Trejkaz Xaoza | 1 Aug 05:29 2004
Picon

Re: Re: Re: Flyspray news

On Sat, 31 Jul 2004 17:53, Remko Troncon wrote:
> > Using cvs2svn implies that we want to migrate.  The authors say that it
> > isn't particularly good for continuous use
>
> What isn't particularly good for continuous use ?

cvs2svn.

> > What would be really good though, would be a mod_svn interface to CVS
> > itself. :-/
>
> That will never happen. Besides, i don't see why you would want this ?

The reason I would want this is because it would give me a WebDAV+SVN view of 
a CVS repository.  You could then use the Subversion client to access a CVS 
repository, so all users could just use one client, instead of having to use 
multiple clients.

I would have thought people on this mailing list of all places, would have 
this ideal.  Or maybe you wish to go back to using 5 IM clients. ;-)

TX

--

-- 
'Every sufficiently advanced technology is indistinguishable from magic' - 
Arthur C Clarke
'Every sufficiently advanced magic is indistinguishable from technology' - Tom 
Graves

             Email: Trejkaz Xaoza <trejkaz <at> xaoza.net>
(Continue reading)

Trejkaz Xaoza | 1 Aug 05:38 2004
Picon

Re: Re: Default profile

On Sat, 31 Jul 2004 21:22, Enleth wrote:
> TX> I don't think that's necessarily an issue.  But on the other hand, if a
> user TX> can understand what an email address is, they can understand what
> a JID is TX> since it's exactly the same difficulty. :-)
>
> Well, I'm afraid there are people using email for years and still
> unable to undestand what it is.

And yet, mail clients still make you enter the full email address.

Even Hotmail requires you to enter your full email address to login these 
days.

But since you seem to be talking about web registration for email accounts, 
how does Psi even enter into it?  The web registration for a Jabber account 
could, and does, work this way.  The Australian server jabber.org.au is a 
good example.  You get accounts on the Jabber server by registering on the 
forums.  The forums make you enter a username to sign up, but give you a 
Jabber address for use with clients.

> The issue is very simple - when you register an email account, you do
> it with web-based interface, you could know nothing about email. In
> Jabber, you register an account trough the client, so you need to know
> the server name and it's relation with final address (JID).
> Then, some kind of explanation and/or wizard would be handy.

I have _always_ said that this should be kept to a wizard.  This is why I'm so 
against it being anywhere in the normal interface. :-)

> BTW, is querying the server for services without an account possible?
(Continue reading)

Remko Troncon | 1 Aug 09:57 2004
Picon
Picon

Re: Re: Default profile

> And remember that the interface for in-client registration is actually 
> deprecated now.  That means that given enough time, in-client registration 
> won't exist anymore and we will all be using pure external registration like 
> web sites.

Oh, that's solved then! Why spend more time and code on the topic if it will be
obsolete soon anyway.

cheers,
Remko
Remko Troncon | 1 Aug 10:03 2004
Picon
Picon

Re: Re: Re: Flyspray news

> cvs2svn.

Of course not, you only use it once. Why keep CVS if you have SVN ?

> I would have thought people on this mailing list of all places, would have 
> this ideal.  Or maybe you wish to go back to using 5 IM clients. ;-)

What ideal ? I really fail to see the point of having 2 access methods
for 1 versioned repository. Either you use CVS, or either you use SVN.

And i also fail to see the link between IM clients. Having 2
command-line tools which you sometimes use (such as CVS and SVN)
installed side-by-side, and two applications you constantly have running
side-by-side (such as 2 browsers or 2 IM clients) are two completely
different things.

cheers,
Remko
Trejkaz Xaoza | 1 Aug 14:07 2004
Picon

Re: Re: Re: Default profile

On Sun, 1 Aug 2004 17:57, Remko Troncon wrote:
> > And remember that the interface for in-client registration is actually
> > deprecated now.  That means that given enough time, in-client
> > registration won't exist anymore and we will all be using pure external
> > registration like web sites.
>
> Oh, that's solved then! Why spend more time and code on the topic if it
> will be obsolete soon anyway.

I _hope_ it's soon, but that's probably being a bit optimistic. ;-)

TX

--

-- 
'Every sufficiently advanced technology is indistinguishable from magic' - 
Arthur C Clarke
'Every sufficiently advanced magic is indistinguishable from technology' - Tom 
Graves

             Email: Trejkaz Xaoza <trejkaz@...>
          Web site: http://xaoza.net/trejkaz/
         Jabber ID: trejkaz@...
   GPG Fingerprint: 9EEB 97D7 8F7B 7977 F39F  A62C B8C7 BC8B 037E EA73
Trejkaz Xaoza | 1 Aug 14:13 2004
Picon

Re: Re: Re: Re: Flyspray news

On Sun, 1 Aug 2004 18:03, Remko Troncon wrote:
> > I would have thought people on this mailing list of all places, would
> > have this ideal.  Or maybe you wish to go back to using 5 IM clients. ;-)
>
> What ideal ? I really fail to see the point of having 2 access methods
> for 1 versioned repository. Either you use CVS, or either you use SVN.

Actually, doing what I was talking about would have resulted in using _one_ 
access method for _all_ repositories.  Right now, I have to use _two_ access 
methods, CVS and SVN, because projects randomly choose which to use. :-)

> And i also fail to see the link between IM clients. Having 2
> command-line tools which you sometimes use (such as CVS and SVN)
> installed side-by-side, and two applications you constantly have running
> side-by-side (such as 2 browsers or 2 IM clients) are two completely
> different things.

Suppose I only need to talk to Yahoo users a few times a day.  Does that make 
it acceptable to have to run their client alongside my other working client?

Or suppose I need to use CVS and SVN 2,000 times a day.  Is it still ideal to 
have to remember which site is using which repository?  I would much rather 
type one command and have it work in all places. :-)

TX

--

-- 
'Every sufficiently advanced technology is indistinguishable from magic' - 
Arthur C Clarke
'Every sufficiently advanced magic is indistinguishable from technology' - Tom 
(Continue reading)

Remko Troncon | 1 Aug 18:25 2004
Picon
Picon

Re: Re: Re: Re: Flyspray news

> Right now, I have to use _two_ access 
> methods, CVS and SVN, because projects randomly choose which to use. :-)

And that is the way it should be: freedom of choice. And if you made a
choice, why provide another way of access (which will be limited
anyway, because of non-compatible features).

> Suppose I only need to talk to Yahoo users a few times a day.  Does that make 
> it acceptable to have to run their client alongside my other working client?

No.

> Or suppose I need to use CVS and SVN 2,000 times a day.  Is it still ideal to 
> have to remember which site is using which repository?  

Yes, i see no problem with that (even though it is a hypothetical situation
that will never happen). 

I can't make my opinion more clear than i did: there's a world of
difference between having to have 2 applications open all the time, and
having to use 2 different command-line tools. 

> I would much rather type one command and have it work in all places.

Good luck.

cheers,
Remko
Jonathan | 1 Aug 19:44 2004
Picon

Psi core dumping even after reinstalling on FreeBSD current

Every time I start Psi on my server I get the error

server% psi
Segmentation fault
server%

I have a core file available but don't know what to do with it.  I can 
put the core file on my site for download if someone wants to grab it to 
look at or I can follow any directions needed.  Vim and gnucash both 
start and work fine so I'm pretty sure it's not an X problem.  The only 
thing I have been able to think of that may be the cause is I am trying 
to run it on a Pentium 233MMX (an original Pentium not P2).

Thanks,
Jonathan
Michail Pishchagin | 1 Aug 21:36 2004

Re: Psi core dumping even after reinstalling on FreeBSD current


Jonathan wrote:
> I have a core file available but don't know what to do with it.  I can
> put the core file on my site for download if someone wants to grab it to
> look at or I can follow any directions needed.  Vim and gnucash both
> start and work fine so I'm pretty sure it's not an X problem.  The only
> thing I have been able to think of that may be the cause is I am trying
> to run it on a Pentium 233MMX (an original Pentium not P2).

Try editing psi/src/src.pro, and make sure that the "CONFIG += debug" is not 
commented out. If so, uncomment the line, and type "make distclean && make".

Now you should have executable with debugging symbols.

Run "gdb ./psi". When it'll segfault, type "bt" and provide the output of that 
command. Thanks.

--

-- 
    Michail "mblsha" Pishchagin     http://maz.sf.net
mailto:mblsha@... xmpp:mblsha@...
Ian Holsman | 2 Aug 01:20 2004
Picon

0.92 is Crashing on OS/X

The backtrace didn't show anything interesting.
it seemed to happen when I logged on to a yahoo gateway..

Qt: internal: bogus 162

Program received signal EXC_BAD_ACCESS, Could not access memory.
0x91bd8fb4 in NewCallBack_priv ()
(gdb) bt
#0  0x91bd8fb4 in NewCallBack_priv ()
#1  0xef68a7fc in QAuServerMac::play(QString const&) ()
#2  0x000f2d18 in ?? ()
#3  0x000efa20 in ?? ()
#4  0x002047e4 in ?? ()
#5  0xef714760 in QObject::activate_signal(QC

Gmane