greetings DS,
Saturday, June 02, 2001, 2:23:58 AM, you wrote:
DS> I think most of the documentation for API2 does more to hide "how the code
DS> works" in the API itself. This is wonderful for people just planning on
DS> using it at it's most primitive level and never extending it.
well sometimes you just need some BASIC (but still - versatile,
cross-browser, reusable etc.) functionality for your not-so-simple
code.
imagine debugging a not-so-simple php4-intense html file which is
about 5-6 times bigger than its parsed version, full of flash, html
tables and styles - it's VERY hard to do ANYTHING complicated in JS
just cause you'll get heaps of errors.
when i have my BASIC functionality i code all the other stuff by hand.
i just need to make sure that my ACTUAL "i/o" functions won't fail in
netscape6 or somewhere else.
DS> This is how you make a DynLayer... (no discussion of DIV's etc..).
DS> This is how you move a DynLayer...
DS> This is how you talk to a DynLayer... (no discussion of the value of
DS> event-bubbling).
:)
DS> Pascal delivered the closest "how the code works" version with his DynaCore
DS> Developers Documentation. But alas even that tends to glaze over the real
DS> inner working of the API and is really intended for a solid coding developer
DS> that has spent a great deal of time studying the code of the API itself.
yes, that's the point.
the same thing i discovered with php - it's a bit offtopic, but, for
example, i found lots of "here you go with arrays" - and after some years of
c/c++ programming those "arrays" seem a great mystery to me. the same
with the rest of the docs - "here you do so and so - and - poof, magic
happens", but HOW it works - that's a real mystery.
i feel a great need to have only half (or even 1/3) of functionality, but with heaps
of useful docs, rather than full functionality and no docs at all.
DS> Dan's original API1 documentation was like working with an excellent
DS> teacher. He defined a problem, discussed both Netscape's and Microsoft's
DS> approach then demonstrated a cross-browser solution using his API.
yes, EXACTLY !
DS> You learned about both browser versions of CSS-P and an API all in one pass.
DS> Wondrous!
well i was used to css1 beforehand and coded html for years,
nevertheless, Dan's docs, both DynDuo and DynAPI, were a TREMENDOUS
help.
DS> With the API2 we act more like a Dreamweaver users manual. Go here. Pick
DS> this. Set that. Magic happens. In the end the person can make a web page
DS> with functionality but has no clue what-so-ever as to the inner workings of
DS> it.
DS> I don't know, maybe this is how it is suppose to be. But I think we miss
DS> the real value this API has if we follow this course. Open Source is so
DS> grassroots and your participation in "code problem"/"code discussion"/"code
DS> solution" thru threaded venues like this do more to teacher you then all the
DS> bound tech books combined.
well maybe. it's okay when i'm asking about something tricky or
complicated, but situation is rather ridiculous cause 1) i'm asking
for SIMPLE solution, instead of complicated (and, to be serious,
rather useless examples) stuff and 2) well, i used DynAPI 1 before so
i'm hoping to have a bit more understanding of the process then a
regular novice :)
DS> I think we could make it better and document the basics of the API better so
DS> that collaborative extending of it happens faster and over a larger group of
DS> contributors. The Pascal's, Roberts and Jordi's all "got there" the hard
DS> way (there are many individuals involved here who also deserve that "got
DS> there" distinction).
well probably. that's what i'm up to - to get the basic stuff done and
to document in out - either as an example with LOTS of comments (if
i'll be short on time) or as a little tutorial.
DS> The hard way requires eat sleep and drinking not only "this" API but
DS> everything you can find on event models, the DOM's, OOP JavaScript, OOP
DS> principals (encapsulation, abstraction), etc... This is not a short path
DS> to take but "if" you take it. You will end up in a very special place.
well understanding the WHOLE process (spending my nighttime with
XML/XSLT - very funny, indeed :) is a great mastery and all, but then
i can come across another question - "what for ?". tools aren't really
critical to the entire process. the problem with (particularly) js is
that js is the only possible solution towards basic and cross-browser
interactivity on webpages without using "closed" technologies (e.g.
flash).
DS> A place few reside.
oh well... :)
peace,
x.
http://xmypoe.com
_______________________________________________
Dynapi-Help mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/dynapi-help
