outsourced to India 
were they could finally get the ton of polishing work on their obsolete 
expensive tech that was required. So I think in 2010 they just start being 
competitive against some proprietary ERP's in rich countries too; finally.

The fundamental difference between Openbravo and OpenERP is pretty much a whole 
different technology at its roots. I won't take time to try to convince you 
here: but Openbravo is a ton (millions of lines) of non OOP PS/SQL lurking 
inside XML CDATA sections, at best verbose Java (and consider the Java language 
being abandoned slowly now for high level business code, even more now with the 
Sun collapse; don't get me wrong, the Java platform on his side with the new 
language running on it will have a bright future. But hey, OpenERP  close to 
running Jython smoothly, so OpenERP wins again here).

Openbravo modularity is really childish compared to OpenERP: no way to add 
screen extension points, no component oriented UI (a lot more clicks), method 
granularity is almost always several hundreds lines of PS/SQL: if a module want 
to override something: he hardly can call super, he should copy/paste/change a 
whole PS/SQL procedure. Now they have a way to package procedures and screen 
together in a module, but those will be hardly compatible. Really far from the 
real tight integration you can achieve with OpenERP modules.

unlike OpenERP, Openbravo has no BPM engine to abstract the workflow logic: 
it's all if/else hardcoded instead.

Openbravo doesn't comes close to the memory session new gen osv_memory OpenERP 
wizards to offer a modular OOP decent UI (even if further improvement would be 
welcome).

Because they don't use an ORM in their core (Hibernate is used by less than 5% 
of the logic, only the most recent stuff), they don't have caching and the hit 
the SGBD much more heavily than OpenERP, their performance suck totally (just 
install it and you will understand).

Their development cycles are really orders of magnitude slower compared to 
OpenERP: their screen are generated static HTML taking dozens of minutes to 
generate and then you have to deploy start your Java server which is really a 
slow startup compared to OpenERP...

Because of those fundamental differences in the platform, being a successful 
Openbravo third party integrator is so much harder (nearly impossible, find me 
just one that his not just an self bought Openbravo showcase). Remember 
Openbravo raised nearly M 18$ and invested heavily in marketing. They say now 
that want to fund raise again ("future financing rounds"  
http://www.openbravo.com/about-us/press-room/news/139/bob-mann-joins-openbravo-as-new-chairman-of-the-board.php
 ). OpenERP is their side made it from scratch due to their own real value 
only. Until 2008, they had a very poor marketing, people hard not here due to 
their marketing (at least the real integrators).

as a result, Openbravo has a very little experimented community, few success 
stories. A very small scope compared to OpenERP (a third of it at best: CRM or 
project is a joke in Openbravo). By mid 2008, I can tell you that Openbrao did 
had way more bugs than OpenERP: doing a simple order/invoice/expedition cycle 
was almost impossible on PostgreSQL (after the marketing it was already the 
leading open source ERP since 2007)...

So, to sum it up: yes OpenERP is really hardcore. But don't fool yourself, 
Openbravo is even harder (and certainly not better). Given the OpenERP (+Tryton 
now) competition, I don't even see the point with Openbravo even if it start 
being better than BO and similar proprietary products (don't get me wrong 
OpenbravoPOS on the contrary is a very decent product OpenERP does not have, 
but hey you could just connect it the same way).

I think OpenERP will improve now that Tiny is getting bigger with the right 
guys being set up. In any case, it has always improved steadily since the 
beginning. If it weren't to improve or were to do a false move, that's not too 
risky either, I guarantee you they are too many involved integrators, there 
would be a successful organized fork (or even Tryon would pass them).
Now it's hard for everybody, including Tiny, so let's be patient and understand 
all the implications of being really open source (upfront investment is really 
hard; generic international ERP is fucking complex to get right), so as long as 
they do their job decently, that's all right, they still do better than any of 
us individually.

We also now have guys taking the test approach pretty seriously with such 
initiative https://launchpad.net/oerpscenario , even some Tiny framework unit 
test were contributed recently... So hopefully out of the box quality pass an 
acceptable threshold soon, meanwhile skilled integrators will close the gap.

------------------------
Raphaël Valyi

CEO and OpenERP consultant at
http://www.akretion.com




-------------------- m2f --------------------

--
http://www.openobject.com/forum/viewtopic.php?p=50451#50451

-------------------- m2f --------------------


_______________________________________________
Tinyerp-users mailing list
http://tiny.be/mailman2/listinfo/tinyerp-users

Reply via email to