1: Right man.I was there when it started. At German Perl Workshop this march in Berlin Richard ignited with his inofficial keynote a lot of controversy. All what said wasn't new or IMO just opinion or chatter/not relevant. Later I spoke with him @ the social meeting in the computer game museum. (seriously, is there a better place for such an event?)
During our conversation I found out: he listens to people, he really loves Perl and he's the right kind of Person to do that, with the right experience set. Even if I don't share some of his fews/considerations what is important.
2: Catch the young (in spirit). Sure nobody like to be agitated or fed some PR bullshit (at least to realize the damns lies afterwards). So why we do that to others and our self. Because we need to reach out to the youngsters who are not experienced enough to see through the stupid badmouthing about Perl (as stated by allison too). Its about giving people who would lovingly join Perl community that chance - nothing else (as far as I am concerned).
3. Better Docs. Is I'm in the business of writing Articles, Wiki, Docs for Perl I see here a lot of improvement possible, because i see here more people than in any other doc effort i joined and having good docs and onramps for newbs and pros is an important part of an good image of Perl in the net and thatswhy also a concern of Pr.pm. We plan also to translate good free boogs such as modern perl and the camel (yes, you pay O'Reilly for the paper).
4. Turn the tide. Frankly, I love mst BUT dont share his notion: "be glad were not in the hype anymore". I mean right but I'm also a bit done with justifying my preference for Perl. I mean so stubborn that i will not get annoyed, but wouldn't it be nice if that just stops and Perl is recognized as one of most interesting languages (for some purposes). It's possible. Lets do it.
P.S. I'm in Bicycle.pm because I like Thomas too. :)
The main idea of DDD was:
reasons were:
I loosely rewrite Kephra with that and I still like it. However there are some flaws. Quick movements in API are still painful. And that proposal didn't even cover the real idea that is behind all of that. You do it to make great software which is well thought through and is not lacking something that you have to fix after shipping.
To achieve that i want to have a style called CP: complete programming.
First of all it doesn't sound as attention grabbing and hipster lame like like "extreme programming". I'm not a terrorist, I like people and if I would be one I would never admit and act as a harmless, bit naive guy. I just like to program but I'm not a nerd and I'm not only fully aware that life is bigger than the box I stare at now, I practice it daily.
Secondly its not only about the docs. You need configs, makefiles, key bindings, and and and .... So when I think of a feature I have to implement all these aspects before moving on. So you have more possibilities to see consequences and polish that feature that its fully usable instead of having to many half made parts.
That's is not the same as making small prototypes. Sure I have them in the sp2 branch too and there is also a larger more overall plan I want to meet. But I always deal with one section of the program and make there one feature but with the whole picture in mind.
Thirdly, this contains a lot of common sense and in your head calculation which can hardly be fully formalized. However I try to work on that to having something to show, maybe already at GPW in Berlin.
You might say now: but I do this already! Great you sound like a reasonable person. Write me your ideas!
Thanks a lot.
PS: originally I wanted to call it holisitic programming: HP. But yeah its sounds like a company already there, too elitist and frankly the first i thought of: holistic - thats sounds like ken wilber and alike and i don't like them very much. These scientist who write about mysticism and meditation feel a bit creepy to me. I prefer to attend the freak show and live life to its fullest and not stand in the corner take notes while others having the fun.
]]>Another area of current development is my layout helper class which will be later a separate module, helping programmer to make GUI with less keystrokes. The ideal testing field for that is the already started config dialog - something old also Kephra never had. There are a plenty pages needed to be styled, that can change quite often.
If you want to play with Kephra just pull from hg repo (link above). No fency dependencies beyond Wx, YAML, Moo and File::UserConfig. More news will follow and my head is so full of ideas.
]]>Don't worry about mistakes my articles are far from being perfect and so are cromatics, randals and so forth. its like giving talks. you start at perl mongers meeting and it sucked badly. but it was a bit fun too so you did another and you got better. I think its not only important to have a CPAN where also a lot of docs are hosted. But we need also scholarly culture were things are discussed which go beyond blog posts, even if there are sometimes of nearly mag quality.
In a magazine are already filtered topics (read almost only the currently important stuff) with a strive for excellence. That spurs other articles, more discussion on deeper level, some part may even move into docs, improve docs. Its the equivalent of a YAPC talk: more audience, more serious preparations and greater immediate impact and a different kind of fun than just talk to your 5-25 pals at the mongers. And you reach a different audience for the overall benefit of Perl. I already know some of my readers who don't track the blogs and don't go to conferences but read the $foo. Plus printed paper is a form factor that some people enjoy most.
so please consider to write about what your best in or care most about and if you have some problems or questions - please write me. Maybe I can help (I written now 53 technical articles and getting a glimpse how it's done).
]]>I know, some think of mst as a bastard with maybe an heart. Its rather the other way around if you know him better. We're quite lucky to have him.
Yes the travelling wasn't that cheap but its so welcoming to arrive there, I almost forgot that its not the norm til I attended two years ago some none-Perl conferences. I found trough Perl conferences people I consider friends in several places and pals in almost every country around and I don't want to miss that. YAPCs are much more overloaded and good just to see people, say hello and recognize new topics. For really learning something are Workshops much better suited.
And its just a great experience (like they say in advertisement just in real) - thanks michele and team. This may doesn't sound special to some but its what makes life really good and rich. Last year it was the small round composed of matt, ingy, gaby and me - just unforgettable evenings. And of course with the wonderful locals which are partially gone to England now. Maybe i should try the London - soon to be renamed into English - Perl workshop too next year.
P.S. don't forget
Lets be honest most people I know see it like my little pet project ("good you allowed to upload to CPAN" but i will use vim/emacs/.. anyway), some like it enthusiastically but that's it mostly. But Kephra is not about to have a Editor written in Perl - Perl is just IMO the best tool for the job. It's not about certain features like nice integration of your favorite Perl tools, not even the TIMTOWTDI interface (having the best of vi, emacs, jedit, komodo, notepad in one place). My deepest satisfaction is to think things through and offer solution that go beyond mostly known and are deeply satisfactory (which is the Perl 6 approach of things too).
As the rewrite is still in development I have for you one examples thats currently being implemented (not fully yet). Its about the docbar - the line of tabs to select the current document. Everybody understands this interface and it can be well handled by mouse and keyboard (
But if you look at vi or emacs, which both got to the tabbar very late you see the problems. If you split it in too many parts it gets messy and the handling, switching rotating and jumping between the views can lead to quit complex manouvres. So what to do. I think jumping between views should be done with (
O course you want to also sometimes change the order of things. Even in Kephra 0.4.n you can move tabs with keyboard just by adding the
Which leads to the real problem not solved in jedit nor even eclipse (not to my fulfillment). There are several possible design and just manage all that with key or mouse seems to be impossible. But its actually easy just enveloping the already know pattern. with ctrl+alt (yeah modern win managers take that but not in combination with pg keys except under ubuntu unity maybe). If i hold that two keys i just add or remove docbars. And if I add Shift i shuffle not the order of docbars but reshuffle the dsign in which they are tiled. This way I'm never a view keys away from what i exactly want and if i miss it i can easily and intuitively go back. It will be even easy with the mouse since you can pick the designs directly in the menu. And a limit of 4 parallel docbars wouldn't hurt anybody.
That was just one small example but I hope you got the idea. And having 20 or 30 more such ideas can make Kephra an editor not seen before and something exciting new done in Perl. Maybe i should have waited to talk about until its full operational but my main motivation is spark imagination. There's lot to be invented in editing. With a modern code base in high level lang like Perl we can move very fast, even the other editors that might copy nice things like in past years some nice little features Kephra had almost since the beginning started to spread, although I believe most people saw it on gedit.
The rewrite called XP (just expressing my feeling of coming late) is done here and has not much to do with the Kephra on CPAN. Please also check farabi and other interesting new stuff.
]]>Yesterday listened to Schwerns talk on youtube and it triggered some memories of things I wanted talk about over and over again but mostly kept quit. I mean my talk about Perl articles in Wikipedia some years ago that was largely motivated by the dysfunctional community there and how prevent this in Perl realm - but the actual talk was mainly about Perl, presenting Perl, writing wiki texts and how to get along inside Wikipedia.
Seriously. nothing made me so angry in decades as how I was treated in wp. But after years of reflection i see the cause in inabilities on both sides. What happened? Many things. But the icing was an incident where I wanted to establish a list of major figures in the field of esotericism. It was trolled even by people I considered friends with entries like Easter Bunny and Santa Claus. On perlmonks I had something experienced something little bit in that direction when we discussed the case about the UFO-hacker who used a little Perl script. I just offered the viewpoint of the recent UFO literature and community since I supposed most people here are not accustomed with that. Just to make clear with what information background this guy might acted. I was massively downvoted. I would not bitch about that but i felt disconnected. But honestly there are a lot of things make feel disconnected to Perl and programming people in general which brings the very loaded topic of faith.
Larry is a christian and never was silent about that, good. In the eyes of some very loud radical atheist I'm a moron - fine that's consistent with their believe system. I don't even feel an urge to challenge that because all the good arguments are out there and they would listen if they choose to. And I know that mental growth is the order of business. The only reason I raised that topic (beside sometimes to exchange about real interesting stuff) is to lift stress: "here look I'm perfectly normal, even if i believe in pink elephants". I think that's what really Schwerns talk was about and to do a simple ++ wouldn't cut it. We have to see the human first, opinions second.
Political correctness makes me wanna puke, I'm especially not fond of gender mainstreaming and find its blind theorization like most other contemporary philosophies. But i love to be nice to people. That's is the line I also to drew when we had that case tinita already wrote here about it. Which touched me double time - secondly me as a moderator in the largest Perl forum. I really don't like to make a heavy problem out of it because life supposed to be fun. But it takes a huge amount of concentration to be fair in cases someone feels insulted because most of the time both sides are right and wrong at the same time. We constantly start new wars if we declare someone right and someone else wrong (that's the sane part of postmodern relativism). And such an incident is never the root cause of feeling hurt - consult some serious literature about that. If you like victimisation, blaming and such - politics is your arena. In the end all of us want mostly the same and its not even that different. and the only value all this discussion can have is when we just remember how diverse we're truly are and that beauty lies in that. I mean There is more than one point of view or something like that. I don't believe much in rules regulations and things that might to lead to less freedom. lao tzu already wrote there are signs of decay. Nevertheless trolls need, like any other infants, strict boundaries set by their parents. To show up in a relaxed way, help if possible the angry person to refine their statement but make clear that there is a better way to putting it and that its no accepted norm here is like community gardening.
But I believe the most potent thing is to release fear of the different "Other" and to find unexpected friends.
]]>Highlight was to me Stevan Littles presentation (no keynote was utterly important), even if I extracted the crucial infos before. I was especially sceptical to declare one line to get the getter/setter like in Moo/s/e. But combined getter/setter are not always right, you might follow the damian or even (god forbid) your own taste. So to declare that and not deliver that part by default seems to be very perlish and sane. And I got the impression that there is not much uprise against it and that and that it might even arrive in 5.18.
Oh and there were so many familiar faces can't name them all because most i don't know by name. But I like to see them anyway. And I'm glad to know now how O'Rear and fglock are looking. But doing "buisness" is also important. Sync ideas with TPF people and other leads. Ask dancer guys for help for the upcoming parts of my perl tut which should also cover some web development (mojo will be covered as well). I spoke with my next grant manager about tablets and with my nominal part time boss renee about future $foo articles. Even spoke with gabor because i plan to make recension about his stuff in mid 2013. I spoke with a guy met in Riga before about translating my perltut from freiesmagazin into Czech. I had also conversations about the state of the perldoc translation into German, which is in a rather poor and outdated state and not as shiny as Spanish. The most fun I had speaking not about Perl and I'm not sorry about that. And btw. sleeping sacks are wonderful.
The most fun talk was not mst's but mäsaks "joy of breaking stuff". Not that I learned that much but his combination of open and childlike fun and an scholarly attitude is something what I want to incorporate more too. State of the parrot threads was to me very interesting and useful for the upcming perl 6 update in winter 2013. I also sat in the last rows with Larry to listen about 3 state logic and flip-flap-flop hardware.
And of course there were my own talks. I had a lightning about the progress toward my GUI DSL called GCL. I planned to speak about examples how certain data structures get translated into GUI hierarchy but because it was a more or less free speech I brought more attention to the concepts. That was on day 1 were I also heard that there is now a free slot. Which I took of course. After recognizing that my Perl 5 quiz was nor online not on my laptop I squeezed my 40 min Perl 6 Regex talk into 20 min. I had to write most slides again but the result was not that bad. Larry helped as backup in the first row anyway. Afterwards my understanding about regex quantifier modifier was much sharper. i edited the slides life during talk and polished them afterwards too. their online now on my new site.
After it there was mst's key note. Like in the first half of the initial keynote from Mr. wall it was more of the social aspects. It left the good feeling that were on a good path but much is left to do. After that max maischein dropped the statistics of the conference and we could see all the busy and great helpers which got their well deserved thundering applause. And suddenly all was over. Look like I will ride next year to Kiew.
I will speak again about it, but since there will be no other track beside it, I can give now away some of the wow points. What I created just the last days (last commit) is a sizer class that has several advantages.
1. greater whipuptitude - less syntax the create something meaningful
2. more introspection
3. new function that let you insert before after a certain widget
accompanying by a panel class that will be mostly what you really want.
it improves on points 1-3 too plus
4. easy handling of recursive structures
5. no hassle with reparenting of widgets
since the input that these both classes take are Perl data structures, (values, arrays , hashes and ref of all of them) we can serialize it. It slowly dawns on me how even to marshal from existing GUI no matter which way you created it. Because it is a Perly data structure it can be stored in YAML (or JSON if you prefer) not XML like XRC or FBP which both can't do marshalling. Maybe this is a dawn of an universal language other can write ports like GCL::GTK or GCL::Prima that can translate the DSL or data structure into another toolkit and even HTML could be doable and we would have true portability for the first time. Its pretty much wet dreams now. All I have now are the Util, Panel, Sizer, menu and Toolbar builder and yes a Splitter Class. The Splitter is also superior to what you have in Wx since,
1. remembers its children , orientation and position. you just say simply resplit or toggle and all is well.
2 it also handles the reparenting
If anyone helps me to put t on CPAN I would be glad.
Don't worry in the upcoming book, which I started to write this week I also will mention and describe other helper modules from CPAN. But I hope you can understand that I see it as one projects with many arms(Kephra, GCL, Book, Docular and the little Apps that will serve as examples), since much of my experience comes from Kephra. Or maybe I just will collapse under the weight of this task. But hey, nearly half of the current Perl mag and the Perl tutorial part 8 and other things that appear just now also could not slow down it.
Tuts are hard work, I know because i wrote some. To set up all the traps that a screwball comedy comes out of it that actually works dramaturgically should be even harder. But maybe someone runs with the idea.
]]>I still work on the appendix (F and G now) rather on the actual tablets for several reasons:
A) work for the beginner
The appedices are more structured toward questions starters might have and maybe the answers there are also easier to find. This helps to bring more people to Perl 6 and for more sophisticated musings are still the specs there, which are much easier to read, once you have every term there explained in the glossary of the tablets.
B) have to learn myself
Even after all these years, I'm still not fond of all corners and depths of Perl 6 (which is still constantly changing - mostly gaining depth and cutting the not practically working ideas). So I write the broader parts first. Not only to let the tablets shine by having at least some parts in decent completeness, but also to be able to contribute while learning myself. once the appendices are written its much easier build upon that the tablets. yes it sounds like having it backwards but to me it feels better to knowing what I actually linking to and knowing what repetetive explanation I not have to do. Also creating a structure where it is clear what belongs where makes it much easier to contribute, which is far more important than having a few good chapters since this task is simply beyond of the scope of one person. I have a life you know. :)
]]>I added a FAQ (Appendix F) page (and moved links to appendix H aka href appendix), but its just questions so far. You migth ask, but there is already a semiofficial FAQ. True. But would you not prefer better sorted anwers with links where all the parts of the answer are even better explained in more detail?
I added also a report.pl which gives you output like:
page title item anchor links lines bytes
---------------------------------------------------------
Appendix-A 30 756 909 2253 4103 184498
Appendix-B 46 0 0 770 850 83065
Appendix-C 0 0 0 0 4 131
Appendix-D 19 0 0 9 136 6127
Appendix-E 8 0 0 4 102 5154
Appendix-F 5 0 0 0 23 639
Appendix-G 24 120 120 302 582 25523
Appendix-H 8 0 0 42 83 3835
Tablet-0 3 0 0 0 73 3274
Tablet-1 16 0 0 5 141 7461
Tablet-10 6 0 0 0 20 287
Tablet-2 33 0 0 31 333 13365
Tablet-3 23 0 0 46 306 16564
Tablet-4 32 0 0 13 154 4289
Tablet-5 30 0 0 7 158 4365
Tablet-6 19 0 0 0 57 513
Tablet-7 0 0 0 0 3 42
Tablet-8 0 0 0 0 11 533
Tablet-9 10 0 0 5 63 1825
---------------------------------------------------------
Sum 312 876 1029 3487 7202 361490
TODO:
empty index entries: 5
unanswered questions: 8
empty glossary terms: 25
Yes almost 3500 links, mostly hand crafted and just recently reformated. Simply not possible without some decent potion of Kephra substitution-fu.
The rest are details . A lot of them, but mostly boring. And hej I started to work on tablet 2.
Further plans: complete Glossary and then FAQ. :)
]]>This page was created just 2 weeks ago to separate all the glossary terms out of Appendix A. This way A is more pure and usable for autogenerated docs, we have a clearer division of labor and G can have slightly different concept and formating, more suitable for a bit longer explanations which take out some repetitive, shorter descriptions out of A which should serve more as a reminder (like johans pocket ref printed by o'reily). But G is still an integrate part of the tablets, linking to related builtins in A, categories in B, explanations in D or into the tablets 0-10. This makes it even more usable by not only explaining the words in the tablets and Perl 6 spec you might not understand, but also linking to syntax where this concept is put into practice.
The first part is exactly where I wish to see more help, because from the initial 30 something entries it has risen to 100 now, no end in sight. Many are just emty entries that can be filled out by people with some better understanding of the programming concepts several languages use. Adding the Perl 6 bits to it will be much easier for me than. So please clone this repo and send github pull requests or come to #perl6 and get your commit bit.
Thank You very much.
]]>As I prepared a piece for the Perlzeitung which will not happen, i interviewed several people - some names you know for sure - about the current state of the core documents. Yes there slowly improving, the d did IMO good stuff, the new ooptut is good but there needs to be much more done. And instead of dropping here a blob of text, just have a look (Appendix A is really impressive). Its hardly started but I'm sure you get the idea (short but comprehencive explanation of everything + 7 different appendices giving alternative entry point to find what you need). Please just think about. The dream needs the hands of many.
Because the really exciting thing is... the tablets moved from socialtext wiki. They moved to their own URL (link above) they moved to a different format (markdown). The HTML you see on the site is generated by pandoc, which proofs python can be useful, because we can create a PDF or other ebook formats with a snap of a finger. And I can write the texts in Kephra which is by far superior to any HTML-widget.
Last night I converted several hunderts links and anchors, which lets me announce proudly that the conversion to that new format is done. This of course was only possible due to the idea, help and initial generator script done by (John M. Gabriele)++ and some admin work by (Moritz Lenz)++.
BTW they both also conspired to create a new Perl 6 wiki. Gitit is powered by (you guess it) and understands almost any modern markup language - no more pain by socialext!