Notices

One thing that might be nice is to change the highlight color settings for the currently selected item on the Input tab. White on a bright teal background makes the items text extremely difficult to read. In general, light colored text on light colored background is more difficult to read than on a darker colored background.

Is the font size on the Input tab staticly defined in the app?

The description was not exact so I assume that you press the shift key and then afterwards the alpha key. This currently in deed does not work. WeightJinni uses libIllumination, which is an non-Gtk GUI library and thus has to handle keyboard stuff directly. It seems like the slight out keyboard does some interesting thinks libIllumination cannot or at leats does not correctly handle (yet).

Not that pressing shift and the alpha key at the same time does produce the correct keys (at least on my german device). I admit that this is a little bit more tricky.

The underlying libIllumination tries to mimic the Gtk style used on the device. However while this works on desktop systems for list selection this does not seem to work correctly with the current standard theme. Font color should be black in this case instead of white. I will take a look at it.

libIllumination does use the default fonts for proportional and fixed characters. However the was an bug that made the font size smaller
then the default. So expect bigger fonts in the next version - and I hope that then the fonts will not be too big :-/. If this will be the case I will have to add some font size configuration option to the application.

Did this ever happen (posting the converted database for WeightJinni)? I'd like to try it, but the .jar converter doesn't work on my Mac (OS X 10.5) either, giving a Java version error.
-Jon

Yes, I managed to generate an WeightJinni import file and import that into WeightJinni. I however found the result inconvincing since the database contains to many entries to be useful and I found no simple way to for example reduce the numerous entries for variants to corn flakes to a usable subset.

You can find the sources for the (very simple) converter here:

http://illumination.svn.sourceforge....htJinni/extra/

Please test it yourself and make your own opinion. If you have changes to the source (addapt to your Mac problems) please send them to me so I can add them to the repository.

Any other usggestion or improvement is still welcome.

It is not a problem to package default data or offer web space for manual download, the problem is that I need data that is free. I'm sure as soon as one person starts to share its data things will gain speed!

There is also a new version of WeightJinni in extras-devel that adds easier to use buttons to the tables instead of the popup menues. I'm also interested in bugs reports further suggestions etc...

Gruß...Tim

Bump...
Just wondering if there is a file floating around yet with the items to be imported. I am very slowly adding items as I eat/drink but having something readily available would be swell If you have been using yours for a while and have a decent list please post that as well. Lets grow a list

Some thoughts.

I am currently using http://www.livestrong.com/thedailyplate/users/myplate/ for my diet tracking needs. It can track foods, excersize, draw graphs, has some shiny charts, and with its help I've lost 10Kg or so since october.

This is an ad right?

Well - not quite.
I have come to the conclusion after 50 days or so of use that the site is coded by ******ed monkeys, with little idea of how to actually use databases, and it has so many little niggles.

The core of the site is a - largely - user contributed food database.
This is however also the main problem - start out with the fact that foods may have multiple regional variations, and that people don't always enter foods carefully.
Then there is the fact that foods being edited means that all users that have historically used the 'old' food value now have screwed up charts.

Rather than coding, or doing something productive, I will now describe:
My ideal diet program. (I have some awk scripts that attempt some portions of this - but nothing like this)

I hit 'new meal'.
I slide open the camera, and point it at the barcode on a bag of onions, It beeps, and I shake the phone twice, to indicate I've used 2.
I grab a couple of cans of tomato, scan one, shake again.
Scan a can of curry powder, weigh out 20g, enter 20.
select 'meat' - lean pork is the third on my list as I use it often. enter 200g as I slice some off and put it on the scales.

...
Now, I get a figure of 1600 calories for this at the end. I edit this to subtract 20g of fat, which I've skimmed off the top, for a total of 1500 calories.
I kmow I want a portion size of around 350 calories, so I set the usual portion size of this to 1/4 - it will last 4 days.

Now, I have a new food 'pork curry', with 350 calories, 30g of fat, 15g of protein, cholesterol n, ...

I create a new meal, and add one 'pork curry', and scan a bag of rice, entering 100g.

Save this as 'curry and rice'.

I now for the next 4 days just add 'curry and rice'.

After the meal, I feel a bit tired, so I tap on the 'coffee' I added in the morning. and add another.

The food database is stored 'on the internet' somewhere - though it can be downloaded too.

A food not in the database occurs:
The user scans the barcode, and it says 'not found'.

A) They check it's not a mis-scan, if it's not, they take three pics, one of the front of the pack, one of the back, and one of any nutritional information on it. They choose not to enter any more than the portion size, and the calories, but they've selected 'auto-share' - so these pics get uploaded, and anyone can type this in.
B) They simply enter the food name and calories/portion into their local db, they aren't interested in sharing.
C) they type in all the info, and share it along with their pics (so anyone can point out an error)

I look at the overview of the day as I'm about to go to bed, and see that I'm still 100 calories down, so I snag a bananna on the way.

The next morning, I look at the totals for yesterday, and find that the onions diddn't count towards my budget, as I hadn't entered a price, so I tap that in, and find I spent 4.20 on food, a little below my average for the month.

It also says I'm low on cheese, so I add that to my shopping list.


Of course, some of these are quite hard to implement.

Does it work on N900?
I tryed... but there was some kind of database error and there was no possability to enter numbers

database errors on first start are normal as there is no database. If you have problems after the first start this is not normal. What do you mean by "no possibility to enter number", there was a bug with numbe rinput, that at leats was fixed in libillumination. I can build a new version of weightJinni to assure that the new library will be used.

I'm still unhappy with the user interface under the N900. The New approach of handling sub windows/dialogs does it make difficult to copy entries from the database to the diary. I'm open to suggestions to improve this.

The other problem is, that there is still no common database, so everybody has to type in everything himself. I have some ideas about that, too, but that would be bigger changes.

I'm still supporting the application but currently lack a bit time to make big changes. Howeever that will change as soon as my other applications are in shape.

no possibility to enter number <-- thats correct. I cant enter numbers