[Koha] [Koha-devel] Koha development

Fri Feb 13 12:54:21 NZDT 2015

While I can understand the motivation behind such a comparison, I think it’s going to be comparing apples to oranges.

After all, what is Koha? 

While the elected release and maintenance teams focus on the “strategic product” fronted by koha-community.org, there is an unknown number of companies and libraries around the world who are working on developing Koha in different ways. Many people will register a bug report, write a patch, test a patch, and upstream a patch without it seeing production beforehand. However, there are many patches which will see production at a local level, before being upstreamed to koha-community.org and distributed globally. There are also many patches in production at the local level which might never be upstreamed for any number of reasons. For instance, here at Prosentient Systems in Australia, we integrate Koha with a lot of our other “strategic products” which are marketed towards Australian libraries in particular. While I try to upstream more and more patches, the global stats won’t ever reflect the amount of development work I actually put into Koha. 

Which gets me back to “what is Koha”?

You could argue that it’s the global product. That since it is “the” version, that it is the definitive product, and thus the thing to be measured. Yet, lots of people who have access to Koha via an entity with development resources (this could be a commercial company but it could also be the IT department or technical library staff) will experience a slightly different Koha. I suppose you could argue that those are all forks of Koha, although the silverware analogy is a bit passé. If you think about Ubuntu and Debian, Ubuntu contributes patches back to Debian. How then to measure development efforts for Debian? Is it current and pending patches for Debian? Is it also time spent developing Ubuntu which may be upstreamed back into Debian? How do you measure time per se? There’s a lot that goes into a commit and there’s a lot that goes into development which isn’t even related to a commit.

I admit that I’m  curious about the “real” numbers myself. I’m curious how much time goes into development for a proprietary product versus an open source product. However, I don’t think comparing is that straightforward. None of us are employed by Koha. I suppose you could try to count all the different companies that support Koha and count how many developers they have working on Koha… which I think is what you’re trying to do… but that goes back to the original problem. What is Koha? What is Koha development? I probably spend 95-99% of my time working on Koha, but some of that is support and some of that is development. Some days it’s all support, and some days it’s all development. And then there are the unpaid volunteer hours. Many of us also put in a lot of unpaid hours which are pretty much impossible to quantify. I’m sure some of us devote dedicated unpaid hours to Koha. However, I know that I tend to put in unpaid hours depending on particular issues or how much free time I might have at a given time. I doubt you’re going to have developers of proprietary products willingly putting in unpaid hours, so there’s no one-to-one comparison there either.

I think we really want to know the numbers, and I think media tries to provide them. When Heartbleed broke, we were told that there were 4 OpenSSL developers and 1 or 2 of those were FTE. If you look at  <https://www.openssl.org/about/> https://www.openssl.org/about/, it lists 15 active developers. There’s no indication of how much time they devote and I imagine it would be difficult to measure it. It’s remarked that GPG has 1 FTE and is working on getting a 2nd FTE. That one might be easier to quantify… but I’m curious about how many contributions are made to that project as well which aren’t indicated much of the time. In any case, both of those projects seems to have relatively small lists of active developers published.  If you visit  <http://git.koha-community.org/stats/koha-master/authors.html> http://git.koha-community.org/stats/koha-master/authors.html and sort by “Last commit”, you’ll see 24 developers have contributed in 2015 (including yours truly). If you go back to the start of 2014 to the present, it’s about 93 developers. Of course, some of those might just be one or two contributions here or there. 

But what about the developers of proprietary products? How much time is spent in team meetings? How much time is spent writing code? Reviewing code? What is development? That’s probably an even more difficult question than the previous ones. This is where you get into the whole argument of development versus operations, and the whole DevOps thing. How much of support is actually part of development and vice versa? 

Even comparing development FTEs between companies with proprietary products might not make sense because the actual work is so variable. You could say this company has this many FTEs working on the ILS product. That’s probably general enough to be true. However, in terms of FTEs on development, I don’t think it’s going to be very reflective :/. 

For instance, let’s look at  <http://www.americanlibrariesmagazine.org/sites/americanlibrariesmagazine.org/files/content/Charts_MarshallBreeding.pdf> http://www.americanlibrariesmagazine.org/sites/americanlibrariesmagazine.org/files/content/Charts_MarshallBreeding.pdf. It states that there was 1 “Dev” at ByWater Solutions in 2013. Ok, there might have been one FTE “developer” employed there. However, I know for a fact that ByWater Solutions sent patches from at least 4 people in that year, and that’s probably an under-estimate. Here at Prosentient Systems, all the contributions to the Koha community are vetted through myself, but we generally have two people working on Koha. However, we have different hats as well. It’s impossible to say how much of our time is spent with the Development hat. 

Yet, if we don’t provide you with a number, then the number that you do report will also be incorrect. So it’s more than 0 but less than 2. 

David Cook

Systems Librarian

Prosentient Systems

72/330 Wattle St, Ultimo, NSW 2007

From: koha-devel-bounces at lists.koha-community.org [mailto:koha-devel-bounces at lists.koha-community.org] On Behalf Of Marshall Breeding
Sent: Friday, 13 February 2015 8:34 AM
To: Chris Cormack
Cc: koha-devel at lists.koha-community.org; koha at lists.katipo.co.nz
Subject: Re: [Koha-devel] [Koha] Koha development

For most of the vendors in the table, development is focused on one strategic product.  In some cases they may also have some other projects, so that whole number may not be devoted to the main ILS.  The column of interest is the one for “Dev” counting only developers, not the total headcount of the company.  That’s the one for which I am trying to derive a comparable figure for Koha.  I know that Catalyst is quite a large company involved with many projects, but I’m specifically interested in the personnel resources allocated to Koha.

-marshall

From: Chris Cormack [mailto:chris at bigballofwax.co.nz] 
Sent: Thursday, February 12, 2015 3:23 PM
To: Marshall Breeding
Cc: Brendan Gallagher; Tomas Cohen Arazi; koha at lists.katipo.co.nz <mailto:koha at lists.katipo.co.nz> ; koha-devel at lists.koha-community.org <mailto:koha-devel at lists.koha-community.org> 
Subject: Re: [Koha-devel] [Koha] Koha development

On 13 February 2015 at 10:14, Marshall Breeding <marshall.breeding at librarytechnology.org <mailto:marshall.breeding at librarytechnology.org> > wrote:

Brendan,

Some personnel FTE numbers from last year can be seen on the tables from the Library Systems Report:

http://www.americanlibrariesmagazine.org/sites/americanlibrariesmagazine.org/files/content/Charts_MarshallBreeding.pdf

The numbers are for the company overall, and not limited to efforts expended on specific products or projects.

In that case for Catalyst there are 235 staff, of which 182 are devs

Chris 