[dojo-contributors] Tentative agenda for today meeting

Adam L. Peller adam at peller.org
Wed Oct 3 12:02:01 EDT 2012


What I'm reading in the deviceTheme.js comments is that this feature may
not work a all and that while the script tag is better, just avoiding this
feature and loading stylesheets directly is the only guarantee.

Given that we've already slipped into this usage pattern prior to
publication of all of these warnings, and we have an implementation that
basically seems to work, we badly need Richard's workaround until 2.0.
 We've scoped out what it would take to restructure the documents with the
script tag in Maqetta.  It would represent a significant amount of work,
not just to special case the model we use to construct our documents, but
to migrate existing usage.  For those worried about reinforcing an
undesirable usage pattern, this feature is already accompanied by warnings,
adding the simple module id is not going to change that.

-Adam


On Wed, Oct 3, 2012 at 10:59 AM, Richard Backhouse <backhous at us.ibm.com>wrote:

> If that's the case then what is the harm with adding a module id to the
> define call :
>
> (typeof define === "undefined" ? function(id, deps, def) { def(); } :
> define)("dojox/mobile/deviceTheme", [
>
>  "dojo/_base/config",
>  "dojo/_base/lang",
>  "dojo/_base/window",
>  "require"
> ], function(config, lang, win, require){
>
> I know it's not recommended to hard code ids but I think this can be
> considered a special case, especially as loading via AMD is discouraged.
> This is the patch we use to get it loading in Maqetta with Zazl
>
> Richard
>
> [image: Inactive hide details for Christophe Jolif ---10/03/2012 10:53:42
> AM---Richard,]Christophe Jolif ---10/03/2012 10:53:42 AM---Richard,
>
>
>
>    From:
>
>
> Christophe Jolif <cjolif at gmail.com>
>
>    To:
>
>
> "dojo dev." <dojo-contributors at mail.dojotoolkit.org>,
>
>    Date:
>
>
> 10/03/2012 10:53 AM
>
>    Subject:
>
>
> Re: [dojo-contributors] Tentative agenda for today meeting
> ------------------------------
>
>
>
> Richard,
>
> That code is just here is to _not_ totally break existing apps that might
> be doing that prior to our recommendation of _not_ doing it. I don't feel
> this would be reasonable to remove that code and totally breaking existing
> in a patch release.
>
> --
> Christophe
>
> On Wed, Oct 3, 2012 at 4:48 PM, Richard Backhouse <*backhous at us.ibm.com*<backhous at us.ibm.com>>
> wrote:
>
>    If it won't load via an AMD loader reliably then I suggest taking out
>    the define call completely and make it only load via a script tag.
>
>    BTW the Zazl issue has nothing to do with plugins. It's when you have
>    code like what is in deviceTheme problems arise with dependency analysis
>
>    (*typeof* define === "undefined" ? *function*(deps, def) { def(); } :
>    define)([
>    "dojo/_base/config",
>    "dojo/_base/lang",
>    "dojo/_base/window",
>    "require"
>    ], *function*(config, lang, win, require){
>
>
>
>    [image: Inactive hide details for Ben Hockey ---10/03/2012 10:39:12
>    AM---On Oct 3, 2012, at 9:26 AM, Christophe Jolif wrote:]Ben Hockey
>    ---10/03/2012 10:39:12 AM---On Oct 3, 2012, at 9:26 AM, Christophe Jolif
>    wrote:
>
>
>
>    From:
>
>
> Ben Hockey <*neonstalwart at gmail.com* <neonstalwart at gmail.com>>
>
>    To:
>
>
> "dojo dev." <*dojo-contributors at mail.dojotoolkit.org*<dojo-contributors at mail.dojotoolkit.org>>,
>
>
>    Date:
>
>
> 10/03/2012 10:39 AM
>
>    Subject:
>
>
> Re: [dojo-contributors] Tentative agenda for today meeting
>
>    ------------------------------
>
>
>
>
>
>    On Oct 3, 2012, at 9:26 AM, Christophe Jolif wrote:
>
>    > Adam,
>    >
>    > On Wed, Oct 3, 2012 at 4:05 PM, Adam L. Peller <*adam at peller.org*<adam at peller.org>>
>    wrote:
>    >
>    >> Could your team take another look at
>    >> *http://trac.dojotoolkit.org/ticket/15901*<http://trac.dojotoolkit.org/ticket/15901> whether
>    for 1.8.1 or 1.8.2?
>    >> It's a blocker for Maqetta, and we had to fork the repository to get
>    >> around it.
>    >
>    > Maybe Eric will have a different stance on this. But to me from our
>    > past experience this is really not a good idea to load that piece of
>    > JavaScript as an (asynchronous) AMD module (that might change in the
>    > future if we are able to get notified on CSS loading but as of today
>    > this might well end up in initialization nightmare, especially when
>    > running in a container like Cordova) so I'm wondering why Maqetta
>    > seems to be reluctant to use it as recommended which, if I'm not
>    > mistaken, would solve the issue without having to change Dojo?
>    >
>
>    i agree with christophe, i've had exactly the same experience.  this
>    file does not work when loaded asynchronously via a loader.
>
>    if you have never seen this issue, the things i found that exacerbated
>    the problem was to do a build and try to load an alternative theme to
>    iPhone (which is the default so iirc sometimes it seemed like it was
>    working) - i was using android.
>
>    the only way this file works consistently is to add it as a separate
>    script tag.  although this is inconvenient, i don't think changing it to
>    suit static analysis is going to help - if anything it just makes it easier
>    to use it in a way that doesn't work.
>
>    as christophe eluded to, this file can't be loaded reliably by an
>    asynchronous loader without knowing when the css is fully loaded - perhaps
>    by using some kind of css plugin.  i believe zazl has difficulties with
>    plugins (maybe i don't understand and it might be able to handle this case)
>    so this doesn't seem like it would move you forward either.
>
>    ben...
>
>    _______________________________________________
>    dojo-contributors mailing list*
>    **dojo-contributors at mail.dojotoolkit.org*<dojo-contributors at mail.dojotoolkit.org>
>    *
>    **http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors*<http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors>
>
>
>
>
>    _______________________________________________
>    dojo-contributors mailing list*
>    **dojo-contributors at mail.dojotoolkit.org*<dojo-contributors at mail.dojotoolkit.org>
>    *
>    **http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors*<http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors>
>
>
>
>
> --
> Christophe_______________________________________________
> dojo-contributors mailing list
> dojo-contributors at mail.dojotoolkit.org
> http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
>
>
>
> _______________________________________________
> dojo-contributors mailing list
> dojo-contributors at mail.dojotoolkit.org
> http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.dojotoolkit.org/pipermail/dojo-contributors/attachments/20121003/c858d089/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 45 bytes
Desc: not available
Url : http://mail.dojotoolkit.org/pipermail/dojo-contributors/attachments/20121003/c858d089/attachment.gif 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 105 bytes
Desc: not available
Url : http://mail.dojotoolkit.org/pipermail/dojo-contributors/attachments/20121003/c858d089/attachment-0001.gif 


More information about the dojo-contributors mailing list