There also is a lot of talk about increasing GNOME market share. I really believe the biggest growth area could be in the business environment, and groupware is one of the most important pieces. Email/Calendaring, Web Surfing and Word Processing are the keys to that market segment.
Improvement Areas:
- Proxy into GroupWise creates a massive folder tree, that is too hard for users. For many if they have to scroll too far, they will never find it.
- People have trouble sometimes with window management, and using the stack and working on multiple messages at the same time.
- Searches are not retained.
- Tomboy notes should be better integrated into email messages and calendar events. It sure would be nice to have notes that you have taken and attach them to a meeting.
- People still talk about and miss the old Summary page from version 1. It was nice to get a high level view of various components.
- More thumbnails please in the attachment area.
- Hide things like .zip files from users and extract them into individual files with thumbs. They shouldn't have to know how to extract the files and where to save them.
So I created mockups. Harish, Chen, Varadhan if you could go ahead and have this done over the weekend, that would be great. :)
35 comments:
I really dont think the fact that its a zip file should be hidden entirely. I often get zip attachments that i do not want to unzip right away. Hiding the fact that its a zip file could also cause some confusion if, for example, the sender later asks if you got the zip file, and you reply "No, there was no zip file attached to the mail".
I'm sure there has to be a way to both allow people to extract and save the compressed files directly while still not hiding the fact that they are part of an archive.
I also would like:
* Tags
* Group messages by date (day, week...)
* Group messages by sender
* Contact groups
* Group/MailingList filters
I would like to click somewhere like "MyTeam" and see all the mails of the people of certain group, let's say. And also something like "SomeMaliningList".
Yeah, Mailinglist support MUST be integrated ASAP.
KMail handles that very well!
Yes, I like your line of thought very much. The mockups are interesting, but they've too many tabs.
Evo3 is definitely something to work towards, though.
@dave: the summary screen was removed in 1.5 because nobody maintained it anymore. it was a nice one at that time, though.
i'm happy to see somebody coming up with ideas, and i share the "no new features" thought. however, this would be bad for marketing, you know...
@felipec:
contact groups have been available for years, also mailing list filters for 2 years now. which old version are you running? ;-)
I would love to have a "pure" calender view without the task list nor the left frame. My problem is that when I use the month view the entries are too small to be readable. Therefore I would like to be able to spend all the desktop area on the calendar.
Anyway I just wanted to wish you good luck with evo 3.0
eek! too many tabs?
But I definitely agree that Evolution needs to develop its own identity. It currently feels more like Windows than GNOME.
I really like the idea of integrating the composer directly into the main evo window rather than having floating compose windows. Very nice.
Please, no more Xin1 applications...
Split out in Mail,Calendar,Tasks,Todo
Window switchers do not belong inside one application. We already have a taskbar?
But expand on the evo data server idea!
(like adding rss/atom handling?)..
And.. we don't need a summary? If Calendar/Mail were two seperate apps i could organize them side by side. Summary is only needed because evo tries to do window management. Don't do that!
Sorry for being so negative - keep up the brainstorming ;)
I've always thought that evolution would be various thousand of times better if it would be three integrated apps and not the konqueror-like monster it's now. There are things called windows and evolution isn't exactly using them. For instance I'd like to have a mail-only evolution, still with the ability of creating appointments from mail using the calendar app, notes using tomboy and using (and automatically creating-on-send) contacts in the contacts-app, right now evolution-data-server makes that totally feasible (in fact, both contacts and dates, the lightwight pim apps show that), and we 'd take out a lot of the ui overhead involved in having such a modular app.
Oh!, also, if I clicked a mailto link in a page, just the message composer should get opened and not the main mail-browsing window!
I agree with Zith - automatically showing the contents of archive files isn't a feature I'd want, as I tend to receive quite complicated archives such as source tarballs.
Although that said, showing the zip file with a little drilldown button next to it would be useful - click the button to show something like your mockup view. Because while it's the less common use case for me, every now and then someone sends me a zipped up spreadsheet or word document that I'd like to access without needing to open the archive in file-roller first...
It's refreshing that people are actually discussing major work on Evolution, as I feel that, at present, it's the one part of Gnome that just doesn't feel up to par. I think, though, that honestly, most of that is because of internals, and not the interface. Locking on network I/O, for example, is really irritating.
As for your mockups, they're nice, but I agree with the "too many tabs" sentiment, and, in particular, I anticipate some issues with the "each account gets its own tab" thing. I would imagine an interface going the *other* direction, in favor of merging more data into a single view. Why, for example, should the search be per-account (as it currently is, and seems to be in your mockup)? Why shouldn't I be able to easily search or filter all of my mail, regardless of the destination account?
Another thought: now that lots of the logic is abstracted away in EDS, it seems like it might be prudent, especially if the whole interface is being reworked, anyway, to move the UI away from C, and into C#, or maybe Python (though, given its connections with Novell, C# seems the most likely). Most of the resource-intensive stuff (searches, parsing through piles of messages, etc.), should still happen in C, but moving the UI to a higher-level language probably wouldn't result in much of a performance loss, and would certainly encourage contributions from others (myself, included), who don't often feel like wading through mountains of C to tweak UI issues.
Re: attaching tomboy notes. In principle a good thing. However, tomboy notes should be handled in the same way as any other email attachments. Just make it easy to attach them as you would attach ZIP, PDF, or any other file. Treating tomboy notes differently makes it more complex for no reason.
The "attachment" column in the mailbox view could show an appropriate icon for the type of attachment, instead of just a paper clip, for each email with an attachment (including an icon for multiple files).
...just my thoughts on the matter.
One vote here in favor of a ("massive") single folder tree ... much faster to scan for new items between my (2) primary email accounts by draging a single scrollbar.
Why should evo depend on Tomboy? I don't want that.
The issue with the top tabs that display email addresses is only used with the GroupWise backend. If you have multiple email accounts that are all your own, they still would appear down the left edge.
Proxies allow you to log into other peoples email and calendar and view it. Right now what it does is merge all of their items into yours...and it's confusing for users and causing the conflict code to not work correctly.
That whole tab bar would not be there if you never Proxy into GroupWise.
I am a daily evolution user and not a designer.
And I tell you that I don't need as a user any additional feature.
I need Evolution to be fast and not to crash.
If this is achieved I would be happy.
But now, everytime I launch evolution I'm frustrated by its slowliness and how buggy it is.
Evolution 3.0 should stands for fast and reliable.
ZIP decompression should not be handled by Evolution. Stuff like that should be handled by Nautilus, transparently. The only way users should notice the difference between a directory and a compressed folder is - at most - the different coloured folder with an embedded compression icon.
IMO evolution should be just discontinued as it is big, bloated and uses an aged codebase.
why do I have to start my calender in order to manage my contacts?
I think evolution should be split up into mail client/ calendar/ contacts manager, which use EDS or better a freedesktop standard for this stuff.
I've gotta say, monolithic evolution is a pain. I thought with the addition of Evolution Data Server we'd be able to break things up more. Having E-mail, contact management, calendar, notes, and tasks all embedded into one big application locks a user into a single application, even though eds allows them to use many. And as a man who's used oo nautilus for a while, I don't mind extra windows, especially if it means I can replace "Tasks" with a smaller program more taylored to my needs.
In no way would I oppose splitting Evolution into pieces. Remember when the old Star Office was one massive 'desktop' and it was split into pieces years ago?
I guess my one big question is why this should be Evolution-based at all?
There are at least 3 competing web-based email/PIM solutions that can do this kind of functionality today.
Throw in the offline support that Mozilla is working on right now (that Zimbra already uses) and you've got the solution you want. And it's not tied to Linux or Gnome.
Alex: I just have never warmed up to web based applications. We looked at Groupwise webaccess, and it's not strong enough to use. Even the newer AJAX applications like Yahoo mail are clunky and slooooooow. We aren't really experiencing any speed problems with Evolution. Evolution provides a consistent look and feel with other GNOME apps. Java and web apps just do not fit in and increase our support calls because they behave differently.
I have two accounts, one for personal and for work-related mails and I only really use one mail folder per account, the inbox. Hopefully, even with the new per-account tabs (if I understand the mockups), Evolution 3 lets me keep the current setup where _both_ accounts are visible at the same time; otherwise the tabs are a terrible usability regression for me.
Next gen email/calendering/contacts etc needs to combine seamless integration with the desktop (ie, not some clunky separate program) with easy remote (web) access.
This needs to be thought through carefully. What happens if a user doesn't have a handy LDAP server to hand and a nice corporate support setup to deal with it? A home user would like to be able to go to a website that works like hula (for example) with the same info that they would get at their local machine.
As regards a monolithic evo and , being able to click an email link and a small email editor pops up is exactly as it should be.
I agree with the general sentiment that splitting up Evolution would be a good thing. It might also make porting the app somewhat easier. It would be HUGE if Evolution were easier to port to Windows and OS X.
Niiice!!!
One thing to polish out is the component buttons (mail, calendar...). Having them at the bottom is strange, not nice for new users.
Pehaps some vertical tabs at the left, to change between components, and distinguish them using some kind ob bg pixmap on the toolbar, so you notice you are in the email component at a glance.
Or separate windows for each component, as someone suggested. They could be tightly integrated (open calendar when you choose save this email as a taks/accept a meeting...) without the need to be opened on the same window.
Yes! I want Kontact ähh GKontact (mail, feed, kalendar, adressbook)
One other thing I'd love to have in future Evolution is capability to do intelligent things.
The kind of things other email clients does currently, some of which I could remember now are:
1. Automatically save drafts, so when I forgot about it, and just close Evolution (or evolution crashed) the drafts will not be missing. I don't know if this is currently already implemented inside Evolution?
2. Automatically save contacts of people I sent mails to.
And then I'd agree to the opinions of making Evolution lighter and have a better integration with GNOME.
Also better interaction (more crash proof would be fine) with MS-Exchange, I know MS-X sucks, but that doesn't mean Evolution need to suck when interact with it also, right?
there are many more ideas that I just forgot about right now.
in reply to arief:
1) evolution does this already, it autosaves the mail from time to time (but not to draft). if evo crashes on next launch, it asks you if you want to recover the mail you were typing.
if you exit evolution without sending the mail you were typing, evolution ask if you are sure (and don't want to save the mail).
2) this is already done too
3) for ms-exchange, well I don't know of any other client that does this currently. It's true that support varies greatly for different versions of exchange server but that's what you get from proprietary closed software right ?
Oh god... nested tabs?! Never, ever, ever. Please think of the children.
Keep it simple. Please.
That is Evolution's #1 feature over Outlook. Outlook has more preferences and options than Windows itself. Its a massive piece of bloatware that even its biggest fans admit is sluggish and awkward at times.
Keep Evolution Gnome-ish: Keep it simple!
have a look at this combined GMail and iTunes email layout.
When Evolution first came out there was nothing else like it in Linux. I switched to it immediately.
Then Thunderbird was released and blew Evolution out of the water with its speed, stability and huge number of Add-ons.
Alas, my employer decided that everyone must use the Microsoft Exchange server (with IMAP turned off!) so I had to switch back to Evolution. Now I have a 15-20 minute wait each morning while Evolution filters my mailing lists and spam.
If you really want to improve Evolution you should be concentrating on speed and stability, not bells and whistles. If you want Evolution to succeed in a business context then you must have rock solid Exchange support. Throw out the clunky OWA based connector and write a proper MAPI compatible plugin.
Yes, evelution should be splitted.
I am a professional user who has to use MS Exchange in my company and I cannot use Evolution because
- the OWA conector is so buggy,
- there is no MAPI support in Ubuntu so from April this year onwards I can't use any Evolution any more.
My colleague posted important bug information in some important bug tracker and he says no developer ever seems to read it.
Post a Comment