dynapi-dev

[Dynapi-Dev] CVS Partially Updated with Makefile

[Dan S. <dy...@fu...>]

I've contributed my Makefile into the CVS, along with updating the /api/ and /util/ files with semi colons (that's enough ;'s for one night).

The JSPack util that is needed to build the compressed DynAPI is released separately:

http://www.dansteinman.com/jspack/

Also, I SERIOUSLY recommend removing the /js/ directory altogether.  We don't want people submitting changes to the files in /js/ because that is the directory that is automatically created by the makefile.

The examples need to be updated to point to /src/.  Also it would be good to ensure that all the examples use the DynAPI include() so that the switch from /src/ to /js/ is a one line change.

The DynAPI packages that are created by the makefile/jspack will not work correctly yet.  If there is even one line that does not have a semi colon in the right spot nothing will work.

The Makefile produces a duplicate copy of the /src/ tree with all the .js files compressed.  It also creates .js packages of all the files in the subdirectories(eg /lib/dynapi.api.js) as well a "all" package which contains everything.  It goes one step further to Jar and Gzip those package files.

As I mentioned earlier the significant semi colon to look for is in this case:

method = function() {

};  // it occurs in event listeners and every prototype in the DynAPI

Also it's better now to never do one line methods like:

obj.prototype.method=function() {code}

Just because it's not easy to spot where the ;'s go:

obj.prototype.method=function() {code;}; // better to unwrap this

The files that need to be updated with semi colons are in /gui/ and /ext/.  At that point we should be able to build and a working compressed DynAPI.  It's really quite a dramatic compression.  From 100Kb of code down to 15Kb Jar and Gzip = 85%.  Pretty darned cool if I say so.

Robert will you be able to use jspack on the sourceforge server for cron jobs?

Let me know of any problems,
Dan

[Dynapi-Dev] Bug with SlideTo in a Viewport

[Raymond S. <dst...@or...>]

Just started doing some research into whats causing it.  Happens in both NS
and IE.  To recreate it make a simple 3X3 array as a gif and set 9 targets,
we use printer registration marks.  Triggering slide to results in "sometime
right, sometimes wrong" animation landings.  This can happen with any of
them.  Being right, then wrong on the next move.  Errors are horizontal or
vertical in the 4-5 pixel range.

Keep you updated.

Worked fine in prior builds

DS

Re: [Dynapi-Dev] Bug with SlideTo in a Viewport

[Raymond S. <dst...@or...>]

The math.round in pathanim.js is creating the error.  If you need precision
moves, remove the rounding.

DS
----- Original Message -----
From: "Raymond Smith" <dst...@or...>
To: <dyn...@li...>
Sent: Tuesday, December 05, 2000 11:27 PM
Subject: [Dynapi-Dev] Bug with SlideTo in a Viewport


> Just started doing some research into whats causing it.  Happens in both
NS
> and IE.  To recreate it make a simple 3X3 array as a gif and set 9
targets,
> we use printer registration marks.  Triggering slide to results in
"sometime
> right, sometimes wrong" animation landings.  This can happen with any of
> them.  Being right, then wrong on the next move.  Errors are horizontal or
> vertical in the 4-5 pixel range.
>
> Keep you updated.
>
> Worked fine in prior builds
>
> DS
>
> _______________________________________________
> Dynapi-Dev mailing list
> Dyn...@li...
> http://lists.sourceforge.net/mailman/listinfo/dynapi-dev
>

Re: [Dynapi-Dev] Bug with SlideTo in a Viewport

[Richard :o <ma...@ri...>]

hi,
If you are using a recent beta release, but not yesterday's, it might still
contain the error I reported in
pathanim.js on line 133, y2==y2 should be y2==y1.
  var dy = y2==y1? 0 : (y2-y1)/N;
As you probably know slideTo is being generated on the fly and handled by
pathanim.js now (AFAIK)

Cheers,
Richard:o

----- Original Message -----
From: "Raymond Smith" <dst...@or...>
To: <dyn...@li...>
Sent: Wednesday, December 06, 2000 8:27 AM
Subject: [Dynapi-Dev] Bug with SlideTo in a Viewport


> Just started doing some research into whats causing it.  Happens in both
NS
> and IE.  To recreate it make a simple 3X3 array as a gif and set 9
targets,
> we use printer registration marks.  Triggering slide to results in
"sometime
> right, sometimes wrong" animation landings.  This can happen with any of
> them.  Being right, then wrong on the next move.  Errors are horizontal or
> vertical in the 4-5 pixel range.
>
> Keep you updated.
>
> Worked fine in prior builds
>
> DS
>
> _______________________________________________
> Dynapi-Dev mailing list
> Dyn...@li...
> http://lists.sourceforge.net/mailman/listinfo/dynapi-dev
> ____________________________________________________________
> Get your FREE personal .com domain name and
> NAMEzero Personal Portal at: http://www.namezero.com.
> For customer service, mailto:cus...@na....
>
>

Re: [Dynapi-Dev] CVS Partially Updated with Makefile

[Robert R. <rra...@ya...>]

I added semicolons to the rest of the files.  However, there could be
mistakes.  Please get the latest snapshot at
http://dynapi.sourceforge.net/snapshot/dynapi.zip, and post to the
list if you find errors in semicolon placement.

-- 
// Robert Rainwater

On 12/6/2000, 4:57:54 AM EST, Dan wrote about "[Dynapi-Dev] CVS Partially Updated with Makefile":

> I've contributed my Makefile into the CVS, along with updating the /api/ and /util/ files with semi colons (that's enough ;'s for one night).

> The JSPack util that is needed to build the compressed DynAPI is released separately:

> http://www.dansteinman.com/jspack/

> Also, I SERIOUSLY recommend removing the /js/ directory altogether.  We don't want people submitting changes to the files in /js/ because that is the directory that is automatically created by the
> makefile.

> The examples need to be updated to point to /src/.  Also it would be good to ensure that all the examples use the DynAPI include() so that the switch from /src/ to /js/ is a one line change.

> The DynAPI packages that are created by the makefile/jspack will not work correctly yet.  If there is even one line that does not have a semi colon in the right spot nothing will work.

> The Makefile produces a duplicate copy of the /src/ tree with all the .js files compressed.  It also creates .js packages of all the files in the subdirectories(eg /lib/dynapi.api.js) as well a
> "all" package which contains everything.  It goes one step further to Jar and Gzip those package files.

> As I mentioned earlier the significant semi colon to look for is in this case:

> method = function() {

> };  // it occurs in event listeners and every prototype in the DynAPI

> Also it's better now to never do one line methods like:

> obj.prototype.method=function() {code}

> Just because it's not easy to spot where the ;'s go:

> obj.prototype.method=function() {code;}; // better to unwrap this

> The files that need to be updated with semi colons are in /gui/ and /ext/.  At that point we should be able to build and a working compressed DynAPI.  It's really quite a dramatic compression. 
> From 100Kb of code down to 15Kb Jar and Gzip = 85%.  Pretty darned cool if I say so.

> Robert will you be able to use jspack on the sourceforge server for cron jobs?

> Let me know of any problems,
> Dan
> _______________________________________________
> Dynapi-Dev mailing list
> Dyn...@li...
> http://lists.sourceforge.net/mailman/listinfo/dynapi-dev