Hello,

I'm considering switching once and for all from windows .
Already did @work, but my home computer is a different story.

The main catch is that I have this MS Access database, which I use to manage my nature photo collection (see for yourself if you like: http://www.floredefrance.com).
Over 10 years of work, more than 30.000 pictures, and not a few lines of code to make it do exactly what I want...

All considered, not something I can give up easily, but I'm still willing to go.

So let me summarize my needs in a few straight (an I hope simple) questions:
Just to help you place my limits: I can code and script (VB, PHP, SQL at the moment) and I'm well acquainted with image magik and such... I probably can learn to use PERL or Ruby to replace VB.

On with the questions (all this (and more...) I can do wih MS Access):
Will I be able to display pictures in forms by using a hard drive path saying where the actual file is ? (I cannot physically store the binary of 30000 pics in the DB)

Will I be able to interact with the form structure on the fly ? (for example by populating a drop down list control based on a search work entered in a box beforehand ? hiding or showing controls depending on field content?)

Will I be able to script call to external programs ? (like launching google earth and make it display the place where current picture was taken? or run a command calling an external program that will make a thumbnail of images returned by a query?)

Will I be able to interact with the file system ? (rename files using db content ? create an XML or CSV file and write data in it ? change IPTC tags within a picture ? and so on...)

OK, I probably can find a hundred other questions, but with these, I will probably have my answer.
I'm really eager to make that switch: been waiting to do so for a long time, so please say YES !

Thanks in advance to anyone who answer me !

I am not the person to give you a complete answer but perhaps I can help steer your thought process.

Microsoft Access is a combination of a number of individual products:

1) a database engine (like Oracle DB but not nearly as robust) capable of storing data and retrieving it with SQL queries.

In the Linux world there are many, including SQLite which Kexi uses for internal metadata. Some of these DB engines are quite capable of storing ALL your data, including images. Check out MySQL and Postgres. I'm not sure but believe Kexi can use both of these for data storage (but not metadata).

2) a programming language which can access the data in a DB (through queries) and the operating system/other programs through external calls.

There are many choices in Linux. From what I can tell from your web site, PHP will be as good as any and will do all you've listed so long as you're willing to live with just responding to input screens with dynamically constructed output screens. Full interaction will require something that runs on the user's PC, like Java or Javascript/ECMAscript. If all you care about is running on YOUR PC, there are many languages available.

3) a form and report generator.

I rarely have a need for these so can't advise, but since you know PHP you can of course generate them with that language. I'm told there are WYSIWYG tools for this but I can't point you to them.

4) a graphical query generator to help you construct your SQL queries.

I hope somebody tells me I'm wrong, but I haven't found one in the Linux world to match the one in MS Access. If you're OK coding SQL yourself you may not need this but I find it far harder than the MS Access way.

5) All of the above running fully integrated in one package.

I think this is what Kexi aspires to be, and it is closer than anything else in the Linux world but it's not yet a drop-in replacement.

TO SUM UP:

YES you can maintain your web site on Linux (and maybe better than on MS products) but there will be differences.

Best of fortune to you.

Thanks for this detailed answer.
Yes your description is quite accurate. And also it confirms that Kexi is (or will someday be) the tool I need to escape Microsoft shiny claws

Just a few precisions because I just assume too much about what people will understand about what Access does for me:
The database I use is NOT the one behind my website: it is far more & much bigger.
The one powering the website is only an export of a very small subset of the one I manage with access.

My Access DB is a tool that cannot be reduced to the website management: think tip of the iceberg, if you will. There are some thing that only become possible when enough data is available (like over 10 years and more than 30k pictures)
I can ask things like:
- plot me a map of where I took pictures of this species
- make me a graph of the time distribution of pictures through the year for this plant (flowering times, you know)
- export me a KML file of all sites with ladybirds, color coded by number of species, and open it in google earth
- show me a list of all names that have been given to this species in the past and are now synonyms
- count how many species I have photographed in a 50km radius around this point
- what is the ratio between fuel cost and number of species for this place (is it cost effective to go there ?)
And if I get a new idea, I just get at it and code until it works...

In fact I use Access as a front end to manipulate several data sets, of different formats and sources:
- some of the data is stored in attached CSV files (lists of plants and animals published by museums for example, these being regularly updated)
- some data is in native Access storage (parameters for example)
- most of the data is in a MySQL server running on WAMP on my machine
- and finally the subset of the data that is used by the website is also on MySQL (and transits there by means of SQL dumps and FTP, my provider not offering direct access)

So at the moment I am not tied to MS Access by its proprietary data structure, but more by its abilities as a flexible data manipulation tool.
And That what I expect Kexi will be able to replace.

On another point, yes I am aware that MySQL would perfectly be able to hold the binaries of my pictures, BUT, transferring via FTP 60k pictures (a medium size for display and a thumbnail for each) is a pain but feasible. On the contrary I don't think uploading a SQL dump of 10 or 15 Gb would be practical or even possible.
And, call me old school, but I prefer to have my pictures in files where I can see them

For the rest of your comments, you are perfectly right:
- Access is a really easy to use query generator and there are very few queries I had to do outside of it due to limitations (some involving unions if I remember well)
- Access is not too bad when it comes to forms and reports
- Access is really powerful when it comes down to desktop interaction
I could probably code most of this in PHP, but I don't think PHP would cover the last point as well or as easily.
Well... I sure hope someone of the Kexi team will come along and read this, because the world really needs an alternative to Access
In the meantime I'm really going to try Kexi.

I'm in a similar position but on a much smaller scale. I used Access for years and is one of a very few apps that are not avaiable in Linux yet. Kexi is not Access. It is far inferior and I doubt it will ever be on par with Access but then again, you have a multi-billion dollar company behind Access.

As I said, I'm on a much smaller scale and I'm sure Kexi will do everything I need it to do. My issue is finding help on exactly how to do stuff. When I get (got) stumped on Access, I could search the net, ask questions in forums and read through books. You won't find that with Kexi. Case in point, this is the ONLY forum I have found for Kexi help. There's 426 TOTAL posts in this forum. That's not much to go on. The handbook is 60 or so pages. Don't get me wrong, I'm not expecting anything more. All this stuff is volunteer based. Folks contribute what and when they can. I think Kexi is awesome for what it is but just be cautious and understanding that when you get stuck on a problem, the answer may not be easily obtanable.

Hi, interesting feedback. I'll leave a note on a bit more general level: yes there is no multibillion company behind Kexi, the idea that repalces this kind of challenge is that like for Linux and KDE and {you name it} -- there are thousands of users behind the product, and users are contributors, not just customers.

Is there important difference? For many, yes. It's all about control of your future. If the current maintainers leave, any user can take over and continue. No billions in budget can assure this. Something to think about

You can make plugins and build business around them, that's also possible from day one.

Another possibility is to build database such as the specialized photos database and benefit from that. Kexi has to be customized, this project is set up here to assist in exactly such customizations.

It all depends, if one thinks more long term.

As for the number of forums, the answer is simple, it was easier to have one forum and thus mailing lists have been killed (other than the calligra-devel), leaving this forum as one place to go. Same for the user-contributed documentation, tutorials etc. - at http://userbase.kde.org/Kexi.

As for the handbook [http://userbase.kde.org/Kexi/Handbook], it covers 80% of features in a way how technical references do. One notable area waiting for contributions is documentation for Reports. More tutorial-like content would be also good.

You will not find scripting in the docs and that's because scripting is experimental and is planned to be completely replaced with the Right Modern Thing[tm]

Hello thierrya,
You have a very nice database (idea and the content of course). I'll answer your question inline below.

thierrya wrote:Will I be able to display pictures in forms by using a hard drive path saying where the actual file is ? (I cannot physically store the binary of 30000 pics in the DB)

Not yet, but this can be added relatively easy, and once it was planned as typical feature. If you get ability of compiling Kexi, you can get updates sooner when such feature arrive. Depending on official deployment on Linux is problematic in so fast moving software such as Kexi (people use Kexi 2.6 when 2.8 is arriving for example).

thierrya wrote:Will I be able to interact with the form structure on the fly ? (for example by populating a drop down list control based on a search work entered in a box beforehand ? hiding or showing controls depending on field content?)

Will I be able to script call to external programs ? (like launching google earth and make it display the place where current picture was taken? or run a command calling an external program that will make a thumbnail of images returned by a query?)

Not in general, because scripting is not there. The answer for this and similar questions is, if you get someone who customizes Kexi for you (probably in C++) you can get usable solution for now instead of waiting for scripting.

thierrya wrote:Will I be able to interact with the file system ? (rename files using db content ? create an XML or CSV file and write data in it ? change IPTC tags within a picture ? and so on...)

This would again be most flexibly done in scripting, so the answer is: not yet.

PS: Requests like this encourage to deliver the scripting module... first we need real resources to make it real. Every single feature truly extend applicability of Kexi.

Hello,
Thanks Jaroslaw and everyone else for your answers.

Yes, I also think that the community approach is better for the user in the long term, than the corporate one.
But I also have to recognize that it is slower, has to be.

I tried Kexi extensively last weekend, and though it is not ready to fulfil my needs (as Jaroslaw pointed out, no external pics and no scripting yet) I'm still impressed by the progress open source DBs made since the last time I looked.

I'm really sorry that my coding abilities are not up to the task: as far as code go I'm at most a scripter (if that) but in no way a real coder.

Nevertheless: I will keep an eye on Kexi in the future, you can be sure of this.
In the mean time: best of luck to all with this project.