Older blog entries for sdodji (starting at number 42)

My adsl connection is working again :) I can now spend time on irc.gnome.org again :)

Did no hacking on MlView yesterday... oh my, how lazy i am.

The mlview code is now based on a shared lib. I've called it libmlview. For the time being, it contains all the core code of mlview. I've also updated the mlview.spec.in to make the rpm build process aware of this modification... looks good.

I will know have to focus on cleaning and stabilizing the api of this new library.

Life

My adsl line is still not working :( dunno exactly what is going wrong.

Diane was here this week end. I didn't spend enough time with her but it was better than nothing.

tatata !!! I've just finished the rewrite of the "save-as" feature and debugged the "rename-view" feature. The new mlview_editor_save_xml_document_as () function is definitely shorter/clearer/simpler than the previous one.

Hmmmh ... I am starting to like this new document/view code.

Finished a first version of the "rename-view" of mlview yesterday night. It is a bit like the "rename-buffer" function of emacs.

I still have to rewrite the "save-as" function because of the new multi view architecture of mlview. This would simplify the function.

After that, in a couple of weeks, i will maybe have the time to work on the plugin architecture of mlview. That would be an opportunity to cleanup the mlview api.

Not so many new lines of code written so far. Have to hack on it again this night.

Life

Dad is leaving today.

Have to make sure that my telco has connected my adsl line. These guys are not serious. They take the money and then, they take their time.... anyway, better late than never.

Debuged the "set-content" fix i wrote before. It seems to work now. I am working on the "rename-view" feature. I hope i will finish that by the end of this week. I should also update the "change-attribute" feature to comply with the new document/view code.

hmmhmm another busy week.

life

Went to see daddy yesterday. He is really cool. We are getting closer. I should call him more often.

Yesterday, the guys of the phone company broke my phone line. They were trying to plug the line to the adsl system and they broke the line... damn. I hope they will fix that today cause i really need to commit my code into the cvs repo.

Went forward in the fix of "set-content" bug fix. The code still needs to be debugged.

Life

Went to my karate training yesterday. It was very hard for me to stand the pace.

When i came back home, i didn't have enough time to hack yesterday... anymay, some few lines of code are better than nothing.

I can smell that this week will be another busy one.

When forward in the document/view enforcement. The last code i have written still need to be debugged.

Life

Did not have time to hack enough this week end. Still don't have any fast internet connection at home yet.

Anyway, I've managed to use my slow old pcmcia modem connection to post this diary using gnome-ppp... better than nothing. I will also use it to commit my code into the cvs this night.

The "automatic-subtree-adding" feature code is there. I would say it's in a beta stage. The code is in the cvs branch devel-0-0-2.

Now, i will focus on the next step of the document/view model compliance: allow multiple views on the same document model... sound interesting.

Life

• Diane called my this week end between 2 and 3 am. I was pleased to talk with her. Her job is tough but i now she will overcome it.

• Eric and Gaelle helped me to move my two other PCs to my new flat. I Still have to get an internet connection soon.

• Helped Gaelle to move to her new flat. Her new place is rather cool. It was a bit difficult to park the car but everything went almost smoothly.

Debugged the "automatic-subtree-adding" feature a bit and started the first version of the "automatic-attributes-adding" feature. Still have to handle the tricky attribute value types such as IDs and IDREFs.

I also have to fix a pb in the inmplem of the mlview_parsing_utils_build_element_name_completion_list () function. The "case ADD_CHILD:" is bugged.

When the node is not a leaf node, the function must cycle through the children node and invoke the function xmlValidGetPotentialChildren () for each child node pair.

Hmmh, i hope i will have the time to write down this fix tonight.

Life

I will be very busy this week end. I have to help Gaelle to move to her new flat.... damn ! all my friends are moving ! It hope i will have the time go to may Karate training.

Fortunately, i have the nights to code :)