Skip to main content.


On 01 July, 2004, the Israeli Perl Mongers held their regular monthly meeting. The program:



  1. Gaal Yahas
  2. Gabor Szabo
  3. Hezi Golan
  4. Jason Friedman
  5. Maoz Etziony
  6. Mikhael Goikhman
  7. Oded Resnik
  8. Offer Kaye
  9. Ran Eilam
  10. Ronen Shemesh
  11. Semuel Fomberg
  12. Shlomo Yona
  13. Thomas Maier
  14. Uri Bruck
  15. Uri Itscowits
  16. Yair Mahalalel
  17. Yuval Yaari

Here we go again. What a nice meeting. The usual gang was there + a couple of new faces, whom unfortunately I did not get to meet.
Shlomi Fish was not there, because he had to get some degree. This was strange. Many people were amazed that some degree is more important for him than a monger meeting. So of course there was no book review.
I missed last meeting, and almost forgot how much fun this can be. We started by diving straight in to UTF-8. There was no motivational speech by Gabor, but despite this morale was high. How long can it remain so, without his fiery words to enflame our hearts?
Shlomo Yona went from "SHALOM OLAM", to the darkest, most hellish corners of the Unicode idea. He showed us the points where Hebrew handling can fail in our code, and explained what is happening. He went over many potential problems, some of which were only barely solvable.
He talked about almost all aspects of Hebrew I can think of, from the character encodings and sets, to regular expressions and tokenizing words and sentences with Lingua::HE::Sentence. Apparently, all this is somehow related to his research.
There is much knowledge in the group of the technological history of Hebrew on computers, so there were many digressions into IBM vs. Microsoft charsets, Einstein (the word processor), Linux Hebrew history, browser history, and more. This really should be written down somewhere.
Then, the customary eating of burekas. Please get some mountain-berries next time. Oh, and nobody likes the poppy seed ones.

Then came Maypole. Is it a drinkning-and-dancing holiday in the land of our leader? Yes, but also other things. Gaal showed us what is, without doubt, the simplest way to create a web CRUD UI over relational models. Then he showed us how to customize it, with emails that can be hidden from the public, and extra fields. Eventually we ended up with an almost functional real estate rental web site. And it does seem to take a little less than an hour.
There were few interruptions, and I think I understood everything However I was left with a feeling of being shown some black magic. I would like to hear more about how it all works.
Both are our top resident speakers, so you can be sure the examples were clear and the drama was intense. The Unicode topic in particular has potential to be a complete bore, unless you come to the right monger meetings.

The highlight of the evening, as usual, was the social gathering. Because the Godfather have a new routine (they look in for our meetings dates, then at 10PM they turn off the lights and hide under the tables, so that we will think them closed), we went to a pizza place, which will remain nameless. It is a 80s style Israeli pizza place, long before there was any competition or knowledge of the domain at all. This was the time when the mere act of eating pizza was considered cool, but before the term 'pizza' was precisely defined. The pizza sucked, and I was hearing nothing good from the sandwich people either. The conversation, a usual, was flowing as smooth as Goldstar.
Gaal told us about Unicode security implications. I thought I understood at the time, but cannot explain it right now.
Shlomo Yona told us of a problem he has. He needs a tool that will allow a linguist with no XML knowledge to annotate his linguistic data, in XML.
We talked about this for some time, and could think of no good tool, or even how such a tool would look. Of course the problem was that we did not use Triz. Not that I know too much about it, but I am certain that it has some kind of conflict-resolution rules for Shlomo's engineering conflict: The user must be ignorant of XML, BUT he should be constrained by some XML schema and the data is in XML.
We started from generalizing (SOME schema). Once we looked at ONE schema, things became clearer. We started looking at the STATICS of the data: how it looks, is it recursive, etc. Because we are designing a UI task, we should have started looking at the DYNAMICS of the data. We started from thinking about the data: how will it be stored, changed, etc. We should have started from the user. Once we applied these 3 rules, we all agreed we have a solution: write a custom UI (we even designed it) for the specific XML transformation task.

Gaal told Mikhael Goikhman and me about Getopt::Declare. Migo was not impressed. Later I looked at it, and thought: woh, this is advanced. Much to my amazement, As I slowly scrolled down the POD, a chapter called 'Advanced Features' popped up. How about setting ranges in one command line option? reading lots of options from a csv file? My head is hurting. Must try to stay on the GUI side of things from now on.
Yuval and migo got quite a beating about the editor holy war they started. Migo's defense was 'he started it'. Yuval did buy him a beer though. And these are just the issues that I managed to keep up with.
Hope to see you in the next meeting!

This meeting summary was written by Ran Eilam.