What I intend to do can be summarized with Share your Kraft. Up to now, Kraft is working fine for a single user. But what if a team wants to use Kraft and share number cycles (which are base for the document numbering like invoice number), documents and template catalogs? Well, as long as they share the same database, it might work (I didn’t test deeply) but if they happen to be on different locations it becomes difficult. I try to make that possible.

My development target for Kraft is simplicity. For the user of course, but also for the setup. The server to share data, which is obviously needed, must work on a cheap hosting offer, and it must work with a weak internet line. So a database connect via internet is not possible.

I decided to investigate in ownCloud and enhance it with a plug-in called KitoC. ownCloud is a project started by Frank Karlitschek and implements a handy but scalable WebDAV Server beside more. Seems to fit my needs perfectly. Yesterday I implemented the number server function in KitoC after good conversation with Cornelius at breakfast in the office. Not very much achieved yet, but had to learn a bit of ownCloud first. I keep you posted.

But the other nine applications are also really interesting, such as goldendict, which combines multiple dictionaries on the desktop or TOra which is a cool database GUI. We do not have them in Factory nor
Contrib.

The next openSUSE release 11.4 is slowly but surely coming up and I think it makes sense to add cool software now. Maybe the listed apps in the blog are ideas to spice up our distro a bit with good software? I volunteer to take care of Kraft

Today, Sourceforge posted a blog about Kraft after they kind of mail-interviewed me. It’s nice, it really focuses on the things also important to me. This might be another step towards a broader user base for Kraft. I say that because one could have the impression that the number of people actually really using Kraft could be larger. A high number of users is one of the fundamental criteria for a successful free software project and thus I am constantly trying to understand whats the reason for the impression or the fact.

The first idea is that the Kraft project simply does something wrong in the way a project should be driven. But there are releases, there is a so far ok website, there are communication channels with information on it and people answering questions. Of course, it always could be done better, but I hope and believe we are not doing too bad. Marketing could be more, that’s granted.

The next thing could be that nobody needs this kind of software. But there are quite some companies doing this kind of software in the commercial space. So there must be a market. Actually I think the market is huge. People are writing invoices all over the world and I bet many of them are not really satisfied with the way they do it usually which makes Kraft at least an option to try for them.

And this might lead to better path: Probably these people do not know that the option exists. They simply haven’t heard about Kraft yet and if they would there is a good chance that they would not believe that it is free etc pp. And this is probably not specific to Kraft but also applies, of course much more weaker, to larger projects like openSUSE or KDE: A lot of people from the ‘real world’ don’t know about free software communities, the ideas behind and the benefits for users of the software. That sounds strange to us, as this is our daily reality, but start with asking your parents or non computer related friends if they really understand what it is about. Imagine what people know who have no computer job nor -hobby nor know you!

What consequences can that have for us? Well, we could decide to skip this group of people. That would mean, beside some other effects, that Kraft would not make sense any more and I don’t like that. It probably should influence the way we see the ‘product management’ aspect of our projects. For me, ‘product management’ is often equivalent to “take care that the result is especially useful to non computer scientists” (which is probably not what PM really is about) and the focus on that is very important and the precondition for the next point.

We might have to take our projects even more out of the geek niche and go to places where the ‘real world’ happens. That is difficult for various reasons. First, it means that we have to start to explain again from start, and maybe also get questions where the answer is not obvious. Furthermore it might have practical issues, because for example fairs for handcrafter utilities charge seriously for software boothes which is not the case if we present projects on FOSS events.
On the other hand its easy because we all just have to spread the word even more and tell everybody about free software, our projects and free culture. And try to think as if we weren’t free software people. I know, most of us do already what they can and that’s great

One of the new features of Kraft 0.40 makes the setup of Kraft easier. The file based sqlite database was added in addition to the mysql backend, so for evaluation no database server setup work is required any more. Furthermore a setup wizard that guides new users through the setup process aims to make it easier for new and unexperienced users.

Please help us by testing the beta version and report bugs back to the Kraft user mailinglist. The sources can be downloaded from the Sourceforge project page. Binary packages for openSUSE 11.1, openSUSE 11.2 and (K)Ubuntu 9.10 are available from the Kraft Beta repository from the openSUSE Buildservice, see here how to add and use it.

Documents should represent the ‘coorperate identity’, which in small size firms probably comes down to printed stationary with a company logo and some other information on it. Kraft has to print  nicely on it. For that it is important that the layout can be configured at all and without compiling Kraft if the customer address should be printed fife millimeters higher for example.

Currently Kraft uses a document template written in RML for the layout. RML is a XML format which can be converted to PDF utilizing a python based command line tool which is called by Kraft. RML is a open source toolkit, quite powerful and mature. However, it does not solve all problems with flexible document creation and sometimes comes a bit unhandy. As a result our eyes are always open for alternatives.

Here are some requirements a template system must provide:

• There is a document template in the file system. It can be changed by the user without recompiling Kraft. Kraft picks it up, fills the document values in and processes it to PDF. Other output formats are optional.
• Layout: Areas where parts of the document are printed can be freely specified, ie. where the address, the date etc. is printed.
• Graphical elements like lines, fixed text, boxes, colors and images can be placed everywhere.
• The system knows at least different layouts for the first page, middle pages and the last page.
• All pages have page header and footer.
• Loops: Since an invoice for example has an unknown amount of items the system must be able to handle that, including clever space management with  pagebreaks. Nested loops are possible.
• Maintain areas which must not be split, i.e. an invoice item should be printed completely on one page and not be split by a pagebreak.
• Text faces, paragraph alignment, width, spacing and these kind of things must be configurable in the template.
• Some variables are available such as a page counter.
• Really great would be if the system provides carryover of calculations, like  on the top and bottom of each page the so far accumulated sum is printed.

Which free layouting and PDF generating system is able to provide that, preferably Qt/KDE based? Kugar was striving to solve it but when I tested it it did not work out.

Another idea is to use the ability of KWord to work with templates. If Kraft could read KWord templates, fill them and automatically generate a KWord doc from it, that would be a great solution, because in addition to automatic PDF generation documents could easily be exported as KWord docs and changed manually if needed. A great ‘template editor’ also would be available. This would in the direction of office suit integration that commercial Kraft competitors nowadays have.

I am not sure how far we are away from that. Something to investigate.

The best thing is that an additional developer works on Kraft: After my last status post Thomas Richard (account trichard) contacted me that he is interested, next days I had the first patch in my mailbox and from that point of time on he constantly contributed high quality changes into the Kraft repository.

His high energy, dedication and fresh ideas gave me a new motivation push after having worked on Kraft basically alone for more than four years. That’s great!

The last months we worked on porting Kraft to the KDE4 platform which is in a quite good shape in SVN already: Kraft compiles without warnings and without Q3 and K3 support classes and works stable again.

We couldn’t resist to make use of the new capabilities of KDE4 here and there and as a result we have a few small feature updates as well. The most interesting might be that the first KDE4 Kraft version will additionally support a sqlite database backend which eases setup and configuration for users tremendously.

Following our friends from the KMyMoney project we will come up with a first Kraft-on-KDE4 beta soon. Please stay tuned.

Kraft is one of the candidates for the KDE group for financial apps which is a consolidating idea and was encouraged in Alvaros article A group to bind them all recently.

Unfortunately it is still a bit tricky to set up. To make it easier to check it out Live Images were created featureing Kraft on an openSUSE distribution with all tools and  interesting demo data. That is perfect to try it out and give it to friends and colleagues and talk about.

Please check the download page of the Kraft Homepage for details.

I say ‘small business’ as a target group and I mean small shops doing crafts like carpenters or plumbers or landscaping working alone or with a few people. I think free software and especially KDE is very good for these kind of companies. Larger companies usually go for specialised software, which is available for nearly all kind of crafts in all levels of usefulness and quality.

When I talk in the community about Kraft (I am of course not as good as Tackat in his best Marble-times) I sometimes feel that the coolness factor of this kind of software is not so big. People seem to think “How can one do this kind of boring, already-there software?”. Well, yes. This kind of software exists. But as free software and on KDE, well integrated into the desktop? Not that I am aware of.

Here are some good reasons to work on Kraft for me personally:

1. I think it is important that this kind of software is available. Not only Kraft, but other stuff people need for their business, for example financial software like KMyMoney. Well integrated in KDE this can enable another huge group of users which now uses other systems.
2. It is serious. Kraft is software people use to get their money. If somebody has done work and wants to invoice it she loves a well working software that saves time for her. But if it does not work, it becomes a serious issue quickly because only written and sent out invoices are good invoices from that POV.
3. Especially because there is much competition from the commercial side, it is fun to try to create free software that is even better from for example the usability perspective. It is real challenging.
4. I am somehow addicted This year I work for twenty years on this kind f software. If you like you can check out an underground video which shows software running on an Atari ST, used for daily business in my brothers landscaping company. I released that version in 1991, I still have an earlier release from 1990 which I could not get to run anymore. I started to develop it in 1989.

But back to the new Kraft-release: Beside other things I did some change to the tax system which make Kraft now useable internationally (shame  on me that the earlier versions where tied to german taxing too much).

So please, tomorrow first thing knock at the door of your handcrafter neighbor and ask him if he has thought about invoicing with Linux – Kraft  is with you, when you support him. Chances to get very interesting insights on how non-geeks see the computer world.