dolibarr-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Dolibarr-dev] [Dolibarr-association] Dolibarr 4.0.1


From: Olivier Geffroy
Subject: Re: [Dolibarr-dev] [Dolibarr-association] Dolibarr 4.0.1
Date: Thu, 20 Oct 2016 11:08:00 +0200

It's depends of the final user 

let's says for a company (who use dolibarr) and make less than 100K per month and don't have a lot of externals modules, 2 update per year is easy

for a big company 1 update per year is enough and with dolibarr isn't a problem to stay in 3.8 and to migrate in 4.0 and squeeze the 3.9 (for example)

my 2 cents

2016-10-20 10:58 GMT+02:00 Laurent Destailleur (aka Eldy) <address@hidden>:
I don't understand. You say "If the major release issued every 6 months was free of bug, stable and did not require another install/update after barely one month to correct the most glaring bugs that will not be dramatic"

Every experimented developper know that this argument is the best argument to ask to have more release than 2 per year. And you ask less. So why using an argue to ask more release: The more is the delay between 2 versions, the more is the bug rate on production (that's why more and more project are increasing the release frequency) and difficulty to have a stable version is an exponential of the number of feature added or modified. So your argue is just incomprehensible.

I used on production each version, as soon as it is release and announced and I have no problem. Also the stability of a version depends on bugs fixes during the beta period and number of unit tests added when added new future. Developers must work on this direction instead of an "against productive" idea.








2016-10-19 21:34 GMT+02:00 Charles Benke <address@hidden>:

OK

 

If I follow your argumentation … I will deliver a brand new version of all my modules each week, because I have decide to planned like this

Even if the version is not enough tested, even the previous release have some know bug, even if the document are not upgraded …

And I will explain to my disgruntled customers that this is a good method to make a better quality and simplify their upgrade ...

 

Release a version every 6 months because FOR YOU is more simple is not acceptable. I do not develop modules dolibarr because it is easy but because it allows users to better manage their company, create growth, the emploies ...

 

If the major release issued every 6 months was free of bug, stable and did not require another install/update after barely one month to correct the most glaring bugs that will not be dramatic

 

The minimum straightforwardness that we can have with users downloading a new major release is to explain that this version DO NOT BE USED IN PRODUCTION.

 

 

Bien cordialement,

Charlie Benke

 

De : Dolibarr-dev [mailto:dolibarr-dev-bounces+charles.fr=address@hidden] De la part de Laurent Destailleur (aka Eldy)
Envoyé : mercredi 19 octobre 2016 17:34
À : Posts about Dolibarr ERP & CRM development and coding <address@hidden>

Objet : Re: [Dolibarr-dev] [Dolibarr-association] Dolibarr 4.0.1

 

Your argue is not coherent.

You say you want less version so you have to test your module less often. It also meas your customer upgrade version less often.

 

So why just don't you make your tests every 2 versions. Result will be same. You will work only every 1 year instead of every 6 month, and your customer would be able to upgrade only every 1 year (once your module is validated for the version) instead of every 6 month.

 

It's just your choice and the choice of your customer.

 

Having a release every 1 year, means nor integrator, nor users have choice. Also it means a lower quality and exponentiel work to make upgrade.

But if you prefer to upgrade your module once per year, just do it. You can, it's just a choice you must do. It is not because there is a new version, that you must upgrade your module. If you prefer to follow a 1 year release, just follow this rythm and ask you customer to follow also this rythm. The only difference is that the ryhtm is defined by you instead of being imposed be a dolibarr low release rythm.

And i think it is better to let integrator to decide their release/upgrade frequency then having this decied/forced by Dolibarr.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2016-10-19 16:49 GMT+02:00 Charles Benke <address@hidden>:

Actually I maintain 22 modules, some are simple, some are complex. To test all of them correctly (use all feature, modify doc, …) each time a new major version of Dolibarr is release is more than 2 full weeks long for Romain an me...

During the month a new version comes out, sales of modules on dolistore are halved cut (according to my information it is not related to my modules only).

 

I could do as some others … , just change the version number and wait for my clients put bugs me but I do not find it honest

 

Most integrators with whom I work no longer wish to upgrade versions as there are no major advances between two versions either-called major

The final version of each major costs money and energy to NOTHING: just to show that development teams are able to release two versions per year, two versions full of vacuum .

 

We have all been waiting for new accountancy module for 2 years. The time spent to release a new version will have better been employed to work on this strategic module…

 

 

Bien cordialement,

Charlie Benke

 

De : Dolibarr-dev [mailto:dolibarr-dev-bounces+charles.fr=address@hidden] De la part de Developpement | Open-DSI
Envoyé : mercredi 19 octobre 2016 16:24
À : Posts about Dolibarr ERP & CRM development and coding <address@hidden>
Cc : address@hiddeng
Objet : Re: [Dolibarr-dev] [Dolibarr-association] Dolibarr 4.0.1

 

Hi

Thanks to Camille for pointing the main problem : Module and ratio time spend / bug / patch
As integrator of Dolibarr, it's not "sustainable" for me to test every six month Dolibarr and the modules I'm commonly using. Today I only install 3.9. Maybe next year, I will uprade to 5.0 or not... depending of what functions will be added or remaining experimental.
Modules are too often broken by new version. On the Dolistore you can see module labeled 3.x-4.0 who are in fact broken with the last version or doesn't exist for the current version of Dolibarr. I think it's not good for the reputation of Dolibarr.
I'll be pleased to discuss about this subject in Valence :-)

Regards
Philippe Scoffoni - Open-DSI




Le 19/10/2016 à 15:14, address@hidden a écrit :

Hi

 

Thanks for sharing this.

I agree, Dolibarr migration is pretty nice !

 

but only core part, modules looks more problematic to update.

 

Regarding communication, this is a work in progress.

 

Yes I saw this :) But looks again difficult. But it's better :)

 

From now on, we'll have systematic annoucement when a major version is released, minor version too, why not. A communication group has been started within the fundation with the goal to better communicate with the community. We already are present on social medias, but this dev mailing-list and the dolistore customers are 2 audiences we poorly communicate with (not to say not at all).

 

I don't understand logic, dolibarr users/community are on forum, mailinglist but piority is social network, strange

 

About your concerns around PRs and plugins, I'm sorry you feel that way. PRs are usually correctly integrated and not lost.

 

Maybe now, I'll try again. But I'm not sure. My fear is to lost again energy to nothing.

 

Plugins are the responsibility of their developers. Personnaly, our plugins are upgraded with the new releases

 

I'm not module developper then I don't know if is complicate or not to follow release and provide. As user, i prefer to have my own script and don't use module. In my use case ratio time spend / bug / patch is too heavy.

Thanks a lot

 

km



_______________________________________________
Dolibarr-dev mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/dolibarr-dev

 


_______________________________________________
Dolibarr-dev mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/dolibarr-dev



 

--

------------------------------------------------------------------------------------

Google+: https://plus.google.com/+LaurentDestailleur-Open-Source-Expert/
Facebook: 
https://www.facebook.com/Destailleur.Laurent

------------------------------------------------------------------------------------

* Dolibarr (Project leader): http://www.dolibarr.org (make a donation for Dolibarr project via Paypal: address@hidden)

* AWStats (Author) : http://awstats.sourceforge.net (make a donation for AWStats project via Paypal: address@hidden)

* AWBot (Author) : http://awbot.sourceforge.net

* CVSChangeLogBuilder (Author) : http://cvschangelogb.sourceforge.net

 

 


_______________________________________________
Dolibarr-dev mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/dolibarr-dev




--
------------------------------------------------------------------------------------
Google+: https://plus.google.com/+LaurentDestailleur-Open-Source-Expert/
Facebook: https://www.facebook.com/Destailleur.Laurent
------------------------------------------------------------------------------------
* Dolibarr (Project leader): http://www.dolibarr.org (make a donation for Dolibarr project via Paypal: address@hidden)
* AWStats (Author) : http://awstats.sourceforge.net (make a donation for AWStats project via Paypal: address@hidden)
* AWBot (Author) : http://awbot.sourceforge.net
* CVSChangeLogBuilder (Author) : http://cvschangelogb.sourceforge.net



_______________________________________________
Dolibarr-dev mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/dolibarr-dev




--
Merci d'avance a tous ceux qui vont partager la vidéo dans ma signature ^^
Olivier Geffroy
Consultant Informatique

-------------------------------------
Jeffinfo SARL
29 rue de la Gare 59320 Ennetieres en Weppes
address@hidden
Gsm : 0608632740
Skype : darkj3ff


reply via email to

[Prev in Thread] Current Thread [Next in Thread]