Erpnext 2017 - My Feedback about conference and free thoughts

Hi community,

I’ve attended at almost all conference from live stream and I would like to give some of my feedback.

First I want to congratulate organizations and all participants. Great event, great presentations and great developer sessions.

Here are my free thoughts:

There was in the conference some arguments about lack of contributions, what a community is, etc.
I totally understand @rmehta reaction when he see the filter solution in ERPNEXT (odoo has it for long time) and not shared. He’s feelings are completely fair - you get all of this and you have a interesting feature and not share?.
@rmehta is correct, this should be shared. But, frappe and erpnext are open source, you share if you want. It’s up with your conscience. @rmehta feels bad because he is the father of this community, i understand that but it is what it is, it’s all about their conscience.
I’m NOT actually use erpnext in my company, I’m slowly analise if it will be implemented in my construction company (very happy to see that there are plan to construction domain). But I’ve already contributed with several PRs. Things that I really need to implement erpnext in my company if it will happen. So every “free” time I have I code something and send it back to the community.
If I don’t use the product and already sent improvements, I think it’s fair that others should share too, but it is with their conscience, I shouldn’t be upset, it’s open source, it’s a community.
To be in open speech, I really like odoo. But why I prefer frappe and erpnext? Because I can contribute and make erpnext better, because you can help and have influence with roadmap, because what I need could be in main branch and not need to have and maintain a parallel branch with my “things”.
So please, send back some of your feature, will help all even the contributor (bug reporting, etc).

But I also think that erpnext should be more gently in comments of issues and PRs. I know what some of you feel “every day I have to explain the same”, “all PR’s lack something”, etc etc. Please understand that there are people that want to contribute and help and if they only receive a comment to their PR with the link of wiki for PR’s isn’t a great help. If there is a PR, there is a intention to help. Don’t close a PR without a convenient conversations. You don’t need to be that accurate, even when people is always asking for more features and not contribute.
Don’t close a issue just because you don’t understand the issue, and try to look “out the box”. I give you an example, there are several github issues and discuss about screen utilization in higher resolutions. When I see people saying “too much free space” I see more “How can we give a better use of this space” and not “How can I stretch everything to fill all screen”.
When someone opens a duplicated issue, close it, but give a comment and link to the other issue.
Once I’ve open a issue with the idea to create more calculations tests to guarantee higher standards of quality because there is always lots of bug reporting about bad tax calculations, rounding errors etc etc and IMO this is a MAJOR thing (I can be arrested for tax evasion and saying that is software will not help). The issue was closed some day after linked to a PR with some tests, have you account the number of bugs reported with calculations after my issue was closed? (I don’t but there are several). (please don’t answer me that I should write better tests, I’ve made some other contributions and here I’ve made one giving the ideia).

This kind of community is made of developers, integrators and users. Developers do code, integrators use and do code (some share others not), users use the product. Don’t get me wrong, but users are the principal of any opensource community. The more users you will see more developers and integrators contributing you have. And a User can contribute a lot with bug reporting (how important is that?) and feature asking (very helpful to create a roadmap), so please all frappe community don’t undervalue the importance of the users.

Someone said that the product is at level of dynamic/sage/odoo and we (even not using in production the product I fell the WE and not the YOU) should marketeer better the product. IMHO, I’ve used odoo and dynamics and eprnext is getting there but is not yet at their level for instance, at User Experience. Yes the features are there, but does the users like to use it? is it easy and intuitive? Frappe stack is what makes difference, its fantastic, but sometimes is the limitation (in UX etc) but we are here to make it grow and IMO frappe should have more work than erpnext.
At the end of the day every feature are there and possible to use, but is it a tool that helps the user to use?
Other example, like I’ve already referred these bugs on price, tax and other calculations, this couldn’t happen on main branch.

Someone say that integrators should have some kind of differentiation and “power”. Do you really like to go in odoo way? That’s why there are a new odoo fork called Flectra, because of their model.

After all this, remember I am not complaining and not trying to be a jerk. I like frappe, I contribute with ideas, bug reporting and PR’s, and i hope to be even more helpful in future, writing this is (in my mind) is a contribution.

Don’t get me wrong for this letter, but it isn’t needed to explain that I like frappe and erpnext and I am committed to this community, If not Why should I take my time to write this?

Paulo Almeida


@almeidapaulopt thanks for sharing your thoughts. I don’t think its a question of anyone being perfect, but that there needs to be a balance of contributions coming in from everyone. As a core dev, is my time better spent in fixing issues or working on improvements or answering the same question again and again. At what point does the cost of doing this out-weigh the benefit of potential new developers

Specially when we see most new developers looking to do their own thing. I am not sure this is sustainable until the community comes up with a better model.

1 Like