12 Feb 2004 shlomif   » (Master)

Aarrggh! My Hebrew has turned into Question Marks

I edited a Hebrew HTML file with kedit, but forgot to set the encoding appropriately. The Hebrew was edited fine. However, when I closed it, I realized (after some experimentation) that the Hebrew characters have all turned to question marks. The actual ASCII question mark letter.

I ended up losing the entire document. What the hell? Fury... fury... fury.

To prevent it once and for all, I added this to my .bashrc:

function kedit
    LC_CTYPE=he_IL.UTF-8 /usr/bin/kedit $*

The reason I don't use an i18n'ed locale by default is because it causes my fonts to become uglier. How lame is that? <sigh />

DocBook continued

Well, Bob Stayton helped me and instruced me the how to modify to the code of the DocBook/XSLT itemizedlist handler to make it propagate the role attribute as a class. Now, it works and I'm happy. Still I think it's a huge mis-feature that role is not propagated for any element.

Version Control Article

My Version control article now has many more comments. Many of them were in the vain of trying to protect CVS and why it should still be used. I think this is not true, and new projects should not start with CVS. I expect CVS to be used in years to come, but be considered hugely deprecated just as RCS and SCCS are still used today, but are also deprecated.

It was referenced on OSNews.com (courtesy of Eugenia Loli-Quero who also contributed to the Better-SCM site and comparison), where there are more comments, and it also hit Whatsup in this feature in Hebrew. There it seems to form a discussion of the Linux Kernel Development Model, why it requires BitKeeper, and whether it's a good thing.

As a result, or just out of word of mouth, I received a great deal of E-mails about the Better-SCM site lately. Someone even sent an addition to put in Visual SourceSafe and CM Synergy to the comparison. (which I bounced because it wasn't a patch to the XML file). I have modified the pages a bit since then. Next, I plan to put its source on a public Subversion repository so people can contribute more easily.

I also started sending a Newsletter about Better-SCM to all the people who contacted me about it. ("Cathedral and Bazaar"-style) It worked pretty nicely so far and I received a lot of useful input. Someone, however, wished to be unsubscribed from it, and suggested I ask before I add someone to the Newsletter. I have to think about it a little, but think I'll go with unsubscribe on demand, given that most people are not very responsive.

Finally, there's now a new mailing list - better-scm-discuss. Several people have joined now, and there was some active discussion. That's good.


I managed to do quite a bit of work on Perl-Begin. I moved its source to a Subversion repository at opensvn.csie.org (which offers a very discreet and unintrusive service, albeit sometimes the connectivity seems slow), and then added a lot of content and changed the CSS somewhat (made it more modular and nicer). Finally, I created a newsletter for it as well, and sent some announcements to the people who contacted me about the site. One of them replied with some useful suggestions. :-)

In any case, I contacted ask and the rest of the beginners workers about contributing to learn.perl.org. Eventually, I found out that perl.org had a hardware failure and so work on learn.perl.org was stalled. What I did do was write a "learn/" section to www.perl.org, which will take some time until it is proof-read, and integrated into the main site. What I did there was simply link to the appropriate places on learn.perl.org and perl-begin.

I'm glad I could find some time to do a lot of really productive work. I missed hacking, and hacking on web-sites, while not being as challenging as hacking on code, is still fun and rewarding.


Check this message of Hackers-IL for a post I made to Hackers-IL in regards to the new and annoying Yahoo Groups limitations. (Hackers-IL is hosted there). There are three possibilities: accept the Status Quo, pay Yahoo Groups for getting a primum service, or move to our own service.

We decided that being hackers, the last option would be best. The problem is that several people have resources at their disposal, but none has the time (except me). So, I went out and contacted sun, one of the beak.hamakor.org.il's admins to host the hackers.org.il domain there. He told us we should send a description of what Hackers-IL is all about to the appropriate list and he'll see if everything can be arranged.

I wrote a charter, which was criticized as being too long to be effective and Arik Baratz wrote a shorter one, which was still very good. I ended up revising both charters and putting them online. We decided we are going to use the shorter charter for contacting the beak admins and for further purposes like that, and the more verbose one for putting on the future site.

The discussion was pretty long, and eventually we debated if it was off-topic or whatever. We reached a conclusion that it was just a "meta" discussion, which was a necessary evil, but that it was still annoying. This was towards the end when everything stabilized, so it just closed the discussion.

Yesterday, I sent the mini-charter in its final form to the appropriate list, with some explanations on what we need in the present and in the future.

Mozilla's Firefox

I'm glad the Mozilla people changed the name of their browser so it won't clash with the Firebird database. I also like the new name. It is almost certainly based on the Biblical story where Samson sets fire to the tails of 400 foxes and drives them to the Philistines camp to throw it into chaos and destruction. There's a nice symbolism with trying to do the same with MSIE, but I don't know how many of the target audience know the Old Testament well enough to realize that.

Still, the name is very cool, even without recalling the story. Of course, I have a feeling that it might start a new trend: Fire+[Animal]. Firecat. Firedog. Firedove. Firebear. Firepenguin. Etc. All these names are very cool, but one will have a hard time distinguishing between them. But who is John Galt?

How Good are my Technion Grades?

I posted a post to the Joel on Software forum about my Technion grades. This seemed to be a mis-calculation as many people took the opportunity to criticize me, claim I was no-good or lazy, etc. They obviously haven't been to the Technion...

Here is the creme de la creme of the discussion by someone calling himself "Caliban Tiresias Darklock":

Shlomi, just a bit of constructive criticism here... and this *is* supposed to
be constructive...
 You come across as childish, inexperienced, and belligerent. Nothing you have
 written interests me. What little sticks out in my mind is not positive. You
 sniff at the college for not having a standards-compliant web page, as if
 anyone really cares. You sneer at CVS for not having a few minor features, as
 if anyone really cares. You solve a few basic puzzles, and build a
 presentation generator that you use to create a bad Perl tutorial. Then you
 come in here and say you know you're a good engineer. 
 Unfortunately, you're not. You're just a competent Perl programmer. But
 engineers solve problems; if you expect to impress me with your work, do
 something that's relevant to someone other than yourself. Why in the world
 would I want to read your ramblings about SCHEME or run some program you wrote
 to solve mazes? I know why I would want to read *my* ramblings about SCHEME or
 run some program *I* wrote to solve mazes -- because I'm arrogant and
 self-involved. I love writing stupid little things for no good purpose, just
 because I can. Every engineer does. 

Maybe I should check the definition of "constructive" in the dictionary again...

Melissa Joan Hart - Pictures

You innocently type "melissa joan hart" into Google and search it for images and what do you get in the first hits? That's right - soft porn.

A few notes are in order. The Google search for MJH (=Melissa Joan Hart) is perfectly fine and clean and leads to various fan sites and pages. Secondly, some of the other pictures in the search are not provocative at all. Thirdly, the first pictures are absolutely horrid, IMO. I don't have a problem with sexual or even provocative photos - some of them are very nice - but these are done in very bad taste. (I guess ESR was right when he noted that most porn is disgusting)

Moreover, this is not entirely an issue of "child-safety", keeping "innocence", etc. I just wish that the Google images search was a bit smarter about this, as the Google regular search is. It is possible that it was the anti-thesis to these pictures that made them get a higher rank.

In any case I found a very nice picture in the first page of hits. Hart (fully groomed) standing among automn leaves. She looks a bit sad, but the picture is very beautiful. Having downloaded it, I noticed it was scanned and of low quality. So I decided to reduce its size using the GIMP. The latter simply opened it in %50 scaling ratio, which made it look very decent, so I just reduced it to this size. Finally, I placed the picture in my Konq+Xmms Virtual Workspace instead of a very nice picture of Gabrielle Anwar, which I had there for several generations now. (I guess it's always time to move on)

Melissa Joan Hart - Silly Etymology

Well, enough Cybersex for now and let's just discuss MJH's etymology. Her full name is "Melissa Joan Catherine Hart". Considering the fact that Catherine is a variant for "Catherina" we can see she has no less than 6 private names:

  1. Meli (short for Emily)
  2. Lisa (short for Elizabeth, or Alisa)
  3. Jo (the "Little Women" short for Josephine)
  4. Anne
  5. Cathy
  6. Rina (a common Hebrew female name meaning sing-song)

Six names for a girl who isn't even thirty yet. (!)


Well, for something a bit more serious. I had some kind of enlightenment in regards to Solving. Now, I simply say that instead of trying to bring the open-source world-like Distributism model into western-world-countries, I am instead going to propose that competing leaders will compete for controlling the countries instead of the "official" elected government. This will create a multi-tude of leaderships all competing for the same niche, and doing so logically and reasonably, instead of one leaderhip who cares more about his public image than about being a honest one.

I revamped the content of the site (without re-writing it from scratch). Then, I linked to it on the IRC in several channels, and also on a web-designers' list I am a member of. From the replies I received, it seems that I did not explain myself correctly enough, and people thought it was Feudalism or even worse. This means something is still wrong there. I still got some useful replies and input. Guess it's time for more revamping... <sigh />

Latest blog entries     Older blog entries

New Advogato Features

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!