I must also find the time to work on my new css parsing library.
I must also find the time to work on my new css parsing library.
I finally released mlview 0.0.3rc1. It fixes a lot of stability problems.
Users of this version must also upgrade to libxml 2.2.26 .
I still have to deliver some other small fixes before 0.0.3. After this, I must kick my ass to start the gnome2 port.
I've been playing with GNU/Linux for 6 years now but I've only installed slack, RH and Mdk... now Debian. Which one will I like most ?
After this release I must force myself to download the sources of gnome2 on my brand new debian system, compile it, admire it and start the porting of (damn even apm works!!) MlView to gnome2. That will take a lot of work since I have to totally rewrite (at least) the MlViewTreeEditor and MlViewNodeEditor widgets which are heavily based on the GTKCTree and GTKCList widgets ... Okay, these widgets are MVC pattern compliant but I have to rewrite the models and the controllers still.
Went forward in the support of iso latin1 edition. It works somehow. I think I should release it ...
I am now working on the foundations of the css support in MlView. I therefore have started to write a standalone C implementation of the Simple Api for CSS. I am expecting some great hacks on that front.
This CSS stuffs are certainly going to lower the release frequency of MlView...
I went to the last "first jeudi" of Paris. It is a friendly monthly meeting of more than one hundred Free Software users and programers who live in Paris. It was great. I met a lot of cool guys there. Going to work the day after was much less cool.
Hmmh, It seems like the people who certified MichaelCrawford as journeyer when he was known as goingware are quicker to certify him than those who certified him as master .... We still have to wait a bit for the system to reach its stable state but if the number of such phenomenon (persons changing their alias) increases, that would be a great test of the reliability of the advogato trust system.
MlView Hacking Still busy with the external encoding issues. It's Not that it is very complicated but I want to handle this aspect gently.
Work Hummh, the atmosphere is getting worse and worse at work. Our managers now want to send the activities of my team offshore. Hmmh, bad weather expected ...
Question. I definitely hate working in a place were the management doesn't understand that people who write software should 1) maintain it 2)document it so that further maintenance can be made at least possible. Instead, the management only insist on the code production part which is actually the smallest part of the software lifecyle. The greater part of a software lifecycle actually being the maintenance and the evolution. But they totally ignore the good practises that will ease code maintenance and evolution in the future.
They went to an extend where they created some division where people do *only* maintenance.
Readers should keep in mind that there is *zero* code documentation ... so an incredible amount of energy is spend to fix the bugs ...
I do think that one day, the company will just collapse because of their being incapable to roll out new products beacause all the energy is spent on maintaining unmaintainable products. That is very bad.
And as time goes, things are getting worse and worse.
Does anyone share the same experience as me ? Does anyone know how to get out of this spiral ?
I Have been busy last week with some family issues. Now, I am back to work and free software hacking again.
Started to work toward the next release of MlView which should be 0.0.3xxx .
Right now I am busy with the support of isolatin1 edition issues. In gnome1.xx I think translating isolat1 strings that come from the text widgets into utf8 before sending them to libxml2 and vis et versa will be enough to support isolat1 edition. However, a more general approach will be usefull for the truly i18n proof text widgets of gnome2 .
After that I hope I will start to work on a text oriented DTD edition view based on the GnomeCanvas. That will probably be a quit big/exciting development. Hmmm, another busy automn/winter is coming.
Long time since my last diary.
I am getting more and more overworked these days. I definetely need some vacations.
The general economic conditions are getting worse and worse. I really don't know where we're going. A lot of companies are still laying people off. Tough time.
Released MlView 0.0.2 last night. It fixes a couple of bugs since O.O.2-rc4 and adds a lot of new features since 0.0.1.11 .
An article of unixreview talks about MlView :) . That's great. Anyway, as always, a huge amount of things still need to be done. But unfortunately, there are only 24 hours in a day.
I've also started to work on a suite of tools to manipulate java class files... nothing to do with xml edition but still interesting. A lot of things to do on that front too.
I finaly got a somehow working version of my gtkdoc2doxygen.pl perl script. I have then converted the gtk-doc style comments of MlView into doxygen style comments. The extracted documentation is available here. I still need to achieve a _huge_ amount of work on it to make the documentation complete. Oh my, that's boring no end but very usefull if i want other programmers to understand and reuse the code... busy coming weeks
If you are a software dev manager and you are planning to buy a software versioning system, STAY AWAY FROM CLEARCASE . This product is _really_ a piece of crap written by true wimps. It is damn bloated, unstable (when it hang, it is likely to totally hand your system), needs an army of administrators out there to make it run and last but not least, it heavily relies on your network to work properly. On unix, they have a crappy merge tool that you are _obliged_ to use ... (yeah, back to stalinism).
You would better ask you developpers to know what tool they prefer before spending that much money into such a dollar sink. One think is sure : don't listen to clearcase sales people. Ask your programmers instead. If they don't know clearcase, they should stay away from it.
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!