Re: Use ecomseo on demo rather than ecommerce

[Jacques Le Roux]

Hi Taher,

Le 25/01/2017 à 08:03, Taher Alkhateeb a écrit :
Hi Jacques,
I think you might be mixing up "multiple webapps" with "duplicate /
entangled webapps" in your examples. Yes we probably have other things to
get rid of, but I think we should avoid working in the direction of
increasing the entanglements instead of fixing root causes.

Sincerely I was really proud of myself when I thought about and implemented this solution (then temporary in my mind). When Anil kinda imposed a veto 
on ecomseo and I had no time to prove that we could still use the contents way with ecomseo.

I still see it as something very clean which allows to easily compare the 2 
solutions w/o any burden on the legacy.

Anyway, my intention was then indeed to compare and argument to replace the ecommerce webapp by ecomseo (which would then be renamed ecommerce). It 
seems we are ready for that, are we?


Also I don't see the point of bringing comments about a different
discussion and references to a fork of OFBiz to this discussion and rant
about it?

Human, I'm an human not a robot.

This remembers a joke we had with Andrew Sykes in the early days. A lot of the code we still rely on has been writing by Andrew Zenesky (you cleaned 
some of it). We were "wondering" if actually Andrew (Zenesky) was not actually a robot. Because of the speed he was writing code!

See? Human, "Community over code"[1][2] :)

Jacques
[1] http://communityovercode.com/
[2] http://theapacheway.com/


On Wed, Jan 25, 2017 at 9:38 AM, Jacques Le Roux <
jacques.le.r...@les7arts.com> wrote:

This argument looks more like a witch hunt to me.

If you both think it's a valid argument against ecomseo then what  about

accounting, ar and ap webapps in accounting component

ofbizsetup and ordermgr-js webapps in commonext component

content and contentimages in content component

marketing and sfa in marketing component

facility and catalog in product component

ical and workeffort in workeffort component

What are your plans for those? We might decide to change that but it's so
far a feature not a bug.

We dropped the ecomclone which was just showing how to reuse a webapp in
another webapp. I agreed about dropping it because ecomseo was also showing
the same feature. We can simply document it w/o showing an example in code.
I would not be against but we then need to document it in both readme and
wiki.

Now I can also agree about unifying the ecommerce component and then we
need to compare the 2 webapps which are very similar. Actually we "only"
need to compare the filters and servlets in both and decide which are the
best.

I think, but have not tested, that you can use content in ecomseo has it
was abruptly advocated by Hans[1] and Anil[2]. Else we can merge this
feature from ecommerce to ecomseo filters and servlet.

BTW, I don't see good practises in Hans's and Anil's answers and sorry to
say but I think the best answer then was from the regretted Adrian[3]

Since then ilscipio has its fork[4][5] and will not contribute anything
anymore, well done experts!

So you see there is some resentments about this.

Yes, people worked hard to contribute it with some other main features
like "Solr" OFBIZ-5042

But I think we can go ahead and find a common ground for the best or the
project, which is only what I have in mind.

Jacques
[1] http://ofbiz.135035.n4.nabble.com/Proposal-URL-Generation-Ch
anges-tp4639289p4639294.html
[2] https://issues.apache.org/jira/browse/OFBIZ-5312?focusedComm
entId=13939116&page=com.atlassian.jira.plugin.system.issueta
bpanels:comment-tabpanel#comment-13939116
[3] https://issues.apache.org/jira/browse/OFBIZ-5312?focusedComm
entId=13942316&page=com.atlassian.jira.plugin.system.issueta
bpanels:comment-tabpanel#comment-13942316
[4] https://en.wikipedia.org/wiki/Scipio_ERP
[5] http://www.scipioerp.com/


Le 24/01/2017 à 21:24, Taher Alkhateeb a écrit :

Small correction, I meant one component with two webapps

On Jan 24, 2017 11:21 PM, "Taher Alkhateeb" <slidingfilame...@gmail.com>
wrote:

I have to agree with Scott, the approach here is wrong in my opinion. My
definition of mess is simple: you should not have two components with two
webapps.  A real root solution is to unify the ecommerce component with
one
webapp exposed instead of having all this hairy code and
interdependencies.
I also think this should be discussed more rather than just committed
without a thorough review.

On Jan 24, 2017 11:12 PM, "Jacques Le Roux" <
jacques.le.r...@les7arts.com>
wrote:

Inline...
Le 24/09/2016 à 06:20, Scott Gray a écrit :

You're asking for opinions and I can't give one without any knowledge of
this feature so I'm simply telling you that the lack of documentation
is
a
hindrance to evaluation and adoption. I took a quick look at
SeoConfig.xml
and have no idea what most of it does.

You don't need to change SeoConfig.xml by default.
It seems to me that the comment there are not worse, and even better,
than you generally find in such config or properties files in OFBiz

Would you not like to have spiders bots continuously crawling it without
any risk?

I don't understand this question.

I guess you know that most of the spiders bots which are crawling the
web
are not doing for the good of websites and users. Better to prevent
them to
hurt in any way. OOTB the ecomseo webapp is better than the ecommerce on
this aspect. And it's also better with and for valuable crawlers
(Google,
DuckDuckGo, youNameIt...). Pour demo instances and OFBiz OOTB at large
would benefit from using only ecomseo.


Replacing ecommerce by ecomseo is another matter, another step I'd say.
It should have been the first step before any of this was committed but
instead you pushed ahead and committed it as an alternative despite
objections so now we have duplicated functionality and a messier
codebase.

Though I did not test it I don't think it prevents to use the content
component in anyway as feared Hans and Anil. I just put it besides to
end
the discussion then, having other stuff to do...

What makes you think we have a messier codebase?

I was going to list out the current set of possible SEO approaches that

now
exist in the codebase but it's all such an overlapping mess I can't
make
sense of it without spending too much time trying to figure it out.

Yes, I agree we need to document that better, starting from OFBIZ-5312
content

People need to stop adding things they can't get consensus on, IMO the

mess
that results is worse for the project than any missing feature.  I'll
never
understand why some committers are so desperate to see something
committed
that they'll push forward at any cost.  Not every feature has to start
it's
life in the OFBiz repo.

I wanted this committed because it's ecommerce webapp improved. I did
it
a way that did not hurt the legacy situation. It just offered an
alternative. For me it's a better solution.


And I'm surprised that you prefer to have "/control/" in each URL by
default :/

That is not the deciding factor in this discussion for me.  I'm against
any
change to the status quo until we get this mess cleaned up.

I'd like you to define what the mess is and how you would want to clean
it.

Thanks

Jacques


Regards
Scott


On 23 September 2016 at 20:34, Jacques Le Roux <
jacques.le.r...@les7arts.com

wrote:
Hi Scott,

Of course there are real world users, did you see the references?

Do we have a such document for ecommerce? No, so It's not a good
argument
to not switch the demo from ecommerce to ecomseo. Would you not like
to
have spiders bots continuously crawling it without any risk?

Replacing ecommerce by ecomseo is another matter, another step I'd
say.

And I'm surprised that you prefer to have "/control/" in each URL by
default :/

Compare
https://ofbiz-vm.apache.org:8443/ecomseo
with
https://ofbiz-vm.apache.org:8443/ecommerce/control/main

Jacques



Le 23/09/2016 à 01:15, Scott Gray a écrit :

Well for me it's a -1 until I hear some positive reviews from other

users
(and ideally committers).  I don't like having two ecommerce webapps
and
my
preference would be to merge the two into one, but I can't promote
that
idea without any group consensus that the SEO approach is good and
well
architected.  A document describing the architecture would make that
much
easier and I'm amazed one wasn't supplied with the proposal, no
wonder
it
sat there without much attention for so long.  But since one doesn't
exist
we'll just have to wait until people have time/care to review it
and/or
use
it and provide feedback.

Regards
Scott

On 23 September 2016 at 00:07, Jacques Le Roux <
jacques.le.r...@les7arts.com

wrote:

https://issues.apache.org/jira/browse/OFBIZ-5312

More athttps://issues.apache.org/jira/browse/OFBIZ-2214?jql=project
%20%3D%20OFBIZ%20AND%20text%20~%20%22seo%22

Jacques



Le 22/09/2016 à 13:25, Scott Gray a écrit :

By the way, is there any technical or user documentation for it?  I

haven't
looked at it and don't have time to review the actual
implementation
right
now.  The link you provided doesn't offer much more than a sales
pitch.

Regards
Scott

On 22 September 2016 at 23:22, Scott Gray<scott.gray@hotwaxsystems.
com>
wrote:

This mostly to somehow battle test it, even if I know it works well

So does it need battle testing or does it work well?  Have you

deployed
it
to any production instances?  Has anyone else?

Regards
Scott

On 19 September 2016 at 01:27, Jacques Le Roux <
jacques.le.r...@les7arts.com> wrote:

Hi,

Maybe you don't know about or did not try it, but we have ecomseo
a

clone
of the ecommerce component tailored for SEO

https://cwiki.apache.org/confluence/display/OFBIZ/Search+Eng
ine+Optimisation,+SEO+in+ecommerce

I propose to use it as the default ecommerce demo. This mostly to
somehow
battle test it, even if I know it works well.

As it's based on ecommerce, users should not experience a big
changes,
apart the changed URLs

What do you think?

Jacques