Re: status of editing

After Feb 13th generally works for me.

On Tue, Jan 17, 2017 at 2:54 PM, Gary Kačmarčík (Кошмарчик) <
garykac@google.com> wrote:

> Feb (in general) works for me.
>
> On Tue, Jan 17, 2017 at 11:33 AM, Grisha Lyukshin <glyuk@microsoft.com>
> wrote:
>
>> How about some time in February?
>>
>>
>> Sent from Outlook <http://aka.ms/weboutlook>
>> ------------------------------
>> *From:* johanneswilm@gmail.com <johanneswilm@gmail.com> on behalf of
>> Johannes Wilm <johannes@fiduswriter.org>
>> *Sent:* Tuesday, January 3, 2017 4:02:02 PM
>> *To:* Piotr Koszuliński
>> *Cc:* public-editing-tf@w3.org
>> *Subject:* Re: status of editing
>>
>> Hey everyone,
>>
>> given the current situation, I think we should have a call within the
>> next next few weeks. How would the week between January 18 and 25 work for
>> others? If not then, do you have alternative suggestions?
>>
>> On Tue, Jan 3, 2017 at 2:53 PM, Piotr Koszuliński <
>> p.koszulinski@cksource.com> wrote:
>>
>>> Hi Johannes,
>>>
>>> > * The undo stack is global, which means it's broken for every editor
>>> we have been able to find on the net, including those managed by all the
>>> browser maker companies. It would be good if we could figure out how to
>>> replace the global undo stack for contenteditable with separate undo stacks
>>> for every contenteditable element (could be an optional setting if this
>>> works best for Safari, even though no existing editor uses the global
>>> setting) [3].
>>>
>>> It's a very good point that this is broken for everyone. I can
>>> understand Webkit's team rejecting the proposals to expose the undo
>>> manager because that would be impossible/hard to integrate with the OS or
>>> browser. But the truth is that the situation is totally broken already and
>>> currently every RTE I checked implements its own undo manager, which
>>> completely ignores what the browser tells it. It's also unacceptable for
>>> RTE authors to have a global undo stack (we've taught users that each
>>> editor handles undo separately [1]). Finally, I don't understand how the
>>> browser's undo stack is supposed to work with RTEs implementing custom data
>>> models and collaboration features. Ryosuke pointed out [2] that W3C
>>> "specifically worked with Google Docs team to ensure their undo worked with
>>> the API", but I don't understand how was that supposed to work. It'd be
>>> interested to see some PoC or discussions, because it may turn out that the
>>> proposed Undo Manager API would be acceptable.
>>>
>>
>> It has been pointed out earlier that the Undo Manager API proposal was
>> too complex for various reasons, but that it it included a way to define
>> the scope of the undo and that this part could be used to make undo more
>> local. This may be a good idea, even if it's just a complex way to say that
>> global undo is never desired anywhere for richtext.
>>
>> I came to the same conclusion as you, and found that also Google, Apple
>> and Microsoft softwares are broken. It's almost not noticeable if one
>> doesn't know where to look, but just find two different places where there
>> is text input (for example the search bar and the email composing part in
>> Gmail). Write a little in one, then click into the second. Type a little
>> more. Now undo all of it. You'll notice that they're all broken, in all the
>> browsers and all the OSes. Either the undo/redo is deactivated when it
>> should be activated, or it doesn't undo when it should.
>>
>> It seems that JS editors really only want direct control voer enabling
>> and disabling the native undo buttons and listen to the beforeinput events
>> for both of them. But I understand that due to OS restarints, Safari cannot
>> do this. Instead they seem to argue that one could get a simple undo
>> manager where JS can manually add items and make the undo scope be local..
>> This seems to be almost as good, although it will likely create problems
>> for collaborate editors, when a change of user A mean that the last 7
>> changes of user B no longer have any meaning.
>>
>> The main point here is that we really need to get going with this. This
>> should be in the interest of all the involved organizations and companies,
>> as the undo/redo menus are broken for all of them, and have been for a very
>> long time.
>>
>>
>>>
>>> > * There is a large, opverlapping menu on iOS giving formatting
>>> options. This is problematic for two reasons: 1. It overlaps the texteditor
>>> 2.
>>>
>>> I agree with everything you wrote, but I'd like to add one thing here.
>>> It's a much broader topic, but we've been researching how we can show our
>>> own controls on Safari@iOS and it turns to be extremely hard when the
>>> on-screen keyboard is visible. As far as I understand, Safari implements
>>> some non-standard viewport mechanics which makes positioning things very
>>> hard (if not impossible). From what we've seen, it all works as you'd
>>> except in Chrome@Android.
>>>
>>> This means that not only the menu is overlapping with our controls and
>>> that we can't control it, but we also can't reliably display something on
>>> the screen when the keyboard is visible. So the situation is broken on 3
>>> levels.
>>>
>>
>> I believe I saw a long description with images in a report written by a
>> CKEditor person some months ago. Do you have the link for this?
>>
>>
>>> >  The issue with the non-available features in editors has apparently
>>> become worse with the "Touch Bar" on Macbook Pro. While Safari always had
>>> some editing options hidden in an obscure menu that don't seem to work in
>>> any of the existign editors, some of these formatting options are now more
>>> prominently placed, which means it will be more obvious when they don't
>>> work [5].
>>>
>>> This is really sad. We've been working to gain more control over the
>>> editing experience and suddenly a font color picker appears in the "Touch
>>> Bar". I can even understand bold, italic and lists which are what more than
>>> 90% RTEs enable (although, not all – see Twitter). But font color doesn't
>>> appear in any modern editor because it's a non-semantic styling option
>>> which, in most cases, content authors should not be able to use. Exposing
>>> features like font color picker in the touch bar moves us back to 00's [4].
>>>
>>
>> Well, I can see that there are 7 Billion people on this planet and with
>> so many different writing styles and needs, there is likely also a
>> community out there that happens to want these features. And having direct
>> access to some of the richtext editing features right on the keyboard
>> sounds pretty neat.
>>
>> But I must agree with Piotr that this isn't what the main editors
>> currently are interested in. I wonder: Has Apple considered whether to open
>> up these various formatting menus (on iOS and macOS) so that the JavaScript
>> editors can enter their own menu items in there and replace the existing
>> ones? It seems like this would allow both for you to keep your menus, while
>> alleviating some of the frustration these editor devs have had when dealing
>> with Apple products.
>>
>>
>>
>>>
>>> [1] https://github.com/w3c/editing/issues/150#issuecomment-249815640
>>> [2] https://github.com/w3c/editing/issues/150#issuecomment-249775255
>>> [3] https://github.com/ckeditor/ckeditor5-design/issues/149
>>> [4] https://twitter.com/reinmarpl/status/815891250612174848
>>>
>>> --
>>> Best Regards,
>>> Piotrek Koszuliński | CKEditor Lead Developer
>>> --
>>> CKSource – http://cksource.com | Follow CKSource on: Twitter
>>> <http://twitter.com/cksource> | Facebook
>>> <http://www.facebook.com/cksource> | Google+
>>> <https://plus.google.com/+CKSource> | LinkedIn
>>> <http://www.linkedin.com/company/cksource>
>>>
>>
>>
>>
>> --
>> Johannes Wilm
>> Fidus Writer
>> http://www.fiduswriter.org
>>
>
>