[Vtigercrm-developers] 2016 thoughts & reflections

Wed Dec 28 10:54:32 GMT 2016

I never thought you would publish Vtiger 7 code :] I must admit that you
surprised me. 

I don't know why but I was certain that people who had developed 6.0
version also worked later on 7.0, in particular, because Vtiger claimed
to have about 30 active developers. I reviewed the code from Vtiger 7.0
in detail and I can see that there aren't too many changes and most of
them are related to the interface, new look and feel and a few new
functionalities. 

In practical terms, there haven't been any new changes between 6.1 and
6.5 [maybe just multi-template support] and trac.vtiger.com and
code.vtiger.com are still being forgotten/ignored. Therefore, I thought
that Vtiger put much effort into Vtiger OnDemand. When I evaluate
version 7, I look at it from the perspective of 3 years because only
OnDemand version has been developed. If I revised it as changes from
version A to version B, I would say changes are alright. However, I
revise it from the perspective of 3 years, so changes aren't
satisfactory, especially for me. If I had to estimate the work time for
people, I would assess it as 1-2 developers [what about the remaining 28
developers?] who have developed OnDemand version [I don't count the
changes that haven't been also added to the community version]. 

When your human resources for the project development are so limited,
you should resort to work or help of others. When I reflect on the
development of the community version, I see hundreds of bugs which
haven't been fixed for years, and basic functionalities that don't work
[e.g. migration between versions - nearly every migration pack has
errors that weren't fixed in the subsequent versions]. 

When you published v7 I saw that all the changes published at
code.vtiger.com were completely ignored [as if they never existed], the
community's work has been disregarded, and many great changes won't be
moved to the new version. Many of you, who are happy about the release
of v7, should first of all understand that this is a marketing move
[first they make you wait for a few months for a new version, then they
publish a version that has nothing to do with community, and then you
got months of stabilization in front of you]. When I recall the
migrating script 5.4 -> 6.0 I get chills and I can already imagine what
the migrating script 6.x -> 7.x is going to look like, even though there
are just a couple of changes. Because how can you prepare a decent
update package when you don't even use it yourself?

The new version has been published, but apart from that nothing has
changed, because there's still nobody to update it, all the questions,
comments and errors from community still won't be taken into
consideration. The community doesn't know what to do, so some will wait,
some will move on, some are happy with what they got [which means
they'll sell paid modules and the fewer changes the better for them].
For the community members to be able to introduce changes into the
system, they'd have to be able to actively communicate with the
developers, and this hasn't changed for years.

I look at projects from a completely different perspective - the quality
of code - which hasn't improved in the 7.0 release. That's why I'm also
disappointed because 6.0 version introduced a lot of changes in quality
and it should be continued. Unfortunately, I don't know why [maybe lead
developers resigned] but the quality hasn't improved since 6.0. However,
it doesn't mean that 6.0 was perfect code-wise but it introduced new
quality and if it was continued, then after 3 years there would be some
really good code in 7.0. Unfortunately, Vtiger 7 is a blend of 6.0, 5.4
and in some places even older versions such as 5.1, 4.x etc. 

Vtiger has the greatest amount of errors in code [over 42 thousand,
including 532 Critical, 4889 Major, 20021 Minor, 17102 Info -
https://insight.sensiolabs.com/projects/302bc9b8-5a04-4405-ac50-ba10d2bc953c
[1]] from all the projects I know [Vtiger, coreBOS, SuiteCRM, SugarCRM,
EPESI, oroCRM, ODOO, YetiForce], and these errors influence the
development, security and transparency. Over 70% code requires
refactoring and 7.0 version only generates additional code instead of
fixing bugs which sometimes have been in the system for 10 years. I
don't know why Vtiger doens't follow any standards in designing PHP
applications that have been described for example here:
http://www.php-fig.org/ [2]. You may think that it doesn't matter, but
when you aren't selected for a very large project only because of the
low quality of the system, you will understand that if you don't invest
enough energy into the quality, it will influence your business sooner
or later. 

What if you stop at one version and just focus on fixing and stabilizing
it? Maybe a custom fork whose only purpose would be to deliver a stable
version? 

A great example would be coreBOS, which for almost 3 years [since March
2014] has been concentrating on fixes only. What effect did they
achieve? Unfortunately at Github they only got 11 issues, 27 people
watching them, and 332 downloads at SourceForge. From the practical
point of view, coreBOS is Vtiger 5.4 with hundreds of fixes as far as
code formatting is concerned, and some minor changes that don't impact
code quality at all, because this project doesn't comply to any
standards whatsoever, ie. http://www.php-fig.org/ [2]. It also includes
over 35.000 errors [659 Critical, 10050 Major, 8278 Minor, 16594 Info].
Additionally, just like Vtiger, this project requires code refactoring
[more than Vtiger does, because Vtiger 6.0 brought many positive
changes, which were not included in coreBOS].   

VTE CRM is the same example as it was forked from 5.2.1 version but they
haven't changed it since then, so all security vulnerabilities are still
present. What novelty has this fork introduced, except dividing the
community and creating a little monster which won't be developed?
SensioLab 50.000 errors [3300 Critical, 13774 Major, 9386 Minor, 23475
Info]. It introduced some commercial modules… and nothing more. I don't
know how many forks like that have been created but if I haven't heard
about them, they probably weren't successful. I understand that these
companies try to earn money from paid modules but, as a result,
customers receive a worse product without any perspectives for the
development and on a low degree of security. I'm trying to understand
why Vtiger pushes us away and those, who can develop a modern CRM
system, build semi-finished products that have no perspective for the
highest possible development level. 

Creating a product that keeps evolving requires a team of programmers
who develop their coding skills on a daily basis, a team that grows
together and has adequate financial resources to be able to keep going
on, regardless of the number of clients in the first years. If you think
you can take Vtiger 7 and start developing it independently, then you
will most likely not be capable of dealing with the burden of
development and responsibility, because it requires looking at it from
the perspective of several years, not just from a one month perspective.

Currently, a few systems are developed very dynamically [especially
commercial systems] because they have enough financial resources. These
systems are Base, SalesForce, Dynamics, SugarCRM… Do you consider where
you are going to be in 2, 3, or 5 years time if Vitiger's community is
already so divided, there are hardly any financial resources and no one
takes care of the quality and security? Last 6 months I have been
negotiating with 2 very large companies that want to deploy a CRM system
on a large scale. Do you know what are their opinions about open source?
In order for us to take part in the procedures, they had to conduct
functional, performance and security tests. We had to make a lot on
changes in the engine over many weeks, so that we wouldn't be ashamed
and our system could pass the tests. 

I understand that the current situation is very satisfying to 2-3
companies that sell Vtiger modules...and they won't be too keen on
accepting changes, because the more flawed the system is, the easier it
gets to sell some tiny feature for 100EUR. However, in my opinion, this
attitude is not only harmful to the Vtiger community, but to the entire
open source community. 

I'm no longer as active here as I used to be, and I guess this is how
it's going to be, and it's due to the fact that Vtiger doesn't have much
to offer anymore. I wish I was wrong, and I wish you could surprise me
in 2017.

-- 
Z poważaniem / Regards 

BŁAŻEJ PABISZCZAK 
_Chief Executive Officer_ 
M: +48.884999123
E: b.pabiszczak at yetiforce.com 

Links:
------
[1]
https://insight.sensiolabs.com/projects/302bc9b8-5a04-4405-ac50-ba10d2bc953c
[2] http://www.php-fig.org/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.vtigercrm.com/pipermail/vtigercrm-developers/attachments/20161228/6afd52d6/attachment.html>