Re: [Evolution] Request for additions/corrections in Evolution

On Mon, 2002-05-27 at 23:06, Jeffrey Stedfast wrote:
it does word-wrap, what do you mean?

        It does it how it wants.  The same way Word puts a paragraph where it
wants.  Also, Word has the decency to disable that.  So does Evolution. 
But doing word wrap by hand it is at least unpleasant.

        Guys.  Evolution is under the GPL right?  In this case I recomend you a
great piece of code.  It is called 'fold'.  And it comes with the
text-utils pack.  Probably it is already installed on your drive.  Check
the source and probably you can make the word wrap usable.  Than just
add the dialog with the fold options.

please see

        Very interesting indeed.  Thanks for pointing that out as I would have
never find that particular page.  But what was the purpose of that link
in this discussion?  To tell me that are the rules you implement?  In
this case I see no slider or whatever to choose that 'soft wrap
configurable column' size.

3. A new line in an unwrapped quoted text leads to two newlines instead
of one.  And I have to edit by hand the extra newline.  Anyone knows
huh? I've never seen this afaik.

        Please give me a hint of what to do to prove this?  It happens from
time to time.  And just forwarding a mail to you (no problem with the
privacy as they are in Romanian ;-) might modify the text.  As I read
below in your answer instead of printing the text you do that conversion
that of course (as pointed in the link above) is modifying the original

4. There's only one type of "quotation".  So if I I have two levels of
quotation there's no change as Evolution doesn't know about it.
the development branch is much more smart about citations, maybe it will
solve whatever your problem is?

        My "whatever" problem is:
Level three quotation
Level two quotation
Level one quotation
No quote

        IMHO it makes sense to highlight them with different colors.  That is
if you are able to code that and this isn't getting to complex.  I'd
hate to see Evolution crashing after it loads for 40 seconds... meaning
waiting another 40 seconds for a new launch and so on.  So please,
implement this
ONLY if this doesn't break something else.

        Anyway, you say there's good news and 1.2.x will have those problems

This is solved in the development branch afaik

        Good to know.  A lot better than filing some bug report.

again, solved in CVS afaik

        Same comment as above.

again, solved in cvs afaik

        And again... same as above.

        It's really good to know these things are actively solved.

just hit return 3 times? that's what I do. some people may like only 1
return when they hit return. this is just you being lazy :-)

        Me using something like Evolution means lazy.  I can just install
postfix+fetchmail+procmail+mailx+vim and I'll be over in no time with
less disk space consumed and a happy CPU that doesn't sweat every time I
hit reply for a large message (like this one for example).  Instead of
making cron jobs, scripts and launch fetchmail and postfix when I get
online I'd rather just click on Send/Receive.

        Now, this comment made a moral dillema arise.  I have no problem in
being called lazy.  But don't you think you can actually hit all your
users with this affirmation?  Come on, let's think together - postfix
and qmail are quite easy to set up and there are documents about this. 
There even is a very good, a bit outdated but else very good Mail-HOWTO
and some related HOWTOs that state in plain English (and the languages
that they were translated) how to install such a system.  There are
lists for newbies and advanced users that can help you with scripts and
configs.  Fetchmail is a 10 minutes read for the config file which is
amazingly human readable (it even has extra words to make it like human
language).  So, as long as you know some English that would be a
breeze.  Procmail is set equaly simple and the only fact that takes more
time with the setup is the amazing complexity of this tool.  Than...
lots and lots of mail clients.  Both graphical and shell/ncurses.  They
are all slim and optimised (well... not counting that thing that comes
with Netscape ;-) For calendar - we have lots of options as well... for
reading news - I had once a cute little tool (if I recall well its name
is ASnews... not sure, it was a long time ago).  That tool (and it
wasn't the only one) read the headlines and even was able to link with
the browser.  For "contacts" again... see Gnome-PIM and other packs.
Hmm... ever the newly released Open Office and Star Office have this
which integrates really well with the office pack.  Only for the task
module I don't know any equivalent as I have no use for that (cron,
anacron and at do their jobs great).  Now, correct me if I'm wrong -
anybody and I really mean anybody with some English knowledge can make
something work as Evolution with only part of that resource
consumption.  But they use Evolution.  Why?  Because of what you said
there... so think again before calling me lazy.  Who knows... maybe
people will start looking for an alternative...

you can always learn the filter expressions and use the "Expresion"

        Hmm... right now I don't have the time for this.  But in case I will,
would you (your team) be interested into a FilterHackingHOWTO for
Evolution?  Just to document things.

2. In case you don't come up with some type of grouping of the
criterions, why not make some "composite" rules.  I mean, so far if I
have two mailing lists to filter for "uninteresting" mails and for
offensive mails I'm lost.  For all simple rules "if... than" type I can
just choose "stop".  this is an excellent option as it speeds up the
process if you have many rules.  But I can't say
rule1: E1 OR E2 - action goto rule 14
rule14: E3 OR E4 - delete; stop.
this ain't gonna happen.

        Agreed.  It's dirty.  Although it is simple to set up.

3. As an alternate, I see Evolution goes for "all in one pack", but how
can I plug in say procmail for processing incoming mail taken by
Evolution.  I see it keeps the mbox...

I've added a filter rule to pipe a message to an external program. You
can use the return code of the external program to base your filters on.
For example:

[Command][/usr/bin/my-filter][is greater than][5]

        Oh.  Interesting.  I'll note that.  One silly question (and lazy too
;-): Evolution won't mind if procmail modifies its mail folders while
Evolution is active?

4. Am I missing something?  If I put 'To:' in the specific field my rule
would be ignored.  It will work only with 'To', no ':'.  Why?
because the ':' is not part of the header name?

        ;-)  Hahahaha, what happened to the user friendly label... or should I
call it idiot proof?  This is when a guy like me (the idiot in this
story) thinks that ':' is part of the field and not just a sepparator.

how do you propose Evolution monitor your internet connections? This is
not Win32 where you can just make an OS call to find out if you are
online or not. Nothing like that exists on Linux (or Unix period).

        You are not fair here.  And you just try to excuse some lame
programmers who preach Unix superiority.  For a guy that has a net link
to the net this is probably a less important issue.  For a guy like me
on dialup you can just check the ppp+ interface.  If the interface is
gone... what is Evolution trying to send?!?  And how?

I think the global Stop button in the toolbar should do the trick?

        Oh!  Thanks.  I really didn't notice that button.  Silly me!  Please
accept my hones appologies for that stupid mistake.  I looked for it
after reading your message and that was the only button that had no
meaning (it is disabled unless there is some traffinc going on). 
Problem solved.


1. it is assumed that you are online


2. if the send fails, it will popup the composer window and let you
decide what to do with it

        right. but in my sittuation, not in all cases (hmm... 10 minutes I
think was pretty much enough for a timeout message)  but I can't
reproduce this issue. :(

3. if we save to Outbox temporarily, it's theoretically possible that in
the interim, your auto-send&receive thread will wake up and send all the
messages in Outbox before the composer's Send could send that message,
thus you might end up sending the message twice.

        you are right. it makes sense.

4. And most importantly, at the time we had no way of finding out the
UID of a message after appending it to a folder (we are still not
guarenteed to be able to get the UID of the message we just appended to
a folder, seeing as how IMAP doesn't guarentee it unless it supports the
UIDPLUS extension and implements it *fully* - which unfortunately, some
do not) and so after sending the message - how would we know which
message to delete from the Outbox and move it to the Sent folder?

        I wasnt' aware of that.  But the way you put it is makes perfect sense.

        Bottom line - Thank you Jeffrey for sheding some light over this issue
as well.

you can't. well, you could just rm -f the executables and then they
wouldn't load :-)

        Hahaha... wouldn't be nice an "Startup" tab in options that says what
to load and what not to load?  Anyway, I like your hack.  Quick and
dirty but it should do the job.

On Mon, 2002-05-27 at 23:52, Jeffrey Stedfast wrote:
My fist complaint (or did I miss something?) is that I want to
another editor than the default.  How can this be done?
It can't.
because no one has implemented it?

        Hmm... what about the gnome-vim project?  The author states there (on
his pages) something about some pathes that have to show up in
Evolution.  Does anybody on this list know anything?

then use gvim to compose a message body and paste it into the editor.

        Hmm... remeber?  I'm lazy.  I could go back to my personal favourite
pine (it has its problems but I have used it for years) which knows how
to use an external editor.  But this is not the point.  We're on the
Evolution list debating how to make this app better.

yes, there is a hack around somewhere. do a search on google.

        Easier - go to to the news section on that first
page than there you'll find a link.

The composer is designed to not auto-wrap quoted paragraphs because
there may be a need for them to not be wrapped. Anyways, there is
already a feature request about this afaik.

        Yes. I agree.  But this should (IMHO) be optional - controlable by a
check box.

what number? the number of characters per line? You can't.

        Yes.  That number.  Why not?

    My text is plain text.  My friend's text is also plain text.  Do
mean Evolution does some sort of conversion behind the scenes
yes. GtkHTML is an html editor.

        GOD! I rest my case.

Once you hit return to modify the quoted text, it's no longer our
problem. You've modified it, you get to deal with it.

        I beg your pardon?

I guess this must mean that the Linux kernel must be so bug free that
is 2.4 * no-bugs, huh? Give me a break.

        Hmm... you know, making a driver can hide bugs that arise in certain
moments.  same goes for the networking part where there might be a
design flaw.  Fixing the coloring and giving some more options... this
IMHO isn't something that can't be tested.  Don't you guys check if what
you wrote is working?  Or you just read the compiler output?

No one else seems to have complained.

        Right.  So if less than X people complain about something than the
issue has no value whatsoever to you or your team.  This looks more like
the scripts I make for my own personal use - I don't complain so they
must be perfect. ;-) Cute.

remember that GtkHTML is not a plain-text editor, it is an HTML
Which means that you must convert plain-text back into HTML. This is a
non-perfect translation. There's no way to make it perfect, which is
we now save drafts in HTML always (even if you choose to only send
plain-text mail) this way formatting is preserved.

        Why not support (not create a new one as there are already plenty) a
real editor?  One that goes faster than WordXP and that has no problem
taking "plain text".

It works fine enough to be in the stable release? Like I said above,
one else seems to mind. You're the only one who has complined about

        You're cute with these religious-like remarks.

Why don't you send us a patch? Then we can all be happy :-)

        I will as soon as I have the time to read/understand the internals of
your app.  But beware I'll complain noisily in case you drop my patch
for reasons like "not complaining with our pretty printing politicy" or
things like that.  Do we have a deal?

    Why?  Even pine which uses ncurses knows this stuff.

Huh? what does this got to do with anything? The quoting colours are
sent via email, each program has to identify what the quoted
are and colouise them itself. Pine's quoting has nothing to do with
Evolution's quoting and vise versa.

        Evolution is about sending mails according to the RFC.  But also to
help me.  I think I'm saying this already too often - nobody would want
to waste time with such a beast if it would be only for the receiver's
pleasure.  Actually what's your point with this answers?  Probably you
never even heard about it, but anyway, the mail app plus ed used to do
that.  People don't use them anymore?  Why?  Because it sent bad mails? 
Not at all.  Because their developers were having the same mentality as
yourself (nobody complains so this is fine)?  Nope.  Because people
wanted more usability.

        Please take your time and answer me honestly: is Evolution just another
Outlook wannabie clone which is trying to reproduce even the badly
optimised routines from the above mentioned app?  Because if the answer
is yes, this would save both me and you a lot of time.  In that case I
would really couldn't care less and if I ever become so desperate I'll
support the Wine group to make Outlook work on Linux too. I'm having
this conversation hoping for an improvement and not for a blindly follow
of the Outlook, Word or Netscape.  Else my friend you're fighting for a
lost cause as Outlook will always be better than any clone just for the
simple fact that it is the original.

Thank you for throwing insults at us left and right. That's always a
sure way to win people over and make them do things for you.

        Oh.  Sorry.  I started with the wrong impression.  I have to win you
over my side instead of showing some improvements that you can take into
consideration or just put them on a TODO list without a deadline.  Silly
me, I thought your purpose was to make a better product.  I judged you
wrong and please accept my appologies.  You probably joined this project
to become famous and to get a nick like Mr. JS the GrEaT or something
like that.

Instead of insulting us, why not start writing code and helping out?
seem to think you could do better. "Put your code where your mouth is"
as they say.

        I didn't call you names.  Please accept my appologies.  English isn't
my native language and my command of it is below any native speaker. 
Besides, I have a job.  And my life is kinda full.  Maybe I'm wrong, but
don't you get paid at Ximian?  If yes, I will gladly dump my job and
even gladlier would accept a job with Ximian because it surely pays
better.  Besides, I get that mailbox which is cool.  So tell
me: how do you Jeff Stedfast earn your living?

Just make fetchmail download your mail and use procmail as normal. No
need to have Evolution pipe it to anything.

        No need for anybody to use this.  Just I'll have to make a python
options dialogue and probalbly one of those nice wizards for setting up
the whole pack.  My solution would have mysql for backing up huge mail
folders ;-)  Ain't that cool?  Actually... and I say this again, I like
Evolution.  But I think it has a lot way ahead before becoming as usable
as the site says.

read the source code and/or the filtertypes.xml file in ~/evolution

        Thank you for pointing this out.  Any help like this might be of use. 
This is one reason why I don't say "okay, you'll have that patch by
Monday".  Because unlike you, I have to dig and hack for understanding
what's going on.

No, it means that the expressions may change some day. They won't
in any 1.0.x release. They have already been extended in the 1.1.x
development branch (but older rules will still work).

        You mean 1.2.x will use other DTD?

Yep, you'll have to reinstall fetchmail. I don't see why you'd need to
install Sendmail though.

        To send my mails and get around that "hanging" problem. ;-)  Ever tried
CronosII?  Please do - you'll find it with google.  Anybody in his/her
right mind will use something like that if he/she has to set up 
fetchmail and procmail.. and eventually sendmail.

The linux kernel hackers maybe? :-)

        Kernel hackers?!?  I write '/sbin/ifconfig ppp0' and I see if the line
is active.  What do you say abut this kernel hacking?  If this can be
considered kernel hacking than I'l greater even than Linus.

We can't guarentee that the patch will make it in, but if your patch
well written and clean (ie, no kludgy hacks) then odds are we'd accept

        Be smart Jeff!  If there's no other patch accept it and label it with a
huge note that says "anybody that can rewrite that piece of code in a
clean manner getts a candy from me".  As long as it is working, put it
there, and later on find somebody who can do it better.  It's a lot
nicer than having bugs.  And as long as you label the "unclean" code as
such, your overall impression will be "good".

        For my friend Zed: you are right.  But I don't see why you are mad at
me?  My intial mail wasn't for YOU but for a mailing list inhabited by
some of the Evolution developers as well.

        One last question: I see both Jeff and Zed didn't bother to cut the
parts that they didn't answer.  My mails being more than two
paragraphs... that means a lot of bandwidth waste for all of the
subscribers.  Are they always acting like this?  Or was it just a way to
impress people with the lenght of their answers?  My overall impression
- lack of netiquette.

Attachment: signature.asc
Description: This is a digitally signed message part

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]