Donut Charts

by Italo Vignoli on March 15, 2012

I have just discovered Donut Charts, a visual feature which has been there for quite a long time, but I have never been able to master. It is a nice feature which comes handy to compare developer’s affiliation of LibreOffice with that of Apache OpenOffice.

In fact, although we have published several times the pie chart showing the affiliation of TDF developers, there are people who keep on hinting that LibreOffice is controlled by SUSE. The Donut Chart seems to show the opposite, especially when looking at SUSE and IBM percentage of the respective communities.

Pippo

Here, I am comparing over 360 TDF developers with 21 Apache OO developers, which are a picture of the situation at the end of February 2012.

I have not included in the count of TDF developers the former Sun/Oracle employees (53) and I have slightly reduced the count of volunteers as I did not want to include people like myself who have committed a patch but cannot by any mean be considered hackers.

TDF has managed to attract over 360 new developers in 18 months to hack the copyleft licensed LibreOffice, at the monthly average of 20 new hackers. Apache OO has attracted 21 new developers in 7 months to hack the now permissively licensed Apache OO, at the monthly average of 3 new hackers. At this pace, Apache OO will reach 360 developers in September 2021.

At the start of the LibreOffice project, when TDF developers were still a small number, someone has said: “Having an active contributor base of 150 coders would be an amazing accomplishment. Even a solid group of 30 coders, if not all from the same company, would be an impressive project”. LibreOffice now greatly exceeds both of those numbers (actually, it did so in less than four months, by the end of December 2010).

And not only that: LibreOffice does so with a broad and diverse base of contributors and supporters, a feat that seems easy to achieve, but which the history of open source has proven again and again a tricky thing to get right.

Creative Commons License
This work, unless otherwise expressly stated, is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

{ 12 comments }

Felipe March 15, 2012 at 4:43 am

Hi Italo, I have a question. The recent announce of Intel supporting LibreOffice, is it going to be translated into hiring full time developers ? or the collaboration is going to follow another pattern ?

Thank you.

Mark__T March 15, 2012 at 8:41 am

I wish you would stop picking on each other and instead concentrate on the product itself. These kinds of post, doesn’t help working together (where possible).

superman March 15, 2012 at 9:34 am

Impressive numbers. But impressive numbers are not everything.
To me it looks this way: over 300 develeopers, but the feature list is still not very impressive. Yeah, you have a lot of features listed. But what are they? A modeless dialoque, display the PPI in an existing label, some new icons/shapes, a multiline input bar, better working import filters. These are no “features”. These are matters of course. Matters of course that are not very hard to implement (to change a dialoque type you simply change the API call to use the other dialoque).
And you needed 300 devs just for that?

Where are the real features? Like:
- Complete GUI overhaul (currently there are a lot of icons and endless text-based menues – that’s so “1994″)
- math input, especially for Draw (enter a function like “x^2 + sin(x)” and have it drawn)
- get rid of Java (I remember this being one of the #1 announced features when LO popped up – I still get the same amount of “Java missing” errors after a clean install)
- Fit arbitrary function for Calc (fit user input functions to data points; the solver needs Java and the hardcoded functions are very limited, could be coupled to the math input suggested two points above)

These ARE features. And THESE are some of the features people want to have.

Contrarian March 15, 2012 at 11:00 am

@Mark__T: It’s a shame Italo can’t control himself, but there’s no doubt that IBM’s Rob Weir is actively provoking him (as well as bullying anyone else he can’t control). This post appears to be a response to Rob’s trolling over on Google Plus – see https://plus.google.com/u/0/107646708505179576030/posts/ViuT8bLgz2p – and it happens so often and Rob is so smart and loyal to his employer I have to believe it’s an active strategy by IBM.

Cor Nouws March 15, 2012 at 1:27 pm

@superman: I understand what you mean. And I know many more features that have to come. However, the first step IMO is building a wide and solid community working on a clean and solid code base. That is what happens now – apart from features and fixes introduced at are already the same time.

superman March 15, 2012 at 3:47 pm

@Cor Nouws: Of course, that’s correct and important. But I don’t like it how minimalistic additions are presented as “super new features”. It’s great that they are finally here, but let’s see it from the other side: It’s an absolute shame that I had to wait for a PPI display and a modeless dialogue since when? 2001 when OOo was released? Or early 1990th when the first Windows version of StarOffice came out? The general look-and-feel of the application is like MS Office 97, maybe Office 2000. And yet such features like a modeless dialog are presented as ultra-cool features…

Let’s make a small calculation: We start with StarOffice 3.0 from 1995 because it was the first suite (instead of only Writer) and it was available on several platforms and it used modern window managers (like Windows). So this can be considered as the grandfather of LO. From 1995 to 2012 (LO 3.5) it’s 17 years. LO exists roughly 1.7 years. So: it took 10% (yeah TEN percent) of the whole lifetime of the suite and over 300 devs to develop a modeless dialog, make import filters (one of the most important things) work ALMOST (yeah… almost) perfectly and create some flickering (…) pop-up menus (header/footer menus) while in the rest of the 90% of the lifetime with far less(!) devs a whole office suite was created?
I don’t think the minimalistic features make 10% of the code base, like they took 10% of the whole development time.

Don’t get me wrong, I use only LO, I don’t use MS Office. LO is a great product. But the whole team behind it acts – IMHO – a little bit to “self-confident”.

It’s great that LO is beign developed and all the devs are working hard, I’m sure of it. But the “degree of efficiency” (if I may call it this way) seems rather poor: 300+ devs in almost two years making only rather small features, compared to 100 devs in 15 years creating a whole office suite.

Bjoern Michaelsen March 15, 2012 at 5:38 pm

“It’s an absolute shame that I had to wait for a PPI display and a modeless dialogue since when? 2001 when OOo was released?”
Indeed! How much did you pay for the developer you sponsored to implement these features?

As for your other calculations: You are very inconsistent there. If you follow that logic, “everything of value” was already in StarOffice 3 which already had Calc, Writer and “nothing of value”(*) was added ever since. The Patero-principle is hitting hard here and when you count in the wide range of bazillion possible features including the like of “PPI display” the theil index for this is skyrocketing.

If you care about a specific feature, you are very much invited to join us. Others did — thats where >300 developers come from. So why dont you? Looking forward to your patches!

(*) Of course, there never was any GUI overhaul at all in those years and LibreOffice is still strictly a Windows 3.1/DOS application.

superman March 15, 2012 at 8:09 pm

@Bjoern: You talk from a dev point of view. I talk from a user point of view. I don’t have to care how much devs did the work. I don’t have to care what happened internally.
I only have to care what I can use see in the final product. And this does absolutly not meet with the things I read in all these statistical blog posts. I read about hundreds of new devs in some months, but I only see minimalistic new features and some new colors.

I – as a user – don’t have to care how the GUI works internally. I know that there was a lot of refactoring and reworking since Win3.1. But the final output that the user sees does not much differ (aport from some more colors) – there are still endless lists of buttons and menus that the user has to work through to find a hidden function. This annoys me – especially when I see that MS Office and even other OpenSource suites do it better.
Another example: I also don’t see any effort in providing a 64bit LO for Windows. And please don’t tell me all the DEVELOPER reasons I hear all the time. The Firefox devs are having the same discussion currently: They argue like “Oh, 64 bit brings no advantage, no plugins would work, more work for us, etc. etc.” But their stats say that over 50% of the Nightly users take the unsupported(!) 64 Bit versions, and more and more Stable users are requesting a 64 Bit version. Yet the devs still ignore this wish. They will have to live with a decrease in their user base (actually, people are already leaving to Chrome and Opera that are already working actively on stable 64 Bit releases just because of this). They can argue from a dev POV as much as they like – but this won’t stop the “stupid no-arguments-for-64Bit users” leaving.

All I’m saying is this: Don’t make the mistake that many OpenSource projects do:
Don’t, never ever!, argue from the dev POV when talking about new features and statistics if the users are not happy – the users WILL leave if their wishes are not fullfilled – no matter how good the dev reasons are. (simple Supply and demand)

Open Source User March 15, 2012 at 10:49 pm

@superman

If you are not happy with LibreOffice feel free to migrate to one of the other open source suites such as Koffice, Symphony or Apache OpenOffice. That is the joy of open source…..put up and help code, or shut up and move on. Or purchase MS Office.

Open source is not about what you want. It’s about the freedom to chose, or modify and contribute back to the community.

I am not a developer…..I am a user also. You saying 300 people coding but not delivering anything I want is pointless. Personally I am thankful for all the work these guys put in so that I can have a decent office suite for free.

Bjoern Michaelsen March 16, 2012 at 3:00 am

@superman: “the users WILL leave if their wishes are not fullfilled”
No, they will not — as this is not something considered in absolutes, but in relatives.

Users would switch if something else fulfills their wishes better. Currently there is nothing that does so in the open source productivity market. And that is because we are channeling and focusing the available volunteer resources quite well. And by doing that we are indeed _creating_ new volunteer resources willing to work on open source productivity.

Of course we should never assume that we could not do better and we do not. We are constantly striving to improve the way we work. LibreOffice 3.5.1 is the best free office suite ever, but it wont stay that — the future has a lot of versions of LibreOffice versions to come — each one improving on the previous one.

superman March 16, 2012 at 9:41 am

@Bjoern: “LibreOffice 3.5.1 is the best free office suite ever”
Yes, you are right. I never said something different.

I say it again:
What I criticize is NOT LibreOffice or OpenSource! It’s great.
What I criticize is the argumentation POV from many OpenSource devs that annoys me (not only me, actually).

If I say (to stay with this example) “I want 64 Bit LO” and a LO dev answers back “Why? it does not bring any advantage and I have to work more for no real benefit!” then I – the user – am not happy. Of course, from the dev POV I’m the stupid one because I don’t know the technical details and from the technical details it’s obvious why there is no 64 bit version.

You have to see it this way: The users reporting back (like I do now) are an absolute minority. Almost all users are NOT reporting back. But even if they are not reporting back, they are still often unhappy with some things.
Some example critics I recently really heard from people in my environment (not bound to LO or OSS):
- “Why does the new FF look different? I don’t like this!”
- “Why does Adobe Suite wants to install PS 32 AND 64 Bit at the same time? Stupid!”
- “Why does FF do not have a 64 Bit version? Why did I buy a 64 Bit CPU then if there’s no software for it?”
- “Why does LO has no incremental updater? Was this update really neccessary?”
- “Why does Linux looks different than Windows? No, I don’t want to learn it, I want to use the things I know. The Linux devs are stupid. I want Windows back.”

The people saying these critics have NO idea about computer tech. They are PURE users of these specific softwares, they don’t know what a bugtracker or a repository is. They only know: There is the software and they have to use it. They don’t know where it comes from.
And you know what happens if I argue from a dev POV against these critics? If I say:
“FF has no 64 bit version because the devs would need to adjust thousand lines of code and need a lot of testing”,
“LO has no incremental updater because nobody developed one and developing one is hard, because the software consists of thousands of files that would need validating in an incremental update and this for each previous version. There are working 300 devs on LO but with focus on more important tasks, like bugfixing and performance increases.”,
“You can install another windows manager for linux that makes it look different. You only have to click there and there and enter this command…”

And you know what they replay? They reply:
“And? I don’t care! I want that it works. Now. All these arguments do not solve my problems!”

I say it again:
You do great work. 300 devs are a lot. It’s good what you are working on. But please:
Never argue with dev reasons (coding terms, raw dev count numbers, comparing to other dev teams, …) if a user is unhappy!
And never say “we are doing great work” if a user is unhappy!
This makes him even more unhappy. Even if your dev arguments are correct, it won’t solve and fulfill the users problems and needs and wishes.

Cor Nouws March 16, 2012 at 9:55 am

@superman: I don’t agree with your ‘calculations’ about features, time invested etc. It’s more complicated obvious.
But I do appreciate the way your explain user experience :-)
Showing the good things we do, of course is important to make more people enthusiastic etc.
thanks – Cor

{ 2 trackbacks }

Previous post:

Next post: