Topic of the day:
Craig Mundie's (Microsoft) speech about open source.
At the beginning I must tell you that I do believe in
Karlgaard conspiration theory made public in
Forbes.
I'm almost crying, we are insiders. Most of us know what is Open Source, how it works and why is good. Unfortunately business leaders, does not. Most of them are taking Bill Gates and Microsoft as some kind of oracle, and follow their
suggestions blindly. It does not make any difference that
Linus is right, that Newton and Einstein has done more for humanity than any company. Business is about profit not about humanity, they don't *care* about humanity.
I guess, Open Source will continue it's way, out of any commercial and business interest, just like it was at the beginning. I'm onlyv sorry for companies who invested in Open Source, because their effort is shadowed by Microsoft. I would like only to mention IBM and Sun for their involvement in Apache projects or Intalio which made: Castor, Open-EJB, Tyrex and many other companies.
Myself I co-founded a
company just to be able to use Linux full time and do Open Source full time. If there was a moment at the beginning (1999) when I show a future for the company, now my hopes are fading away, Open Source in Business game is over.
Perens: "There are companies that see Free Software, especially GNU/Linux, as an interloper to be shut down, a competitor to be eliminated. Some of these companies have increased the rate at which they file new patents. It's not impossible that these companies and their business partners could start going after Free Software developers, en masse, with patent infringement lawsuits. Since essentially none of us can afford to defend ourselves, most developers would be forced to cave in, withdraw their software, and stop participating in Free Software development."
discover: "Fullerton hoped the transmitter would do what he had been told was impossible- send time-coded ultra-wideband electromagnetic pulses, rather than continuous radio waves, via the backyard antennas. Ever since a professor at the University of Arkansas told him that such pulses could not be used to transmit a clear signal from one antenna to another, Fullerton had been obsessed with proving they could. Five years later, here he was, standing in his yard under the moonless sky with a smile on his face, listening to his favorite band: "Waiting for the break of day, searching for something to say . . . "
There are two things I like about this article. First it proves that if somebody thinks something is impossible that is not necessarily true. Second the implications of the technology. Imagine a wireless home with 40 Mb/s transfer rater or an personal radio station.
May 4th, 2001
Great time in nature. The whather was beautiful as the peisage. Never thought how exhausting can be relaxation. Yesterday I wasn't good at anything: sleepy, tired, unable to get a decent thought. The funny thing is that I managed to sleep around 9 and half hours, which is definitely record for me. Maybe that was the reason. Pictures, available (as soon as I scan them).
Happy that we don't have to many holidays around here, still this month there is another workless day, 14 May, the official day of my town.
Weird problem, with JDK 1.3 from Sun and KDE 2.1. If there is any setBounds involved (and there is more than you can imagine, setLocation for example) the window or dialog is totally thrashed. This doesn't seem to apply to the JDK 1.3 from IBM, or in Enlightenment.
The consequence of this fact is that I dropped KDE 2.1 as my default window manger, and returned to my old love Enlightenment. This time, amazing things are finally possible. I run the Enlightenment as WindowManager, KDE 2.1 applications as Quanta, KWrite, Kedit and for sure the best of all Konqueror, which eats 1/3 of the memory consumed by Netscape and 99% of the pages are rendered just like by Netscape 6. And on top of this the Gnome panel which holds my menus and applets. Cool, real interop here.
Runned out of pipe tabacco. The weird is that when I have you can also find it in every shop, but when I don't have and I'm on my way to buy, there is none left
. Is this a sign
What's new on the brain bench? Well,
miniXML.
May 3rd, 2001
So new look again
, It just was relaxation time. As I see it is starts becoming a habit. Every couple of months redesign of the site
. Actually, I hope you like it more, for sure I do, it's more an evolution. Except for the minor aspect changes, now, the web log has some real functionality.
I will not put it up, until I see is stable, beside I would like to add a few more candies.
Today (it's 02:00 AM
it's a kind of labor day, so, everybody is free. Time for some beer, nature + "relaxation". It's just funny do to celebrate labor day with relaxation. *I AM WORKALCHOOLIC*, and I don't mind that
May 1st, 2001
This is so funny, or my humor is bad. I've got this while searching for an Linux client for MSN Messenger
March 24th, 2001
Even if it is hard to believe, I'm still alive
. So, time for excuses: I've been so busy. Of course nobody will buy that. The truth is that I've been busy building something, which I like to think, needed a lot of creative work and the time frame was kind of relaxed, you can't rush creativity. It required such an effort
that I missed my schedule with around three months. I know that 10 minutes is not so much time to spend for this page, but my big ego would disagree with you.
So, events! Well, in November or December last year I gave an
interview for the
Application Development Trends magazine, and then in the same time we had a TV crew from
NHK public television for some Internet related documentary. What really impressed me about the TV shooting, among the fact that they filmed around 2 day/10 hours per day for around 5 minutes of broadcast, was the kind of focus they where having. Starting every morning at 9 AM till 10 PM, no coffee breaks, no lunch break. No wander Japan is where it is (or was ?). I never got the tape with the show, so I have no clue what was that they actually broadcasted
. And after that came the relaxation time, spending the New Year's eve up in a very remote area of the Romanian mountains. Hope I will have the time to scan some pictures, because
the peisaje was really impressive.
I'm the kind of guy who likes gadgets. So, now I have an
Compaq Armada M700, an
Psion Series V and an
Nokia 6210 mobile phone. In this idea now I'm mobile (TM) and perfectly interconnected, you know my notebook talks with my phone, my phone talks to my Psion, in one word everything talks with everything. But of couse one is theory and other is reality. The Nokia Data Suite is a crap piece of software, which really doesn't matter since I don't use Windows anymore (I don't have to much against Microsoft, so that's not the reason), but under Linux there is little software available and what exist is not for my devices. Bad luck. I would write them, but no time. Anyway the morale of the story: there is no such as smart devices if you don't have the smart software.
The Open Source and Linux in Business game is over. This hurts, because I believed that this model really offers (if it's done properly) a plus to the client. It's true it's hard to find a revenue model in the movement, I'm not even sure if there is one. I agree that Open Source might not be appropriated to every type of software but for some the benefices offered to both the companies and to clients is huge. What is most annoying is that I feel that not necessarily was a model which didn't found a place in the business sphere because it lacked real value and therefore died. I have the feeling that it died because of bad conjuncture (the dot-com phenomena, and the problems with the American economy) and because of an aggressive marketing with no real value behind.
The truth is that now maybe the whole software industry might get a bit calmer. I always had the feeling that what happened lately was not OK. We are not a special industry, and the software engineers are not a separate species, who deserves special thratement. We are workers like peoples in any other field. We have to produce tools, which other people will happily use. Customers don't care if it is JNDI, J2E, Java, C# or Net based. They are still using Cobol and 10 years old applications because of the simple fact that they work and does the job which is supposed to do. Wake up people.
And of course one game ends another one begins. The name of the new game is:
Hailstorm, .Net, SOAP, and
Web Services. A new madness begins. Aldo not sure about the benefits of the
Hailstorm for the average user (except that if he would use real cash, his wallet would be a lot lighter
, the new game will change a bit the Internet, which might be good. Just like the Open-Source/Linux game revitalized the movement (lot's of new software, lot's of new quality peoples) maybe the Internet will be a bit more after the new game.
Related to
RUE, without any update since October, people begin to be more and more interested in it. Great daily activity and downloads, that's good, the 2 year prediction seems to be true.
March 23rd, 2001
Eazel, a company in which
many (including myself) has put their hopes to improve Linux GUI and usability is
going down.
Unfortunately not many Linux
companies remaining, I'm wandering when
Ximian
is going to close their gates.
Even if the unfortunate dot.com bang wouldn't be associated with the OpenSource and Linux,
Microsoft war machine is all over Linux these days. There seems still that one thing they can
not understand. Yes they can demolish Linux (OpenSource) based companies (what's left whatever)
but they can not make Linux disappear. Linux is not a company and doesn't depend on corporate
financing, Linux is volunteer work and passion, you can't fight that )), in fact I have the
feeling that they will grow our numbers ).
What amazed me this morning that I found an
article about Linux in an
magazine. Then I show
the date August 1999. Which magazine have now the guts to write about Linux, even the truth ?
Still I learned one amazing thing from it, Jon Hall, the executive director of
Linux International (an organization which promotes Linux)
is an manger at Compaq !.
March 14th, 2001
Spent some time reading from
Joel on Software archive, I like the guy (not to mention that I found that we have the same idea about how a
software company should be, sure, he has more chances than me, not everyone had the "luck" to be a Romanian). He is simple, pragmatic, and I found the things he discuss conform my experience. Also he does not try to build a whole theory around his ideas like some do.
Still related to software management. I read a
ebook about
Extreme Programming. I'm in no position at this moment to critique anything related to
XP (since I didn't even finished the book), but even after I listened the
interview with Ron Jeffries on
technetcast, I still have trouble understanding the
two programmers one machine concept.
I have some concerns with this issues:
- Programming is an intimate thing, and I'm not the only one who thinks this way. I have no problem with others watching my programs, but I can not work if someone is staring at me. At least not that efficient. However I do agree that reviewing is absolutely necessary. So after a portion is considered ready, somebody else look at it. Also I think that sharing implementation plans (before the coding) are good.
- Productivity. You will hardly find two programmers who have the same programming rhythm. If the "pair programming" might work at the beginning, in time one programmer (the one which review) wich have a more static role, will lose rhythm, become distracted, and finally his participation will end up being only physical. This is even worse, because you will consider code as reviewed.
- If I understood that right, the major reason for this idea is that when someone start the coding, he loses the overall aspect of the problem and concentrate on syntax, and other issues. I don't think this is true, and I base this affirmation not only on my experience but on talks with other developers. I leave this issues on my development environment, who does auto-completion, special indentation (I notice when I forgot something), and the compiler.
For anyone who wants (dreams) about joining an open source project here is a
thread from the
Tomcat mailing list, in which the real big ones describe their background and what they consider as necessary.
The conclusion ? "
You guys all make it sound like much less pain than I had previously thought.". And what it takes is:"...
maybe it is not as much pure expertise as it is willingness to learn and contribute to the project...".
Open Source ! Last week, again on the
Tomcat mailing list was an interesting
thread on the Open Source licensing issue. I finnaly understood (exactly) what is the difference between X kind of licenses (X, BSD, APL) and GPL licenses (GPL, LGPL, ...).
What is about ? To quote a
kuro5hin posting "GNU's "freedoms" taking away MY "freedoms". Basically
JBoss (GPL)included some
Tomcat (APL) code. Perfectly right (the X kind of license basically says, here it is my code, do whatever you like with it). The issue started when the JBoss guy's offered some code to Tomcat (nice gesture). The conclusion ? Not possible. Including GPL code in APL code would result in APL changing in GPL (not good).So here we are both teams are fighting the same war, but still one has to reinvent the wheel all the time, and not because peoples are against sharing code (on contrary), but because the legal problems faced by the two different licenses APL and (L)GPL. My questions is, deep inside, isn't this just a personal war between the two leaders (
RMS,
ESR).
This war is really not helping our cause.
With this occasion I also found a good (human understandable) article on licenses, named
"The Open Source Definition", written by
Bruce Perens who also happens to be the initiator of the famous "Open Sources: Voices from the Open Source Revolution" book.
This was a big posting, sorry
November 4th, 2000
Well, the bug fix release of
RUE is out.
As usual lot's of downloads no feedback. However now I'm not that confused about that, since I found a
thread on
Slashdot on this
subject.
Only a few peoples who makes the download actually install the software, and even fewer whom install it give feedback or contribute. The idea is that one project needs around one to two years (I'm not sure this is true also for Java) to have contributors and peoples interested in the project.
The big problem with
RUE is that I really don't know what should it be, considering that the new standard in monitoring (and management)
JMX and the way it can use SNMP
and TMN as protocols, there are plenty
tools to do that even open source and Java ones. The way it works now is obsolete. This would not be an issue, but the amount of time I have to spend on it is definitely small compared to other projects, so features will come in quite slow.
Anyway, I won't give up, people have to have options where to choose from, just started to have a few ideas how to make
RUE useful and unique, even if the development will go a bit slower. But first specifications, because I don't want to start the development of the new release until I don't make a clear set of specifications, that's why feedback would be so important.
What else ?
Tyrex, playing around, and this for multiple reasons. First to have the
Tyrex sensor I promised, second doesn't hurt to know a bit about transactions. There aren't to many documents out there about this, but I managed to find a few I liked. In case you are interested, here are the links to the documents:
November 1st, 2000
I plan that till Mon, 16 2000 I will have the 0.52 version of RUE out. It does not contain any improved functionality it is only a bug fix release. Well maybe one thing in plus that the sensors are not anymore integrated in RUE but separate modules. Also I hope that I will have the Tyrex sensors by then.
October 11th, 2000
Spent my weekend with some RUE bug fixes, the patch version 0.52 it's on it's way.
Spent my week trying to solve the "certificate" problem for signing RUE (being able to run as applet in Netscape). I just start to hate more and more those Netscape guys, I don't know why are standards for if nobody respects them. Neither Microsoft does but at least in IE I can completely deactivate the security and run RUE. I just had to insert all over RUE special code just for Netscape-applet mode, when rue is mainly intended as an application, I really hate doing that. But the problems does not stop here. Now Netscape does not trust the Sun packages like jndi.jar and rmiregistry.jar, because are not signed. The ugly part is that I either force the user to consider all the classes signed (case in which I don't need a signed version of RUE, off course that if it finds one signed jar will not work, don't you love Netscape). The second solution would be to sign jndi & rmiregistry jar's which I don't know if I'm allowed to do.
About Netscape and test certificates after a very frustrating week I finally managed to hack a method to produce X509 test certificates for Netscape valid for as long as I want. And I'm sure I will write it down in the next day's just to make the life easier for others who might have the same problem.
October 9th, 2000
Next Posts
Previous Posts
