Hello everybody.
My name is Francisco Chacón, am a web developer and am working on the development of a mobile free skin for roundcube.
I realized that there is no one free skin, so i have no clear basis for start to develop it(my plan is make it free).
I woul appreciate a lot if somebody helps me with some dudes and problems i have.
One of my major dudes, is the markup language used for rounduce costumizable? I read the documentation of the languageś elements, but most of them are oriented to desktop visualization, for example the * messagelist * element is a table, but, in my opinion, the table is not practical in a mobile device (needs too much work whit css).
I appreciate previously the attention and any comment, help or advise. Thanks
On 04/12/2013 10:25 PM, Francisco Chacón wrote:
One of my major dudes, is the markup language used for rounduce costumizable? I read the documentation of the languageś elements, but most of them are oriented to desktop visualization, for example the * messagelist * element is a table, but, in my opinion, the table is not practical in a mobile device (needs too much work whit css).
Messages list widget is indeed very desktop-specific and changing it to other HTML element would be very hard. There's a lot of logic (selecting, dragging, threading, sorting, etc.) that depends on TABLE structure. Many actions depends on this and list.js code.
If you really want to do it. You could keep the table, but make it invisible (display:none) and bind to insertrow event to create your own list (and add rows to it). However, you'll still need to write a lot of javascript code to handle all actions on the list. You'll need at least to re-write the whole list.js code.
I'm not sure it is worth a try.
On Sun, 14 Apr 2013, A.L.E.C wrote:
Messages list widget is indeed very desktop-specific and changing it to other HTML element would be very hard. There's a lot of logic (selecting, dragging, threading, sorting, etc.) that depends on TABLE structure. Many actions depends on this and list.js code.
If you really want to do it. You could keep the table, but make it invisible (display:none) and bind to insertrow event to create your own list (and add rows to it). However, you'll still need to write a lot of javascript code to handle all actions on the list. You'll need at least to re-write the whole list.js code.
Is there any mileage or interest in implementing a minimal interface for smartphones which doesn't try to handle all actions on the list?
I've been looking at Horde/IMP over the last few days. Their smartphone interface appears to implement:
Compose new message (with simple integration with addressbook/LDAP, and the list of identities).
Select a mailbox
List messages in current mailbox
Search for messages in current mailbox
View message in currently selected mailbox
Reply to, Forward and Save the (single) currently viewed message
and that is about it. No drag and drop or manipulation involving multiple messages. No way to manipulate the users' addressbook or preferences.
This seems like a sensible compromise given the limitations of mobile devices. Apart from the addressbook and identity integration in (1) it could be an entirely separate application. Indeed I believe that is how it came about.
I apologise if this subject has been discussed to death in the past: I've only been on the list for a couple of months.
I apologise if this subject has been discussed to death in the past: I've only been on the list for a couple of months.
IMO, the only reason to have a mobile skin (with limited features) is contacts and may be calendaring integration. The preferable way to serve mobile devices is to use CalDAV/CardDAV. Almost all mobile devices have fully integrated clients for IMAP/CalDAV/CardDAV. So, is it really necessary to have a (feature limited) web interface? Just me 5 cents.
On Sun, 14 Apr 2013, Rosali wrote:
IMO, the only reason to have a mobile skin (with limited features) is contacts and may be calendaring integration. The preferable way to serve mobile devices is to use CalDAV/CardDAV. Almost all mobile devices have fully integrated clients for IMAP/CalDAV/CardDAV. So, is it really necessary to have a (feature limited) web interface? Just me 5 cents.
I agree that the integrated clients will almost always offer a better user experience as they are optimised for the device in question.
The one significant advantage that I can see is that the integrated clients have to be configured to use a given IMAP server, while people can just click on a URL to get the (limited) web interface.
So I can borrow someone else's mobile device to check my mail, if I happen to have left my own at home.
Am 14.04.2013 16:05, schrieb David Carter:
On Sun, 14 Apr 2013, Rosali wrote:
IMO, the only reason to have a mobile skin (with limited features) is contacts and may be calendaring integration. The preferable way to serve mobile devices is to use CalDAV/CardDAV. Almost all mobile devices have fully integrated clients for IMAP/CalDAV/CardDAV. So, is it really necessary to have a (feature limited) web interface? Just me 5 cents.
I agree that the integrated clients will almost always offer a better user experience as they are optimised for the device in question.
The one significant advantage that I can see is that the integrated clients have to be configured to use a given IMAP server, while people can just click on a URL to get the (limited) web interface.
So I can borrow someone else's mobile device to check my mail, if I happen to have left my own at home.
That's a point. From a developers point of view I wouldn't be encouraged to spend time for the development of a mobile skin (that's my point of view, I'm not a member of the Roundcube Dev Team). There are commercial mobile skins available. May be you want to check http://roundcubeskins.net/portfolio/iclassic/.
Well, the major reason of the development is just that there is no one yet (free). Also is a good point that smart phones bring up integrated facilities of webmail client.
Taking consideration of this i think that could be reasonable to develop a basic functionality mobile skin, the question would be how much basic should be......
2013/4/14 Rosali myroundcube@mail4us.net
Am 14.04.2013 16:05, schrieb David Carter:
On Sun, 14 Apr 2013, Rosali wrote:
IMO, the only reason to have a mobile skin (with limited features) is contacts and may be calendaring integration. The preferable way to serve mobile devices is to use CalDAV/CardDAV. Almost all mobile devices have fully integrated clients for IMAP/CalDAV/CardDAV. So, is it really necessary to have a (feature limited) web interface? Just me 5 cents.
I agree that the integrated clients will almost always offer a better user experience as they are optimised for the device in question.
The one significant advantage that I can see is that the integrated clients have to be configured to use a given IMAP server, while people can just click on a URL to get the (limited) web interface.
So I can borrow someone else's mobile device to check my mail, if I happen to have left my own at home.
That's a point. From a developers point of view I wouldn't be encouraged to spend time for the development of a mobile skin (that's my point of view, I'm not a member of the Roundcube Dev Team). There are commercial mobile skins available. May be you want to check http://roundcubeskins.net/**portfolio/iclassic/http://roundcubeskins.net/portfolio/iclassic/ .
______________________________**_________________ Roundcube Development discussion mailing list dev@lists.roundcube.net http://lists.roundcube.net/**mailman/listinfo/devhttp://lists.roundcube.net/mailman/listinfo/dev
On 14.04.2013 16:44, Rosali wrote:
That's a point. From a developers point of view I wouldn't be encouraged to spend time for the development of a mobile skin (that's my point of view, I'm not a member of the Roundcube Dev Team). There are commercial mobile skins available. May be you want to check http://roundcubeskins.net/portfolio/iclassic/.
My problem with such skins is that they aren't just a mobile add on to the stock skins but whole new user experience which might cause some confusion when switched on by default.
i'd rather like to see an out of the box solution for larry and a fallback option for skins that do ship with a mobile skin.
i would spend some money on this kind of skin
cheers, Raoul
On Sun, Apr 14, 2013 at 12:35 PM, A.L.E.C alec@alec.pl wrote:
On 04/12/2013 10:25 PM, Francisco Chacón wrote:
One of my major dudes, is the markup language used for rounduce costumizable? I read the documentation of the languageś elements, but most of them are oriented to desktop visualization, for example the * messagelist * element is a table, but, in my opinion, the table is not practical in a mobile device (needs too much work whit css).
We're actually happy to hear that someone is willing to contribute a free skin for mobile devices as this was requested several times and we even put it on our roadmap for future releases of Roundcube. But as you now experienced, some refactoring is necessary to make this happen. We'd really appreciate your work on this. But please consider to more focus on a responsive skin that scales for a variety of devices from desktop / tablets down to mobile phones. A pure mobile skin isn't the most sustainable solution for today's needs.
Messages list widget is indeed very desktop-specific and changing it to other HTML element would be very hard. There's a lot of logic (selecting, dragging, threading, sorting, etc.) that depends on TABLE structure. Many actions depends on this and list.js code.
That's absolutely true and that's (hopefully) the only component that would require changes in the core to make a mobile skin reality. Beside the table structure which isn't not very handy, I assume having something like checkboxes for row selection might be necessary as well.
If you really want to do it. You could keep the table, but make it invisible (display:none) and bind to insertrow event to create your own list (and add rows to it). However, you'll still need to write a lot of javascript code to handle all actions on the list. You'll need at least to re-write the whole list.js code.
That's something we should do because getting a mobile skin is of substantial interest for Roundcube, I'm willing to support you as good as I can. Please give us some time to get over the list code and we'll see how we could bring support for a <div> or <ul> structure to our core list widget.
I'm not sure it is worth a try.
We generally agreed that we want a mobile skin after the recent survey and discussion and this is IMO the first step to take.
~Thomas
On Mon, Apr 15, 2013 at 01:50:28PM +0200, Thomas Bruederli wrote: ...
happen. We'd really appreciate your work on this. But please consider to more focus on a responsive skin that scales for a variety of devices from desktop / tablets down to mobile phones. A pure mobile skin isn't the most sustainable solution for today's needs.
We have a lot of interest in this. It is clear that a lot of users access their mail via webmail on their tablets. We all know that they'd be better served using the native app but there are cases where this may not be an option. We'd be happy to sponsor this. We're going to see what we can do to make the larry template at least function on an ipad - I don't have the details but apparently there are issues which renders it "unusable."
It is exciting to see that there is community interest in a mobile skin for RC. As it so happens I am on contract with a large American retailer with a Kolab Server based groupware environment, and my current project is developing a mobile skin for RC that will be used in house, but will also be released as an open source project once we have it production ready. Our approach is to create a UX that is much like a native mobile app, and not just a smaller, scaled down "Larry" or something like that. In fact our goal has been to refactor the whole client side (as much as is necessary at least) of RC from a mobile first perspective, and stay as far away from looking, and feeling like Larry as possible. Note that there are NO changes to the RC code base. Everything is accomplished within the scope of the skin.
We have a requirement to support "all" possible mobile devices (At what point does that request become unreasonable?), because with a couple of hundred thousand users all owning different devices we cannot realistically limit what the users would be "required" to use in order to functionally use the application, nor can we predict what they are using now, or will be in the future. Besides, for it to be truly useful to the RC community at large the skin must be as platform independent as possible.
On 4/15/13 10:58 AM, Kelsey Cummings wrote:
On Mon, Apr 15, 2013 at 01:50:28PM +0200, Thomas Bruederli wrote: ...
happen. We'd really appreciate your work on this. But please consider to more focus on a responsive skin that scales for a variety of devices from desktop / tablets down to mobile phones. A pure mobile skin isn't the most sustainable solution for today's needs.
We have a lot of interest in this. It is clear that a lot of users access their mail via webmail on their tablets. We all know that they'd be better served using the native app but there are cases where this may not be an option. We'd be happy to sponsor this. We're going to see what we can do to make the larry template at least function on an ipad - I don't have the details but apparently there are issues which renders it "unusable."
On 04/15/2013 08:16 PM, Aaron Stephanus wrote:
It is exciting to see that there is community interest in a mobile skin for RC. As it so happens I am on contract with a large American retailer with a Kolab Server based groupware environment, and my current project is developing a mobile skin for RC that will be used in house, but will also be released as an open source project once we have it production ready. Our approach is to create a UX that is much like a native mobile app, and not just a smaller, scaled down "Larry" or something like that. In fact our goal has been to refactor the whole client side (as much as is necessary at least) of RC from a mobile first perspective, and stay as far away from looking, and feeling like Larry as possible. Note that there are NO changes to the RC code base. Everything is accomplished within the scope of the skin.
Great. Here's the list of known bugs related to mobile usage: http://trac.roundcube.net/query?status=assigned&status=new&status=re...
On Monday 15 of April 2013, Aaron Stephanus wrote:
a mobile skin for RC that will be used in house, but will also be released as an open source project
That's great...
once we have it production ready
while that's bad. Make it open development, so other can help and test early.
Thomas Bruederli wrote:
Messages list widget is indeed very desktop-specific and changing it to other HTML element would be very hard. There's a lot of logic (selecting, dragging, threading, sorting, etc.) that depends on TABLE structure. Many actions depends on this and list.js code.
That's absolutely true and that's (hopefully) the only component that would require changes in the core to make a mobile skin reality. Beside the table structure which isn't not very handy, I assume having something like checkboxes for row selection might be necessary as well.
If you really want to do it. You could keep the table, but make it invisible (display:none) and bind to insertrow event to create your own list (and add rows to it). However, you'll still need to write a lot of javascript code to handle all actions on the list. You'll need at least to re-write the whole list.js code.
As of commit https://github.com/roundcube/roundcubemail/commit/517dae3e74 we have the list handling code refactored to move towards what you need for a mobile skin. With the recent changes you can now generate <ul> or <div> style lists with
<roundcube:object name="messages" id="messagelist" tagname="ul" noheader="true" />
Please note that some functionality such as column-rearrangement will only work with the default table tag but I assume that doesn't matter on mobile devices anyway.
We'll continue with an option to show checkboxes for row selection which currently still happens on simple clicks. So stay tuned...
Please let us know if you can advance with mobile skin development now and what's still missing.
Regards, Thomas