Hi all,

I've been (slowly) working on a new API doc tool, as a part of the new
DTK site redesign effort; an alpha is now up at
http://api-staging.dojotoolkit.org/ . If you have a few minutes,
please take a look and shoot me feedback here.

Known issues:
1. The version switching does not change the tree on the left side.
2. Privates are not being filtered.
3. The scrolling animation is not getting to the correct point in any tab panes.
4. A search function is not implemented yet.
5. Only the 1.4 docs are fully functional (have not regenerated docs
for 1.3 or for HEAD).

Features:
1. Supports full deep-linking; bookmarking the "Permalink" at the top
of any doc page will kick you directly to that doc later on.
2. A caching mechanism is in place that stores all ajaxified portions
of the app; as the site is used, HTML files are being saved that
populate tab panes.
3. Fully degradable (i.e. if no JS, the site should still function).
4. Somewhat themeable; basically the header and footer can be modified
without much hassle.
5. Fully printable; hitting control/command-P will print the currently
focused tab pane without extra things.
6. Deep-linking supports the current API tool URL format (i.e.
http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
just fine).
7. Runs/is based on the XML serialization generated by Neil's docscript parser.

The site is running off the Google CDN for 1.3.2; I briefly switched
it to 1.4 but there were issues with the scrolling tab bar and I
didn't have the time to figure out why.

Things I'm interested in:
1. Help =)
2. Any bugs you might find that I didn't list above.
3. How useful you find it; is it a bit easier to use than the current API tool?

The goal of writing this was to be able to put together a tool that
anyone with the right XML format can use for their own documentation;
I'm successfully using this for a current SitePen client at the
moment, and it seems like it works pretty well.

Dependencies:
A PHP/Apache stack
mod_rewrite under Apache
XSL and cURL enabled in PHP (only for the uploading capability, the
rest of the site should work with a bare-bones PHP install).

Let me know what you think, and report any bugs on this thread (for now).

-- Tom

Search Discussions

  • Tom Trenka at Dec 15, 2009 at 5:17 pm
    Oh, I forgot: some of the examples formatting is off (that's kind of
    ugly stuff right there), and I'm not currently capturing/showing info
    about constructors that have been documented correctly.

    Note also that some issues you might run across may be due to the XML
    serialization format and not a UI bug.

    Thanks--
    Tom
    On Tue, Dec 15, 2009 at 4:15 PM, Tom Trenka wrote:
    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
  • Shane O'Sullivan at Dec 15, 2009 at 6:55 pm
    Hi Tom,

    Very nice, I'm surprised at how fast it runs. Is this intended to be
    used as well as the searchable API tool that we have now, or is it
    intended to replace it? Does it borrow anything from the very cool
    doc tool the Uxebu guys did at dojodocs.uxebu.com?

    Thanks

    Shane

    2009/12/15 Tom Trenka <ttrenka at gmail.com>:
    Oh, I forgot: some of the examples formatting is off (that's kind of
    ugly stuff right there), and I'm not currently capturing/showing info
    about constructors that have been documented correctly.

    Note also that some issues you might run across may be due to the XML
    serialization format and not a UI bug.

    Thanks--
    Tom
    On Tue, Dec 15, 2009 at 4:15 PM, Tom Trenka wrote:
    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
  • Tom Trenka at Dec 15, 2009 at 7:59 pm
    Right now the intention is to replace the API tool we have now, and
    no, it doesn't borrow anything from uxebu; it's actually a pretty
    bare-bones PHP-based site that uses minimal Dijit to set up the UI.
    Not knocking on anything Wolfram did, but the goal here was more to
    get something that anyone using the Dojo doc parser could use for
    their own documentation, which this does (it's based on the XML
    serialization that the doc parser generates).

    In theory, this *should* lend itself to indexing by Google, etc.; it's
    actually page-based, but the minimal JS I wrote intercepts any link
    clicks to set up the tab pane stuff. I have not tried it yet for that
    purpose, but the way it's set up should make it pretty easy to do.

    You can take a look at the source at
    https://svn.dojotoolkit.org/website/trunk/api . The entry point is
    index.php, and the relevant files are lib/item.php and
    lib/class-tree.php. There's still a bit of work to do (I kind of
    hacked it together fairly quickly) in terms of cleanup, like with the
    config stuff but I think the majority of it is there.

    -- Tom

    On Tue, Dec 15, 2009 at 5:55 PM, Shane O'Sullivan
    wrote:
    Hi Tom,

    Very nice, I'm surprised at how fast it runs. ?Is this intended to be
    used as well as the searchable API tool that we have now, or is it
    intended to replace it? ?Does it borrow anything from the very cool
    doc tool the Uxebu guys did at dojodocs.uxebu.com?

    Thanks

    Shane

    2009/12/15 Tom Trenka <ttrenka at gmail.com>:
    Oh, I forgot: some of the examples formatting is off (that's kind of
    ugly stuff right there), and I'm not currently capturing/showing info
    about constructors that have been documented correctly.

    Note also that some issues you might run across may be due to the XML
    serialization format and not a UI bug.

    Thanks--
    Tom
    On Tue, Dec 15, 2009 at 4:15 PM, Tom Trenka wrote:
    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
  • Eugene Lazutkin at Dec 16, 2009 at 12:08 am
    Looks very nice and works fast!

    I tried to find dojo.attr and somehow expected it to be in the tree, it
    was not (obviously). I remembered that it is not a namespace, so I
    clicked on dojo and started to look alphabetically and ... didn't find
    it. After a minute or so I realized that there are several separately
    sorted lists, so I looked at "methods" and found it. All in all I am a
    little bit bummed that a Dojo Base function, which is always included
    and frequently needed, is hidden so deep. On the other hand, if we add
    live filtering or live search, it would be ok because there is an easy
    way to find it. Any thoughts on that?

    Looking at docs I noticed some formatting problems and I suspect that
    the reason is a formatting mistake in comments. It would be really nice
    to be able to do quick round-trips: change the file, run some tool and
    see the output the same way it is shown in the API tool. How hard is it
    to implement something like that?

    Eugene Lazutkin
    Dojo Toolkit, Committer
    http://lazutkin.com/
    On 12/15/2009 04:15 PM, Tom Trenka wrote:
    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
  • Tom Trenka at Dec 16, 2009 at 12:54 am
    Inline...

    On Tue, Dec 15, 2009 at 11:08 PM, Eugene Lazutkin
    wrote:
    Looks very nice and works fast!

    I tried to find dojo.attr and somehow expected it to be in the tree, it
    was not (obviously). I remembered that it is not a namespace, so I
    clicked on dojo and started to look alphabetically and ... didn't find
    it. After a minute or so I realized that there are several separately
    sorted lists, so I looked at "methods" and found it. All in all I am a
    little bit bummed that a Dojo Base function, which is always included
    and frequently needed, is hidden so deep. On the other hand, if we add
    live filtering or live search, it would be ok because there is an easy
    way to find it. Any thoughts on that?
    I'm planning on adding a search to this, just haven't gotten to it yet.
    Looking at docs I noticed some formatting problems and I suspect that
    the reason is a formatting mistake in comments. It would be really nice
    to be able to do quick round-trips: change the file, run some tool and
    see the output the same way it is shown in the API tool. How hard is it
    to implement something like that?
    This is run off the doc parser that Neil wrote, which sits in the
    util/docscripts directory (generate.php). Unfortunately with
    something as large as DTK, it takes a while to run. However...with
    other smaller packages, it's quite fast (I've been using it this way
    on a project) so doing what you're asking for is pretty simple. The
    site includes an upload function to push new XML files up and process
    them, so I think what you're looking for is covered.

    -- Tom
    Eugene Lazutkin
    Dojo Toolkit, Committer
    http://lazutkin.com/
    On 12/15/2009 04:15 PM, Tom Trenka wrote:
    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
  • Mike Wilcox at Dec 16, 2009 at 8:30 am
    Eugene are you aware of the Doc Preview Tool that comes in util? After
    writing your doc comments, hit this page (adjust the localHost of
    course):

    http://test.local/dojotoolkit/util/docscripts/_browse2.php


    Mike Wilcox
    Software Engineer
    SitePen, Inc.
    http://www.sitepen.com
    On Dec 15, 2009, at 11:08 PM, Eugene Lazutkin wrote:

    Looks very nice and works fast!

    I tried to find dojo.attr and somehow expected it to be in the tree,
    it
    was not (obviously). I remembered that it is not a namespace, so I
    clicked on dojo and started to look alphabetically and ... didn't find
    it. After a minute or so I realized that there are several separately
    sorted lists, so I looked at "methods" and found it. All in all I am a
    little bit bummed that a Dojo Base function, which is always included
    and frequently needed, is hidden so deep. On the other hand, if we add
    live filtering or live search, it would be ok because there is an easy
    way to find it. Any thoughts on that?

    Looking at docs I noticed some formatting problems and I suspect that
    the reason is a formatting mistake in comments. It would be really
    nice
    to be able to do quick round-trips: change the file, run some tool and
    see the output the same way it is shown in the API tool. How hard is
    it
    to implement something like that?

    Eugene Lazutkin
    Dojo Toolkit, Committer
    http://lazutkin.com/
    On 12/15/2009 04:15 PM, Tom Trenka wrote:
    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the
    new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in
    any tab panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be
    modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the
    currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should
    work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's
    docscript parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the
    current API tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for
    now).

    -- Tom
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
  • Mike Wilson at Dec 18, 2009 at 4:33 am
    A cool feature (needing a fair amount of work though) would be
    to allow inline editing of docs for logged in users, and posting
    the edits as patch files to trac or similar.

    Best regards
    Mike Wilson

    Eugene Lazutkin wrote:
    It would be really nice to be able to do quick round-trips:
    change the file, run some tool and see the output the same
    way it is shown in the API tool. How hard is it to implement
    something like that?
  • Chris Mitchell at Dec 16, 2009 at 11:08 am
    Nice work!

    Here some feedback based on a brief tour of the site:

    The api web site looks very nice. It runs very fast for me in all browsers.


    Is there a quick way to search/filter the api's by keyword?

    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. See the detailed api
    page for dojo.data.api.Identity to see the problem.

    Also, many private objects appear by default in the tree--most of these have
    no docs at all. Shouldn't they be filtered out in the default view?

    Once you navigate through a number of api's, the tabs become a nuisance. A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.

    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?

    Good stuff!
    -Chris

    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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/20091216/f6dc071d/attachment.htm
  • Tom Trenka at Dec 16, 2009 at 11:24 am
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
  • Shane O'Sullivan at Dec 16, 2009 at 11:31 am
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
  • Tom Trenka at Dec 16, 2009 at 12:28 pm
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
  • Tom Trenka at Dec 16, 2009 at 12:32 pm
    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
  • Micheil Smith at Dec 16, 2009 at 10:32 pm
    You know, the other alternative to using PHP here might be to use Node.js, which should allow us to process the docs on the backend using the similar code to that which the uxebu guys wrote.

    Just an idea :D

    Otherwise, this is looking rather nice.

    - Micheil
    On 17/12/2009, at 4:32 AM, Tom Trenka wrote:

    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors
  • Tom Trenka at Dec 16, 2009 at 11:31 pm
    It might be; since my origins with JS lie on the server and not on the
    client, I can see that. I've tried to make sure that shifting this
    system to something other than PHP would be relatively painless as
    well...either way, what I'm looking to do here is to provide a
    fast-as-possible system for people to be able to effectively use
    documentation provided by the toolkit, in a way that can be used by
    anyone else writing their own systems, using this particular
    documentation format, that can be consumed with as few tools as
    possible.

    Ideally what I'd like to see is ports to your system of choice, be it
    node.js, something Rhino-based, something Python-based, something
    <insert your system of choice here>, and at the same time use this as
    a demonstration of what someone can whip up using Dijit *and* good,
    degradable practices. But first we need to show that it's possible,
    demonstrate how, and actually provide something useful (and at the
    same time do it in a way where others that grok the XML formats that
    Neil put together can duplicate). In other words, create something
    where people looking to find out about how something works is not a
    frustrating experience, and therefore moves onto something else.

    Forgive me if the tone sounds harsh; I have a long history with
    computer-based music where people expect something to be accepted not
    because of the final result, but because of the tools used to make it.
    I'm kind of hoping that we can buck that trend. And thank you for
    saying it looks kind of nice =)

    -- Tom

    P.S. For the record, I did it in PHP mostly because I know it's
    available to pretty much everyone, and (aside from Windows) is pretty
    easy to setup. Most of what I've done with this is really pretty
    basic stuff.

    P.P.S. Christ, this email is downright Faulknarian =)

    On Wed, Dec 16, 2009 at 9:32 PM, Micheil Smith wrote:
    You know, the other alternative to using PHP here might be to use Node.js, which should allow us to process the docs on the backend using the similar code to that which the uxebu guys wrote.

    Just an idea :D

    Otherwise, this is looking rather nice.

    - Micheil
    On 17/12/2009, at 4:32 AM, Tom Trenka wrote:

    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    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
  • Tom Trenka at Dec 17, 2009 at 12:51 pm
    An update: I just got version switching going with the tree, and am
    now displaying version info in the actual documentation.

    -- Tom
    On Wed, Dec 16, 2009 at 10:31 PM, Tom Trenka wrote:
    It might be; since my origins with JS lie on the server and not on the
    client, I can see that. ?I've tried to make sure that shifting this
    system to something other than PHP would be relatively painless as
    well...either way, what I'm looking to do here is to provide a
    fast-as-possible system for people to be able to effectively use
    documentation provided by the toolkit, in a way that can be used by
    anyone else writing their own systems, using this particular
    documentation format, that can be consumed with as few tools as
    possible.

    Ideally what I'd like to see is ports to your system of choice, be it
    node.js, something Rhino-based, something Python-based, something
    <insert your system of choice here>, and at the same time use this as
    a demonstration of what someone can whip up using Dijit *and* good,
    degradable practices. ?But first we need to show that it's possible,
    demonstrate how, and actually provide something useful (and at the
    same time do it in a way where others that grok the XML formats that
    Neil put together can duplicate). ?In other words, create something
    where people looking to find out about how something works is not a
    frustrating experience, and therefore moves onto something else.

    Forgive me if the tone sounds harsh; I have a long history with
    computer-based music where people expect something to be accepted not
    because of the final result, but because of the tools used to make it.
    ?I'm kind of hoping that we can buck that trend. ?And thank you for
    saying it looks kind of nice =)

    -- Tom

    P.S. ?For the record, I did it in PHP mostly because I know it's
    available to pretty much everyone, and (aside from Windows) is pretty
    easy to setup. ?Most of what I've done with this is really pretty
    basic stuff.

    P.P.S. ?Christ, this email is downright Faulknarian =)

    On Wed, Dec 16, 2009 at 9:32 PM, Micheil Smith wrote:
    You know, the other alternative to using PHP here might be to use Node.js, which should allow us to process the docs on the backend using the similar code to that which the uxebu guys wrote.

    Just an idea :D

    Otherwise, this is looking rather nice.

    - Micheil
    On 17/12/2009, at 4:32 AM, Tom Trenka wrote:

    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    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
  • Tom Trenka at Dec 17, 2009 at 3:40 pm
    I started a Wave for this; if anyone wants in, send me your Wave info
    and I'll add you to it.

    -- Tom
    On Thu, Dec 17, 2009 at 11:51 AM, Tom Trenka wrote:
    An update: I just got version switching going with the tree, and am
    now displaying version info in the actual documentation.

    -- Tom
    On Wed, Dec 16, 2009 at 10:31 PM, Tom Trenka wrote:
    It might be; since my origins with JS lie on the server and not on the
    client, I can see that. ?I've tried to make sure that shifting this
    system to something other than PHP would be relatively painless as
    well...either way, what I'm looking to do here is to provide a
    fast-as-possible system for people to be able to effectively use
    documentation provided by the toolkit, in a way that can be used by
    anyone else writing their own systems, using this particular
    documentation format, that can be consumed with as few tools as
    possible.

    Ideally what I'd like to see is ports to your system of choice, be it
    node.js, something Rhino-based, something Python-based, something
    <insert your system of choice here>, and at the same time use this as
    a demonstration of what someone can whip up using Dijit *and* good,
    degradable practices. ?But first we need to show that it's possible,
    demonstrate how, and actually provide something useful (and at the
    same time do it in a way where others that grok the XML formats that
    Neil put together can duplicate). ?In other words, create something
    where people looking to find out about how something works is not a
    frustrating experience, and therefore moves onto something else.

    Forgive me if the tone sounds harsh; I have a long history with
    computer-based music where people expect something to be accepted not
    because of the final result, but because of the tools used to make it.
    ?I'm kind of hoping that we can buck that trend. ?And thank you for
    saying it looks kind of nice =)

    -- Tom

    P.S. ?For the record, I did it in PHP mostly because I know it's
    available to pretty much everyone, and (aside from Windows) is pretty
    easy to setup. ?Most of what I've done with this is really pretty
    basic stuff.

    P.P.S. ?Christ, this email is downright Faulknarian =)

    On Wed, Dec 16, 2009 at 9:32 PM, Micheil Smith wrote:
    You know, the other alternative to using PHP here might be to use Node.js, which should allow us to process the docs on the backend using the similar code to that which the uxebu guys wrote.

    Just an idea :D

    Otherwise, this is looking rather nice.

    - Micheil
    On 17/12/2009, at 4:32 AM, Tom Trenka wrote:

    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    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
  • LiuCougar at Dec 17, 2009 at 3:49 pm
    count me in
    On Thu, Dec 17, 2009 at 12:40, Tom Trenka wrote:
    I started a Wave for this; if anyone wants in, send me your Wave info
    and I'll add you to it.

    -- Tom
    On Thu, Dec 17, 2009 at 11:51 AM, Tom Trenka wrote:
    An update: I just got version switching going with the tree, and am
    now displaying version info in the actual documentation.

    -- Tom
    On Wed, Dec 16, 2009 at 10:31 PM, Tom Trenka wrote:
    It might be; since my origins with JS lie on the server and not on the
    client, I can see that. ?I've tried to make sure that shifting this
    system to something other than PHP would be relatively painless as
    well...either way, what I'm looking to do here is to provide a
    fast-as-possible system for people to be able to effectively use
    documentation provided by the toolkit, in a way that can be used by
    anyone else writing their own systems, using this particular
    documentation format, that can be consumed with as few tools as
    possible.

    Ideally what I'd like to see is ports to your system of choice, be it
    node.js, something Rhino-based, something Python-based, something
    <insert your system of choice here>, and at the same time use this as
    a demonstration of what someone can whip up using Dijit *and* good,
    degradable practices. ?But first we need to show that it's possible,
    demonstrate how, and actually provide something useful (and at the
    same time do it in a way where others that grok the XML formats that
    Neil put together can duplicate). ?In other words, create something
    where people looking to find out about how something works is not a
    frustrating experience, and therefore moves onto something else.

    Forgive me if the tone sounds harsh; I have a long history with
    computer-based music where people expect something to be accepted not
    because of the final result, but because of the tools used to make it.
    ?I'm kind of hoping that we can buck that trend. ?And thank you for
    saying it looks kind of nice =)

    -- Tom

    P.S. ?For the record, I did it in PHP mostly because I know it's
    available to pretty much everyone, and (aside from Windows) is pretty
    easy to setup. ?Most of what I've done with this is really pretty
    basic stuff.

    P.P.S. ?Christ, this email is downright Faulknarian =)

    On Wed, Dec 16, 2009 at 9:32 PM, Micheil Smith wrote:
    You know, the other alternative to using PHP here might be to use Node.js, which should allow us to process the docs on the backend using the similar code to that which the uxebu guys wrote.

    Just an idea :D

    Otherwise, this is looking rather nice.

    - Micheil
    On 17/12/2009, at 4:32 AM, Tom Trenka wrote:

    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. ?It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. ?See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. ?I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    ?Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. ?Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. ?A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. ?I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . ?If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors


    --
    Frontend Lead, teampatent.com
    ?????????
    "People's characters are strengthened through struggle against
    difficulties; they are weakened by comfort."
    - Old Chinese adage
  • Micheil Smith at Dec 17, 2009 at 10:37 pm
    Yeah, count me in too, micheil at brandedcode.com

    On 18/12/2009, at 7:40 AM, Tom Trenka wrote:

    I started a Wave for this; if anyone wants in, send me your Wave info
    and I'll add you to it.

    -- Tom
    On Thu, Dec 17, 2009 at 11:51 AM, Tom Trenka wrote:
    An update: I just got version switching going with the tree, and am
    now displaying version info in the actual documentation.

    -- Tom
    On Wed, Dec 16, 2009 at 10:31 PM, Tom Trenka wrote:
    It might be; since my origins with JS lie on the server and not on the
    client, I can see that. I've tried to make sure that shifting this
    system to something other than PHP would be relatively painless as
    well...either way, what I'm looking to do here is to provide a
    fast-as-possible system for people to be able to effectively use
    documentation provided by the toolkit, in a way that can be used by
    anyone else writing their own systems, using this particular
    documentation format, that can be consumed with as few tools as
    possible.

    Ideally what I'd like to see is ports to your system of choice, be it
    node.js, something Rhino-based, something Python-based, something
    <insert your system of choice here>, and at the same time use this as
    a demonstration of what someone can whip up using Dijit *and* good,
    degradable practices. But first we need to show that it's possible,
    demonstrate how, and actually provide something useful (and at the
    same time do it in a way where others that grok the XML formats that
    Neil put together can duplicate). In other words, create something
    where people looking to find out about how something works is not a
    frustrating experience, and therefore moves onto something else.

    Forgive me if the tone sounds harsh; I have a long history with
    computer-based music where people expect something to be accepted not
    because of the final result, but because of the tools used to make it.
    I'm kind of hoping that we can buck that trend. And thank you for
    saying it looks kind of nice =)

    -- Tom

    P.S. For the record, I did it in PHP mostly because I know it's
    available to pretty much everyone, and (aside from Windows) is pretty
    easy to setup. Most of what I've done with this is really pretty
    basic stuff.

    P.P.S. Christ, this email is downright Faulknarian =)

    On Wed, Dec 16, 2009 at 9:32 PM, Micheil Smith wrote:
    You know, the other alternative to using PHP here might be to use Node.js, which should allow us to process the docs on the backend using the similar code to that which the uxebu guys wrote.

    Just an idea :D

    Otherwise, this is looking rather nice.

    - Micheil
    On 17/12/2009, at 4:32 AM, Tom Trenka wrote:

    Actually, you might want to check it with the nihilo theme--that may
    be part of the issue.
    -- Tom
    On Wed, Dec 16, 2009 at 11:28 AM, Tom Trenka wrote:
    Hi Shane,

    If you take a look at that index.php file, you'll see the basic setup
    (in fact, I think I might have just lifted most of it right out of the
    Dijit mail demo); toss a bunch of content panes in the main container
    under 1.4.

    -- Tom

    On Wed, Dec 16, 2009 at 10:31 AM, Shane O'Sullivan
    wrote:
    Hi Tom,

    If you could come up with a minimal test case for the tab container
    issue, I'll take a look at it.

    Thanks

    Shane

    2009/12/16 Tom Trenka <ttrenka at gmail.com>:
    Inline.

    2009/12/16 Chris Mitchell <ccmitchellusa at gmail.com>:
    Nice work!
    Here some feedback based on a brief tour of the site:
    The api web site looks very nice. It runs very fast for me in all browsers.
    That was one of the major goals =)
    Is there a quick way to search/filter the api's by keyword?
    As I mentioned to Eugene, I haven't implemented a search function yet.
    There appears to be a problem with object param documentation, especially
    where there is a large amount of docs for the param. See the detailed api
    page for dojo.data.api.Identity to see the problem.
    Hmmm. I think in this case, the documentation in the Identity API and
    the Read API are wrong; the keyword arguments objects in question
    should have been defined as "virtual" constructors outside of the item
    descriptions (see how it's done with dojo.__xhrArgs).

    Right now, I am *not* applying any Markdown to parameter descriptions.
    Perhaps that needs to change.
    Also, many private objects appear by default in the tree--most of these have
    no docs at all. Shouldn't they be filtered out in the default view?
    There's an issue with that but for now I'm simply allowing the
    privates in the class listing (see my initial email about Known
    Issues).
    Once you navigate through a number of api's, the tabs become a nuisance. A
    Close All Tabs, Close All Other Tabs is needed to reset the tab bar.
    I agree; I tried to upgrade this to 1.4 to take advantage of scrolling
    tabs but for some reason it was not being rendered properly in the
    BorderContainer it lives in. I didn't take the time to try to find
    out why.
    Where is the doc devs can use for repeating the process for creating similar
    api docs for their own code?
    When this is decent enough for a beta, I plan on (at least) writing a
    blog post about how to install this and how to generate your own docs.

    -- Tom
    Good stuff!
    -Chris
    It's good to see this using the XML data file as input.
    On Tue, Dec 15, 2009 at 5:15 PM, Tom Trenka wrote:

    Hi all,

    I've been (slowly) working on a new API doc tool, as a part of the new
    DTK site redesign effort; an alpha is now up at
    http://api-staging.dojotoolkit.org/ . If you have a few minutes,
    please take a look and shoot me feedback here.

    Known issues:
    1. The version switching does not change the tree on the left side.
    2. Privates are not being filtered.
    3. The scrolling animation is not getting to the correct point in any tab
    panes.
    4. A search function is not implemented yet.
    5. Only the 1.4 docs are fully functional (have not regenerated docs
    for 1.3 or for HEAD).

    Features:
    1. Supports full deep-linking; bookmarking the "Permalink" at the top
    of any doc page will kick you directly to that doc later on.
    2. A caching mechanism is in place that stores all ajaxified portions
    of the app; as the site is used, HTML files are being saved that
    populate tab panes.
    3. Fully degradable (i.e. if no JS, the site should still function).
    4. Somewhat themeable; basically the header and footer can be modified
    without much hassle.
    5. Fully printable; hitting control/command-P will print the currently
    focused tab pane without extra things.
    6. Deep-linking supports the current API tool URL format (i.e.
    http://api-staging.dojotoolkit.org/jsdoc/1.4/dojo.NodeList should work
    just fine).
    7. Runs/is based on the XML serialization generated by Neil's docscript
    parser.

    The site is running off the Google CDN for 1.3.2; I briefly switched
    it to 1.4 but there were issues with the scrolling tab bar and I
    didn't have the time to figure out why.

    Things I'm interested in:
    1. Help =)
    2. Any bugs you might find that I didn't list above.
    3. How useful you find it; is it a bit easier to use than the current API
    tool?

    The goal of writing this was to be able to put together a tool that
    anyone with the right XML format can use for their own documentation;
    I'm successfully using this for a current SitePen client at the
    moment, and it seems like it works pretty well.

    Dependencies:
    A PHP/Apache stack
    mod_rewrite under Apache
    XSL and cURL enabled in PHP (only for the uploading capability, the
    rest of the site should work with a bare-bones PHP install).

    Let me know what you think, and report any bugs on this thread (for now).

    -- Tom
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    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
    _______________________________________________
    dojo-contributors mailing list
    dojo-contributors at mail.dojotoolkit.org
    http://mail.dojotoolkit.org/mailman/listinfo/dojo-contributors

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupdojo-contributors @
categoriesdojo
postedDec 15, '09 at 5:15p
activeDec 18, '09 at 4:33a
posts19
users8
websitedojotoolkit.org

People

Translate

site design / logo © 2022 Grokbase