Work around for the "start up bug"

["Sergey I. Panov" <sipan mit edu>]

I have realized that Subject was misleading. So, I am reposting my
message under different subject. New and important part is that there
is a way around "start up bug" which affects panel, gmc,
gnome-help-browser, gnomecc, etc.  An important part is a work
around. I noticed that if you kill application that get stuck, you
still have "gnome-name-service" running and as long as it is there
you can start you application second time and it will run.

E.g. the following start up sequence should work:

 gnome-name-service &
 panel &
 killall gen_util_applet
 gmc &
 gnome-help-browser &
 

Here is my original message:

> Sorry, I mailed an empty  mesage by mistake.
> Somehow it escaped my computer, even though I decided to to mail anything :)
> 
> I think the problem I was about to report is the same as one reported
> by Peter Wainwright . I can add backtraces, but doubt now it will add
> any incite (I did get panel and gnomecc backtraces identical to those
> from Peter Wainwright).
>   
> The only difference is that, if you leave gnome-name-service running
> after first unsuccesfull start, you can start gmc, gnome-help-browser and
> gnomecc, but not panel -- it just shows up as a gray strip. Killing 
> "gen_util_applet --activate-goad-server gen_util_applet" process
> revives panel as well.
> 
> Well, just in case you'll find it usefull. Here are backtraces for  
> gmc:
> 
> #0  0x404616f4 in __read ()
> #1  0x4048f6cc in __DTOR_END__ ()
> #2  0x4043356c in _IO_file_underflow (fp=0x81618a8) at fileops.c:303
> #3  0x40434148 in __underflow (fp=0x81618a8) at genops.c:249
> #4  0x40431723 in _IO_getline (fp=0x81618a8, buf=0xbfffef78 "", n=2047, delim=10, extract_delim=1)
>     at iogetline.c:47
> #5  0x40430f11 in _IO_fgets (buf=0xbfffef78 "", n=2048, fp=0x81618a8) at iofgets.c:47
> #6  0x40017ec5 in name_server_by_forking (ev=0xbffff798) at orbitns.c:150
> #7  0x400180f3 in gnome_name_service_get () at orbitns.c:242
> #8  0x8065e9c in try_to_activate_running_copy () at gcorba.c:205
> #9  0x806546e in handle_args (argc=1, argv=0xbffffc6c) at main.c:2697
> #10 0x806595b in main (argc=1, argv=0xbffffc6c) at main.c:2897
> 
> and gnome-help-browser:
> 
> #0  0x404f36f4 in __read ()
> #1  0x405216cc in __DTOR_END__ ()
> #2  0x404c556c in _IO_file_underflow (fp=0x812e7b0) at fileops.c:303
> #3  0x404c6148 in __underflow (fp=0x812e7b0) at genops.c:249
> #4  0x404c3723 in _IO_getline (fp=0x812e7b0, buf=0xbfffd24c "  Q@\020I6@8 Q@#", n=2047, delim=10, 
>     extract_delim=1) at iogetline.c:47
> #5  0x404c2f11 in _IO_fgets (buf=0xbfffd24c "  Q@\020I6@8 Q@#", n=2048, fp=0x812e7b0) at iofgets.c:47
> #6  0x40017ec5 in name_server_by_forking (ev=0xbfffda6c) at orbitns.c:150
> #7  0x400180f3 in gnome_name_service_get () at orbitns.c:242
> #8  0x400173db in goad_server_register (name_server=0x0, server=0x812e458, 
>     name=0x8056031 "gnome-help-browser", kind=0x805602a "object", ev=0xbfffdc20) at goad.c:1066
> #9  0x804e0fb in main (argc=1, argv=0xbffffc4c) at gnome-help-browser.c:234
> 
> 
> ******************************************************************
> * FOR THOSE WHO HAVE THIS PROBLEM, BUT STILL WANT TO RUN   GNOME *
> ******************************************************************
> The following start up sequence works (consider it a hack :) )
> 
> gnome-name-service &
> panel &
> killall gen_util_applet
> gmc &
> gnome-help-browser &
> 


                                  Sergey Panov