JAN
18
2001

Aethera Messaging Client Beta 1

Good news from theKompany.com keeps pouring in. As several of you have pointed out, the first beta of Aethera has been announced. In case you haven't been following, Aethera is theKompany's fork of the greatly hyped/anticipated Magellan project. Beta 1 of Aethera sports POP3, SMTP, HTML, DnD, a contacts interface, sticky notes, and more. IMAP, Calendar support, etc are promised for the next beta. There is no mention of the license although source is available from the website -- most of the source files seem to be under the BSD license. For more details on Aethera, Kevin Reichard also points us to this favorable review on LinuxPlanet by Dennis E. Powell. Be sure to check out the screenshots as well.

Comments

I'll second that. Shawn could have taken Magellan, polished it, made it binary-only and sold it. He didn't. He re-licensed it under the GPL, protecting his derived works.

The BSD license has its place, as does the GPL. What has happened here is becuase of the original choice to go with the BSD license. We're all just lucky that Shawn GPL'd his changes.

Besides, didn't he fund Teddy for a while?
(Sorry if I'm wrong..)

Ben


By Ben Hall at Sat, 2001/01/20 - 6:00am

This wasn't an easy decision to make, and we didn't do it lightly, but we tried to do it correctly. Also just to be clear, I didn't fund Teddy in any way.


By Shawn Gordon at Sat, 2001/01/20 - 6:00am

Well, the issue is that it is just the polite thing to do to ask the maintainer of the project for his opinion and try to work with him as much as possible.

It seems to me that from Teddy's post theKompany did not devote any sort of real resources to the project and it is just building on the work from Teddy.

It just seems very rude


By Anonymous Coward at Mon, 2001/01/22 - 6:00am

Please don't make opinions or pass judgement without knowing all the facts. I really don't want to get into details because it is not productive for anyone. We spent many many hours on Magellan, but the bottom line is that we both had different ideas of what to do with it. From your perspective it gives you a choice of what you want to use.


By Shawn Gordon at Mon, 2001/01/22 - 6:00am

Hey Shawn,

If you are not sharing your code with Teddy, then in my opinion you are using the GPL as a weapon against his email client rather than co-existing in "open source". You are in other words using his work against him. That's just my opinion however. You could explain yourself better on this subject to clear up all misconceptions.


By ac at Mon, 2001/01/22 - 6:00am

At the point in time that we put the GPL notice in Aethera, the Magellan site had conflicting information, at some points it said GPL and some it said MIT/X. The code actually said MIT/X, at least the version we had as the Magellan CVS became suddenly unavailable about a month before we did our fork. Given this conflicting information, we weren't sure if Teddy had changed the license and it seemed safer to make sure that the Aethera code was GPL just to be sure.


By Shawn Gordon at Mon, 2001/01/22 - 6:00am

"Within a month there will be very little of the original Magellan code in Aethera"

until then, you should have given the proper credits. shame!


By Evandro at Wed, 2001/01/24 - 6:00am

Evandro we have done that. Look at our web page at http://www.thekompany.com/projects/aethera/overview.php3 the very first sentence gives this credit, and the credit was in the story header. I've given Teddy credit at every turn.

Please check your facts before you point your finger of shame.


By Shawn Gordon at Wed, 2001/01/24 - 6:00am

While other companies just talk and talk the Kompany delivers! Good work Shawn and the guys.


By matt at Thu, 2001/01/18 - 6:00am

There seem to be two classes of PIMs.
The first are ones like Aethera. The second are ones like www.thebrain.com or www.memes.net (OK they aren't usually called PIMs).
The ideal PIM would be one that combined both of these approaches into one application.

For instance, entries in the contact list could have links to other contacts, emails, or notes.
So that when I receive an email from Bob, I could pull up the contact entry for Bob and discover I had added links that point to his girlfriend Jill's contact record, links to organizations he belongs to, a link to a note I entered with info about our last conservation, links to the emails we exchanged, links to contact records for our common friends,
links to calendar entries involving Bob, etc. Each of these items could then have its own links.

Links could have have user entered descriptive info like:
"works for", "first cousin of", "interested in",
etc.

The structure of links would not be preordained but could evolve so that each user could grow their own web of information.

Jim


By James H. Thompson at Thu, 2001/01/18 - 6:00am

Yes! I agree thats what we need to make Aethera leading edge - lets not just aim to have features comparable to M$ Outlook, lets inovate!

Talking about inovation, I like the idea of mobile SMS messages staight from Aethera, WOW!!! I hope the next preview will have some of this, or has this being dropped and only available in Magallan?

Thanks again theKompany for yet another Kick-ass app!


By iCEmAn at Thu, 2001/01/18 - 6:00am

If you look at the Aethera roadmap you will see that we have a number of unique and innovative features, there are more that we are working on but didn't want to talk about because it seems like every time we do the feature suddenly shows up somewhere else :).

SMS and all the other things will be in Aethera, don't worry.


By Shawn Gordon at Thu, 2001/01/18 - 6:00am

What?! Other projects are copying an open-sourced project's features? I'm shocked!

Seriously... I don't understand your logic. If Aethera is one of your GPLd (i.e. non money-making) products, why are you worried about others copying your features? Isn't that what open-source is all about, the ability to copy other people's ideas and/or code?

confused,

-Avdi


By Avdi at Tue, 2001/01/23 - 6:00am

> www.thebrain.com

My God!! A dynamic mind-map interface to all my data .. that's [stunning] !! The difficulty of trying to keep all the years of information & email I've collected in some form of organasation that remains useful over time has been driving me nuts for ages. Heirachical folders run out of steam beyond a certain point and info just gets lost or fogotten about.

My first eye opening experience to the possibility of organising data differently came several years ago in the form of "ML" - an IMAP mail client from a guy at Stanford Uni that allowed you to use filtering rules to create "Virtual Folders". It was amazingly powerful ; but died a death because 1) the guy stopped developing it and went to work for Netscape ; 2) it staticly linked in Motif; and 3) its dependance on large flat file mail files (behind IMAP) ran out of steam performance wise when they got too big (which they quickly did). Realisticly it needed a back end that a) used a database and b) had some built in smarts that enabled it to cooperate with the Virtual Folder generation.

It will be interesting to see if the Gnome folk can re-visit this with Evolution and take the concept onto the next level.

I do like theBrain's mind-map interface though .. I just can't way WOW enough times. This could be the answer to my prayers. I wonder if they've ever heard of KDE ;-)

John


By John McNulty at Thu, 2001/01/18 - 6:00am

This is the direction on which gofai-wmt is headed. It is already doing similar things for webpages with konqueror, try it in autopilot mode.

In fact, if you would have a contact list client with a sufficiently reach DCOP interface, the functionality could be added in an afternoon.

Lotzi


By Lotzi Boloni at Thu, 2001/01/18 - 6:00am

Is gofai-wmt still alive? -- its web pages seem to have vanished. (I tried: http://www.cs.purdue.edu/homes/boloni/GofaiWeb/)

The key to making something like this work is ease of use. It's not clear you could accomplish that easily using a separate program.

There is also the issue of how objects get annotated/extended. For instance for a contact you would like to be able to put in as many phones numbers as you wish, each with some note: 'home', 'summer home', 'tuesdays work place', 'european cell number', etc. The only PIM I've seen that handles this gracefully is http://www.ajsystems.com/amigo.html


By James H. Thompson at Thu, 2001/01/18 - 6:00am

Yes, it is alive.

http://www.geocities.com/boloni2/GofaiWeb/index.html

or try kdenonbeta/gofai-wmt


By Lotzi Boloni at Thu, 2001/01/18 - 6:00am

If Outlook Express is stripped down Outlook, why doesn't Outlook support newsgroups? I ask myself that every time I have to load that crappy program at work and then load a newsgroup reader to check my subscribed newsgroups. Ugh! I'd like to see Aethera support newsgroups. And I think that the previous post about making the different chunks of your program into KParts is essential! Looks pretty damn cool though!


By james at Thu, 2001/01/18 - 6:00am

It's also frustrating that OE has better Mail Rules that Outlook is missing. I'm amazed that MS couldn't use ALL the features of OE in Outlook PLUS some. In many ways - OE is superior to Outlook in the email program part of it.


By Laura at Thu, 2001/01/18 - 6:00am

who told that oe is a descendant of outlook? as far as i know both programs do not share a common root...


By thePooh at Fri, 2001/01/19 - 6:00am

Hi Shawn,

I downloaded and installed this last night, and it looks real sharp. My only complaint is that it seems to store it's data in binary format. I have a large flat file containing 150+ contacts, and I was hoping to convert it into Aethera's format with a python script. Is there any reason that this is not stored in plain text? XML would seem to work very well here. If there is some reason for the binary format, is it at least documented somewhere?

Just my $.02.


By Larry Wright at Thu, 2001/01/18 - 6:00am

Well, we are working on some import features, so we will more than likely make it possible to create custom import routines.


By Shawn Gordon at Thu, 2001/01/18 - 6:00am

That's good to know, but was there a reason for choosing a binary format?


By Larry Wright at Thu, 2001/01/18 - 6:00am

Don't bother about import features, that will be an endless nightmare. Just provide a dcop interface to add new contacts. Then any importing script can just do

dcop aethera add_new_contact firstname lastname address yada yada yada


By Guillaume at Fri, 2001/01/19 - 6:00am

So, now what we need is for kmail and kdepim to start rolling in the good parts of Aethera, so that when the proprietary modules start rolling in, open source alternatives can be made available.


By Neil Stevens at Thu, 2001/01/18 - 6:00am

And that will be the perfect way to make sure that we don't make money on anything and go out of business.

What bothers me is the notion that open source must also be free of charge. I honestly have no problem with including the source to an application that someone has purchased, or that we are giving away for free. What bothers me is this misconception that the software should also be free of charge. This is what forces a company to create closed source so that they can protect their financial interest.

I've been coding for 22 years and creating commercial software for 15 years now, and it seems I've spent half my life coming up with copy protection devices to avoid piracy. This was one of the lures of Linux to me was the idea of Open Source, but I've been disappointed to a certain degree on the misunderstanding of "free".

95% of the email I get is very supportive in both financial and cheerleading, but I do get that 5% that are angry at me for having the audacity to charge for our work. It really blows me away.

So you see, the "for sale" modules of Aethera can be open source as well, but that depends on the public.


By Shawn Gordon at Thu, 2001/01/18 - 6:00am

Right on, Shawn.
Don't worry, as soon as all the "all open-source must be free" fellows get their first mortgage, you'll be amazed how quickly they'll become more realistic in their expectations :-)


By Jacek at Thu, 2001/01/18 - 6:00am

Well, i don't think you have a problem in this case, since the Server version will be Proprietary.

btw: why don't you create a new licence in which you will give the proprietary software, in which anyone will be able to modify the code for personal use, but only *IF* he was purchaced a copy, and he will not be able to (legally) distribute it or create something else with the code without agreement from your company...


By t0m_dR at Thu, 2001/01/18 - 6:00am

The server is only one of the pieces we will sell. All the normal features of a Mail, PIM and Newsreader will be free, it's some, not all, of the unique features that will carry some sort of price.

Another license is an interesting thought. I've been skiddish on this because of the heat Trolltech (wrongly) got over the QPL, but we can dual license it. I'll have a talk with my high priced attorneys and see what we can come up with :).


By Shawn Gordon at Thu, 2001/01/18 - 6:00am

If you come up with a license that will allow paying customers to modify and improve the code, and allow those customers to share their improvements (but without the problems that were the downfall of Minix), that would completely change things.

It would change things for the better. Customers would retain freedom. You could get that revenue from selling copies that you want. In this situation, everybody wins, right?


By Neil Stevens at Fri, 2001/01/19 - 6:00am

Hi Shawn.

I've been using Linux for four or five years now, and I have yet to find an e-mail client that holds a candle to those found on other OSs. The closest thing I've got so far is Pine, which, while amazing, sucks on many levels.

If you come up with _the_ client, I'll gladly pay for it. Especially if it's Open Source!

I personally think that there's plenty of money to be made in OpenSource software. I also think that developer time should be paid for, not the actual products. This is why I like the CoSource idea so much. You've got your plans, and I think they may work too.

The idea of giving the source along with the binaries is a fantastic first step. We need steps like these before development can make the transition from proprietary to Open Source. Your company is one of the few that seems to be making an honest attempt at walking this fine line.

So, if you come up with this PIM and it has Palm and IMAP support, I'll happily send you $50CDN. This is a promise from me to you.

Oh, on a related note, I was VERY excited about your release of KDEStudio. I'd also be happy to pay something for that. Sadly, I've been unable to get it to compile on my Debian Woody system running KDE 2.1 betas. If you were to statically compile it, and offer binaries with docs (thereby giving an IDE approaching MSVC++) I'd also happily pay for that.

Maybe give the source and sell the binaries?? I guess that's what Helix Code is kinda trying to do..

BTW, as an ex-Be user I simply will not pay for Binary-only software anymore. So, if BlackAdder also comes with the source, then I may be interested in buying that too..

Good luck, thanks for reviving all of these half-dead KDE projects,

Ben


By Ben Hall at Thu, 2001/01/18 - 6:00am

I'm leaving for an overnight holiday in a few minutes, so if I don't respond to things in my standard rapid fashion, I apologize.

Are people reading my mind lately? Watch for an announcement about KDE Studio on Monday or Tuesday next week.

What about BlackAdder exactly do you want to see open source? Some of it is, but we don't have control over all of it. I've gotten this question 2 or 3 times, but no one can tell me what it is they want open (I really am curious).


By Shawn Gordon at Thu, 2001/01/18 - 6:00am

Use source from CVS. I had some problems also compiling KDE Studio in Debian (Potato). Problems are fixed already in CVS.


By Hasso Tepper at Fri, 2001/01/19 - 6:00am

I've been trying (and pretty much failing) to explain to my business partners and employees (programmers and html artists, mostly) what open source is all about, and how you really can make money off of it.

I'm hoping that in six months, I can point to theKompany and other true open source success stories. Red Hat et al. were tied up into the dot com stock market bubble... open source really hasn't got some solid success stories for corporations or individuals.

--
Evan


By Evan "JabberWok... at Thu, 2001/01/18 - 6:00am

How many times must I repeat: I don't care if you charge money.

Have you not heard of the expression: "Free speech, not free beer"? Without the freedoms outlined in the Debian Free Software Guidelines (a subset of the terms demanded by the GPL), my computer, and therefore my data, is under the control of another.

I cannot accept that. If there are enough people who can (and Microsoft has proven there are), then you will make money, Shawn. If you, with your funded resources, can provide a high-quality product and good support, better than an open source project can, then you will make money, Shawn.

If, however, you assume that you cannot successfully compete with an open source project, and if you think that you cannot provide better software and support than an open source volunteer project, then that doesn't say much about your confidence in your coders and staff.


By Neil Stevens at Thu, 2001/01/18 - 6:00am

I've often heard you say things to the effect of "If you like what we're doing for the community, please buy or other products."

This actually strikes me as something that should be looked into much more in the Open Source community. As the sentiment goes around here, I have absolutely no qualms about paying for software. I just want the source. I've often thought that a donations-based model for driving Open Source development would be interesting.

After all, if people want to see quick implementations of features and new ideas, full-time developers are required. And these people do have to eat.

My guess is that if an organization such as the KDE League, or GNOME Foundation, or even the Free Software Foundation, were to setup a fund to support free software development, people would contribute. I certainly would.

It has been said that Linux is used by 2 million people in the desktop market alone. If 1 out of 5 of these people were to contribute 20$ once a year, that would be 8 million dollars to fund Open Source development. That seems to be very conservative to me - with the efforts of a large organization behind things, I bet more could be raised.

Hell, Amnesty International has no problems getting 50$/year from me. I'm sure the Open Source community, in which I have much stronger convictions, could get quite a bit more.


By James Morton at Sat, 2001/01/20 - 6:00am

It might have even been mentioned here, and I think it's a good idea. Basically a license that gives you the source code, but prevents you from distributing the code to other people. This gives you the ability to modify the code as you want, you could even submit it back if you wanted, but it also protects our interests by preventing the wholesale distribution of our code to people who didn't pay for it.


By Shawn Gordon at Sat, 2001/01/20 - 6:00am

So your solution is to discourage free software development? This is the main point of the kde project, a "free" desktop environment.
That's not to say you can't charge for your software if you wish to, but if kde developers feel like "rolling in" the good gpl'ed parts of your application into kmail they should be able to do so, and I hope they do. That's the whole point.

Frankly I'm surprised more people haven't raised this point.


By ac at Mon, 2001/01/22 - 6:00am

Well,there has to come a point where commercial software companies have to have a chance to make some money of Linux software as well, right?
This is one thing I do not understand as well: if no software companies can make money off Linux products, who will employ Linux developers?
It would be pretty sad if the majority of Linux developers had to develop for Windows from 9-5 and only do Linux at night as a hobby...sounds like a recipe for mass burnout.
If theKompany plays by the rules, bases their product on open standards and does not lock in the users into a proprietary closed solution, then what's wrong with them making some money?
Seems like the market for KDE apps has come a long way in a short time just because of a little profit-based motivation that theKompany has behind it. If they can maintain superior product quality, I would have no problem paying for their product. If their products will suck and free alternatives will be available, then they'll go out of business, but that should not be the case if they manage to provide good value...which seems to be the case here.
No one seems to have a problem with IBM making lots of $$$ thanks to Linux, selling DB2, WebSphere, hardware, etc...a little profit never hurt anybody.


By Jacek at Mon, 2001/01/22 - 6:00am

That's fine. There is absolutely nothing wrong with it. Nothing. But Shawn's original reply to the suggestion that the good parts of Aethera were incorporated into kdepim, and kmail was:
"And that will be the perfect way to make sure that we don't make money on anything and go out of business"

I don't have any problems with him trying to make a buck selling software. As a matter of fact, I would probably pay for Aethera. But to discourage developers (let's face it, he's getting pretty influential in the KDE community) from taking the parts they like about Aethera is unthinkable. If he actually succeded in doing so, I would switch to gnome (as much as I hate that prospect).


By ac at Mon, 2001/01/22 - 6:00am

Oh come on, let's not go overboard here. The point I was making was that if the only intention of the public is to wait for us to come up with a really good piece of software and then peel it off into something else, then it will most certainly do us no good, and then we will be gone and that will be that.

Think closely about what you are saying here. For the sake of argument, say we charged $50 for Aethera (it is free, but for our example), but it is open source and so all the code gets put into KMail and KPIM. Why would you buy Aethera?

Try not to over-parse my statements as I don't always have time to craft an elegant carefully worded response. I make it a point to reply to all mail and as much of these feedbacks as quickly as possible, which means my responses will not always come across correctly.

I've said time and again that I have no problem with having our stuff open sourced, but I also can't to that to the detriment of my business. I'm doing my best to respect the open source community, and I just ask for the same in return.

As far as influece in the KDE community, well I have virutally none and I don't try to exert any. All we are doing is releasing as many products as we can for KDE that plug the gaps (as I perceive them). Almost all of our employees came directly from the KDE community as well.


By Shawn Gordon at Mon, 2001/01/22 - 6:00am

Using a Caldera eServer 2.3.1 (all updates). I can't compile aethera/libs/vparsers/rdate.cpp:

g++ -I. -I. -I../.. -I/opt/kde2/include -I/usr/lib/qt2/include -I/usr/X11R6/include -I../../src -I../../data -I../../icons -O2 -Wall -c rdate.cpp
In file included from rdate.cpp:34:
rdate.h:49: warning: ANSI C++ forbids declaration `Period' with no type
rdate.h:49: parse error before `;'
rdate.h:52: type/value mismatch at argument 1 in template parameter list for `template QValueList'
rdate.h:52: expected a type, got `RDate::Period'
rdate.h:52: warning: ANSI C++ forbids declaration `periods' with no type
rdate.h:80: syntax error before `('
rdate.h:88: syntax error before `('
rdate.cpp:44: syntax error before `::'
rdate.cpp:46: syntax error before `;'
rdate.cpp: In method `void RDate::parseString(class QString)':
rdate.cpp:51: warning: comparison between signed and unsigned
rdate.cpp:60: `tb' undeclared (first use this function)
rdate.cpp:60: (Each undeclared identifier is reported only once
rdate.cpp:60: for each function it appears in.)
rdate.cpp: In method `void RDate::setFinshed()':
rdate.cpp:77: `Finished' undeclared (first use this function)
rdate.cpp: In method `int RDate::getNextOccurencePrivate(class QDateTime)':
rdate.cpp:89: request for member `count' in `RDate::periods', which is of non-aggregate type `int'
rdate.cpp:93: invalid types `int[int]' for array subscript
rdate.cpp:106: invalid types `int[int]' for array subscript
rdate.cpp:108: invalid types `int[int]' for array subscript
rdate.cpp:108: invalid types `int[int]' for array subscript
rdate.cpp:109: invalid types `int[int]' for array subscript
rdate.cpp:117: invalid types `int[int]' for array subscript
rdate.cpp: At top level:
rdate.cpp:129: syntax error before `::'
rdate.cpp:137: syntax error before `::'
make[3]: *** [rdate.o] Error 1
make[3]: Leaving directory `/home/storm/CVS/kompany/aethera/libs/vparsers'

This error is in the kompany cvs for atleast two weeks. I post this here because i don't see any bug reporting address in the web pages or source.

Is this known ?

PS: KDE2.1 from CVS. QT 2.2.3. i can compile everything else flawlessly.


By Storm at Thu, 2001/01/18 - 6:00am

Two weeks? We just made this available yesterday. I would suggest using the source RPMS, which is what I did on my COL 2.4 system.

The support address is aethera@thekompany.com, I put it in the press release, but forgot to put it on the web page, we are correcting the mistake now.

I've also forwarded your problem to the above email address. Please try the SRPM and see how that does for you, any other problems, please let us know at the above address.

Thanks!


By Shawn Gordon at Thu, 2001/01/18 - 6:00am

Hey! is there still development going on ? or is theKompany doing all New stuff , LOL !!!

go kde go


By ch at Thu, 2001/01/18 - 6:00am

Now that we all have huge hard disks, how about allowing the use of a full-blown database for storing mail, calendar, and contacts data?

Pronto (and others) use MySQL. It's riduculously fast with folders (and vfolders) containing, say, 20-30,000 messages, and rock-solid.


By Michael O'Henly at Thu, 2001/01/18 - 6:00am

This would be great as an option. That way you could easyly access all data through other means, too. Let's imagine some kind of webfrontend for contacts. I'd love this.


By RIchard Stevens at Fri, 2001/01/19 - 6:00am

This is my usual response to this:

Look at how KRN 0.6.11 does it. It's just as insanely fast with folders containing up to hundreds of thousands of messages (ok, you need 256MB of RAM for that ;-)

It uses Gigabase, an embedded DB backend, which is free, and needs no setup at all.

It's cool, it's fast, its stable, it's nice, and the author gives great support. It can even be used in commercial stuff, I think.


By Roberto Alsina at Fri, 2001/01/19 - 6:00am

i'm sorry but what's KRN? i'd like to take a look at what you refer.


By ac at Fri, 2001/01/19 - 6:00am


By Roberto Alsina at Fri, 2001/01/19 - 6:00am

ah. i've been using knode. will give this a try.
tks.


By ac at Fri, 2001/01/19 - 6:00am

Pages