1 |
Dale schreef: |
2 |
> Holly Bostick wrote: |
3 |
> |
4 |
> |
5 |
>> Did you re-emerge gentoo-sources after removing the "doc" USE flag? |
6 |
>> |
7 |
>> |
8 |
>> |
9 |
>> |
10 |
>> |
11 |
>> |
12 |
>> |
13 |
>> |
14 |
>> |
15 |
>> |
16 |
>> |
17 |
>> |
18 |
>> |
19 |
>> |
20 |
> |
21 |
> Yup, I sure did. |
22 |
> |
23 |
>> What is the format of the relevant entry in |
24 |
>> /etc/portage/package.use? |
25 |
>> |
26 |
>> If it does not look like this |
27 |
>> |
28 |
>> sys-kernel/gentoo-sources -doc |
29 |
>> |
30 |
> |
31 |
> Mine looks like this: O_O |
32 |
> |
33 |
> |
34 |
>> sys-kernel/gentoo-sources -doc |
35 |
|
36 |
OK, that eliminates that possibility; it must be something else, then. |
37 |
> |
38 |
>> Also, try |
39 |
>> |
40 |
>> emerge -uDNptv world |
41 |
>> |
42 |
>> emerge --update --deep --newuse --pretend --tree --verbose |
43 |
>> |
44 |
>> (with --tree being the important change) |
45 |
>> |
46 |
>> to see what packages are requiring xmlto. We're just guessing that |
47 |
>> it's gentoo-sources, really; maybe it's not. |
48 |
>> |
49 |
> |
50 |
> |
51 |
>> root@smoker / # emerge -uDNptv world |
52 |
>> |
53 |
>> These are the packages that I would merge, in reverse order: |
54 |
>> |
55 |
>> Calculating world dependencies ...done! [nomerge ] |
56 |
>> sys-kernel/vanilla-sources-2.6.12.5 -build +doc -symlink [ebuild N |
57 |
>> ] app-text/xmlto-0.0.18 0 kB |
58 |
>> |
59 |
>> Total size of downloads: 0 kB root@smoker / # |
60 |
> |
61 |
|
62 |
And so, in fact, it is something else. Or something more, that we didn't |
63 |
know about. |
64 |
|
65 |
Do you use/need vanilla-sources? If not, then you might consider |
66 |
unmerging it, so it does not appear in your world file and attempt to |
67 |
update every time you do an emerge -whatever world. |
68 |
|
69 |
And you might also consider adding -doc to /etc/make.conf, as noted |
70 |
previously. |
71 |
|
72 |
> |
73 |
> Something new that didn't show up last time. My new package.use: |
74 |
> |
75 |
> |
76 |
>> sys-kernel/gentoo-sources -doc sys-kernel/vanilla-sources -doc |
77 |
>> |
78 |
|
79 |
That should solve that, then. |
80 |
|
81 |
>> Great. No more need to deal with that atm, then. |
82 |
> |
83 |
> What about the broke stuff? |
84 |
|
85 |
According to your last output: |
86 |
|
87 |
>> root@smoker / # revdep-rebuild |
88 |
>> |
89 |
>> Checking reverse dependencies... Packages containing binaries and |
90 |
>> libraries broken by any package update, will be recompiled. |
91 |
>> |
92 |
>> Collecting system binaries and libraries... done. |
93 |
>> (/root/.revdep-rebuild.1_files) |
94 |
>> |
95 |
>> Collecting complete LD_LIBRARY_PATH... done. |
96 |
>> (/root/.revdep-rebuild.2_ldpath) |
97 |
>> |
98 |
>> Checking dynamic linking consistency... broken /usr/lib/gaim/tcl.so |
99 |
>> (requires libtcl8.3.so libtk8.3.so) broken |
100 |
>> /usr/lib/libgtkmm-2.0.so.1.5.9 (requires libsigc-1.2.so.5) broken |
101 |
>> /usr/lib/libgdkmm-2.0.so.1.5.9 (requires libsigc-1.2.so.5) broken |
102 |
>> /usr/lib/libpangomm-1.0.so.1.5.9 (requires libsigc-1.2.so.5) broken |
103 |
>> /usr/lib/libglibmm-2.0.so.1.5.9 (requires libsigc-1.2.so.5) broken |
104 |
>> /usr/lib/libatkmm-1.0.so.1.5.9 (requires libsigc-1.2.so.5) broken |
105 |
>> /usr/lib/python2.2/lib-dynload/_tkinter.so (requires libtk8.3.so |
106 |
>> libtcl8.3.so) broken |
107 |
>> /usr/lib/libgtkmm_generate_extra_defs-2.0.so.1.5.9 (requires |
108 |
>> libsigc-1.2.so.5) broken /usr/bin/icewm-session (requires |
109 |
>> libungif.so.4) broken /usr/bin/icewm (requires libungif.so.4) |
110 |
>> broken /usr/bin/gnome-panel-properties-capplet (requires |
111 |
>> libcapplet.so.0) broken /usr/bin/icewmtray (requires libungif.so.4) |
112 |
>> broken /usr/bin/icehelp (requires libungif.so.4) broken |
113 |
>> /usr/bin/icewmbg (requires libungif.so.4) broken |
114 |
>> /usr/X11R6/lib/gaim/tcl.so (requires libtcl8.3.so libtk8.3.so) |
115 |
>> broken /usr/X11R6/lib/libgtkmm-2.0.so.1.5.9 (requires |
116 |
>> libsigc-1.2.so.5) broken /usr/X11R6/lib/libgdkmm-2.0.so.1.5.9 |
117 |
>> (requires libsigc-1.2.so.5) broken |
118 |
>> /usr/X11R6/lib/libpangomm-1.0.so.1.5.9 (requires libsigc-1.2.so.5) |
119 |
>> broken /usr/X11R6/lib/libglibmm-2.0.so.1.5.9 (requires |
120 |
>> libsigc-1.2.so.5) broken /usr/X11R6/lib/libatkmm-1.0.so.1.5.9 |
121 |
>> (requires libsigc-1.2.so.5) broken |
122 |
>> /usr/X11R6/lib/python2.2/lib-dynload/_tkinter.so (requires |
123 |
>> libtk8.3.so libtcl8.3.so) broken |
124 |
>> /usr/X11R6/lib/libgtkmm_generate_extra_defs-2.0.so.1.5.9 (requires |
125 |
>> libsigc-1.2.so.5) broken /usr/X11R6/bin/icewm-session (requires |
126 |
>> libungif.so.4) broken /usr/X11R6/bin/icewm (requires libungif.so.4) |
127 |
>> broken /usr/X11R6/bin/gnome-panel-properties-capplet (requires |
128 |
>> libcapplet.so.0) broken /usr/X11R6/bin/icewmtray (requires |
129 |
>> libungif.so.4) broken /usr/X11R6/bin/icehelp (requires |
130 |
>> libungif.so.4) broken /usr/X11R6/bin/icewmbg (requires |
131 |
>> libungif.so.4) done. (/root/.revdep-rebuild.3_rebuild) |
132 |
>> |
133 |
>> Assigning files to ebuilds... done. |
134 |
>> (/root/.revdep-rebuild.4_ebuilds) |
135 |
>> |
136 |
>> Evaluating package order... done. (/root/.revdep-rebuild.5_order) |
137 |
>> |
138 |
>> Dynamic linking on your system is consistent... All done. |
139 |
|
140 |
"Dynamic linking on your system is consistent... All done." means that |
141 |
nothing needs rebuilding, in the opinion of revdep-rebuild (meaning no |
142 |
libraries are disconnected from the programs that depend on them). |
143 |
|
144 |
By the way, what version of gentoolkit do you have installed? If the |
145 |
last stable (0.2.0-r2), you might very well want to consider unmasking |
146 |
the unstable version for this package only -- add |
147 |
"app-portage/gentoolkit ~x86" to /etc/portage/package.keywords-- as |
148 |
revdep-rebuild is vastly improved (though still not perfect) in the most |
149 |
recent unstable version. |
150 |
|
151 |
It is within the realm of possiblility that your version of |
152 |
revdep-rebuild is less trustworthy than mine (I use the unstable |
153 |
version), so the reason why you're receiving untrustworthy reports, and |
154 |
I'm saying the application is trustworthy despite this is because my |
155 |
version *is* reasonably trustworthy and yours is not (which would make |
156 |
all of the following completely incorrect). |
157 |
|
158 |
I understand that all those "brokens" seem like they should be of |
159 |
concern, but the important thing is that *revdep-rebuild is not offering |
160 |
to rebuild anything*. What I suspect has happened is that you have |
161 |
changed your USE flags and possibly run revdep-rebuild based on a |
162 |
revdep-rebuild -p that you had done prior to changing your USE flags. |
163 |
|
164 |
I say this because I notice that gaim and python both have the tcltk USE |
165 |
flag available; they must have been compiled at least once with this |
166 |
flag active for them to require libtk8.3.so and libtcl8.3.so, but now, |
167 |
even though these libraries are broken, gaim and python are not offered |
168 |
to be rebuilt-- therefore the optional dependency on these libraries |
169 |
must have been removed in the meantime. This could occur if you had run |
170 |
revdep-rebuild -p (which makes a list of the proposed rebuilds based on |
171 |
the system at that moment), changed your USE flages, removing "tcltk", |
172 |
then re-emerged gaim and python during the course of a emerge --newuse |
173 |
world, and then run revdep-rebuild without the -p (which reads the list |
174 |
created from the previous --pretend run rather than re-evaluating the |
175 |
system). |
176 |
|
177 |
All the broken gnome libs depending on libsigc++ (which is where |
178 |
libsigc-1.2.so.5 comes from) seem to be deep dependencies (no packages |
179 |
are offered to be fixed related to them); since you appear to be a KDE |
180 |
user, the only way I see for these libraries to be on your system is |
181 |
that they were optional dependencies emerged via the +gnome USE flag |
182 |
(since the direct dependencies for these libraries are not being |
183 |
mentioned as broken, or offered to rebuild). Perhaps this USE flag has |
184 |
also been disabled and the applications previously using it rebuilt. |
185 |
|
186 |
The icewm applications depending on libungif are also not offered to |
187 |
rebuild, although they are broken w.r.t the libraries, which suggests |
188 |
that icewm is not in your world file, but is a dependency or deep |
189 |
dependency of something that is. |
190 |
|
191 |
But what's interesting is that icewm does not depend on libungif, but |
192 |
giflib, which is apparently not broken, and certainly not broken w.r.t |
193 |
icewm. |
194 |
|
195 |
So what it looks like to me is that your system is very sloppy at the |
196 |
moment, but nothing in fact is broken by some miracle of fate, which is |
197 |
why revdep-rebuild put out all this output, but offered to fix nothing. |
198 |
|
199 |
Or it's a gigantic bug. but if that was the case, you would likely have |
200 |
noticed; you would have rather a wide range of programs and DE/WMs that |
201 |
would not start, such as GNOME, IceWM, and Gaim, which I think you would |
202 |
have reported if it had happened. |
203 |
|
204 |
So I think that revdep-rebuild is telling the truth, and although many |
205 |
things are bent, nothing is broken and nothing in fact needs rebuilding. |
206 |
|
207 |
I would consider an |
208 |
|
209 |
emerge depclean -p (do *not* forget the "-p"!!!) |
210 |
|
211 |
to begin the process of getting rid of some of these apparently |
212 |
no-longer needed packages and libraries. |
213 |
|
214 |
*Absolutely* look at the list and make sure nothing essential is |
215 |
suggested to unmerge before attempting an emerge depclean without |
216 |
--pretend!!!! |
217 |
|
218 |
> |
219 |
> |
220 |
>> *Will* you stop trying to get authorization for emerge -e at every |
221 |
>> opportunity!!!??? :-) |
222 |
>> |
223 |
> |
224 |
> Well, that was the command I was given and copy and paste works. I'm |
225 |
> a bad typer remember. I copy and paste all I can. It's safer. |
226 |
|
227 |
|
228 |
That's the command you were given *when*? About the only time you are |
229 |
"given" that command is when you first install the system. |
230 |
|
231 |
Emerge --emptytree rebuilds every package in the system, as if nothing |
232 |
was installed at all (as if the "tree" was "empty"). Now I don't know |
233 |
how long it took you to emerge your system, but personally, I have no |
234 |
particular desire to emerge gcc and glibc and X and KDE and Mozilla and |
235 |
every single other package on my system all over again, since each one |
236 |
of those named packages takes at least two hours (some significantly |
237 |
more, like glibc and X, not to mention something like KDE) to emerge by |
238 |
itself, so emerging everything would probably take closer to 24 to 36 |
239 |
hours than like....10 to 18 (which is still a da*m long time, btw) . And |
240 |
that's assuming everything went perfectly, which it often doesn't. |
241 |
|
242 |
The point is, emerge -e "should" only be used in the very most extreme |
243 |
of emergencies; it's a pretty desperate measure, not to be taken |
244 |
lightly, and very very rarely necessary. |
245 |
|
246 |
> |
247 |
>> It's really not necessary. And you're getting yourself all worked |
248 |
>> up over a relatively minor issue (or in fact a couple of them). |
249 |
>> |
250 |
> |
251 |
> I'm not all worked up here. I'm ROTFLMAO though. LOL I'm OK as |
252 |
> long as I can figure it out OR get help fixing it. |
253 |
|
254 |
I'm not clear what, at this point, you're trying to fix. The original |
255 |
issue, revdep-rebuild, is solved, insofar as revdep-rebuild has |
256 |
completed with no re-emerges necessary, and the xmlto thing should now |
257 |
be solved as well. |
258 |
|
259 |
If something of substance remains, perhaps you should start another thread. |
260 |
|
261 |
If you're just upset about the revdep-rebuild output with no offers to |
262 |
re-emerge, I wouldn't worry about that (in the sense of thinking there's |
263 |
a 'problem' as opposed to an annoyance). |
264 |
|
265 |
Revdep-rebuild may not be perfect, but if something was actively broken, |
266 |
it would catch that. I would trust that if it says nothing needs to be |
267 |
re-emerged, then nothing needs to be re-emerged. |
268 |
|
269 |
> |
270 |
> |
271 |
>> Basically, you seem to be upset because Portage is having a fit |
272 |
>> when you try to update world. Not because a program is broken, or |
273 |
>> because you can't do some specific task (because a program is |
274 |
>> broken). If that is a correct assessment of the situation, then |
275 |
>> have some perspective. |
276 |
> |
277 |
> |
278 |
> I'm not mad at portage. I love portage. I still remember Mandrake. |
279 |
> I'll never forget that mess. At least with this, it can be fixed |
280 |
> without a re-install. |
281 |
|
282 |
No, I think you're missing my point. My point was that if the only |
283 |
problem with your system is that you have some anomalies when you do an |
284 |
emerge -whatever world or revdep-rebuild, but all the applications you |
285 |
use work, and you can in fact use your system for what you use it for, |
286 |
you don't have a problem, /per se/, you have a minor annoyance. If you |
287 |
didn't update world (which is not necessary), you would never notice a thing |
288 |
(presumably, correct me if I'm wrong). |
289 |
|
290 |
I'm certainly not saying that we don't fix minor annoyances here (we |
291 |
most assuredly do), but there's no need to treat it like The End of The |
292 |
World (which is about the only case in which you might care to |
293 |
reinstall/emerge -e world). |
294 |
|
295 |
Portage constantly up/downgrading giflib/libungif, or even refusing to |
296 |
install one of them, is far from such a state, unless the issue |
297 |
breaks.... (I dunno) Evolution, which you need to communicate with your |
298 |
clients at work, for example. Being unable to use Evolution if you need |
299 |
it is a problem. Portage being annoying is... just annoying. |
300 |
|
301 |
> |
302 |
> |
303 |
>> You don't have to update world every day, or even every month. So |
304 |
>> don't. If things work OK for what you need them to do, then the |
305 |
>> fact that you can't update easily right now is *not a problem*. |
306 |
>> Certainly not one needing a reinstall of the entire system. |
307 |
>> |
308 |
>> |
309 |
> |
310 |
> |
311 |
> I do mine each night because I'm on dial-up and it is easier to get |
312 |
> little tidbits than to wait until there is a new KDE and Open Office |
313 |
> at the same time. o_O It takes me about three days to get just Open |
314 |
> Office so I like to nibble on it a bit. |
315 |
|
316 |
Again, do you really *need* to be on the cutting edge every minute of |
317 |
the day? Especially when on dial-up? What good does the new version of |
318 |
KDE and Oo.o do for you that you have to upgrade every day? And you |
319 |
might also consider alternatives-- smaller alternatives-- like abiword |
320 |
or kwrite. |
321 |
|
322 |
Heck, you might consider using icewm or xfce (or openbox or Windowmaker |
323 |
or fvwm) instead of KDE. |
324 |
|
325 |
It's not that I'm telling you what to do, but you seem to be living |
326 |
outside your means, practically speaking. If you're on dial-up and you |
327 |
really can't afford (either in terms of time or money) the cost of the |
328 |
huge downloads that KDE and Oo.o entail, then you don't *have* to use |
329 |
them, you know. KDE is modular; you can use xfce and a selection of KDE |
330 |
programs-- just the ones you need. You don't really *need* konqueror, |
331 |
for example, since you use Mozilla for web browsing, and there are tons |
332 |
of other file managers for the file management part (xfce itself has |
333 |
one, but I like Krusader, despite it being a KDE program when I don't |
334 |
use KDE). Heck, you can use mc in a terminal (midnight commander is |
335 |
quite good, actually). And then you wouldn't have to download or compile |
336 |
konqueror anymore. |
337 |
|
338 |
If your means are limited (in terms of bandwidth), then it is well worth |
339 |
your time to consider how to maximize the effectiveness of what you |
340 |
spend that bandwidth on. Gentoo is about choice, but choice requires |
341 |
flexibility, and if you *can't* do everything you want (because you |
342 |
flatly don't have the bandwidth), then flexibility is a key skill to |
343 |
develop (how to get the most of what you want with the minimum of outlay). |
344 |
|
345 |
> |
346 |
>> If something specific is broken due to the gif/libungif issue, then |
347 |
>> tell us what that is. It may be that gif/libungif needs to be |
348 |
>> sorted out to fix whatever is broken, but we can cross that bridge |
349 |
>> when we come to it. |
350 |
>> |
351 |
> |
352 |
> I'm not sure if anything is broke or not. It was a block thing that |
353 |
> I thought may be messing up revdep-rebuild since it was complaining |
354 |
> about it. Everything seems to be working OK. |
355 |
|
356 |
Revdep-rebuild wasn't able to perform its suggested actions because of |
357 |
the block, but there are no more suggested actions, so everything is |
358 |
fine as far as that goes-- nothing is broke. |
359 |
|
360 |
> |
361 |
>> It's really not a big deal. Relax. |
362 |
>> |
363 |
> |
364 |
> I re-emerged vanilla-sources, it had the -doc on it too. Now I get |
365 |
> this: |
366 |
> |
367 |
> |
368 |
>> root@smoker / # emerge -uDNptv world |
369 |
>> |
370 |
>> These are the packages that I would merge, in reverse order: |
371 |
>> |
372 |
>> Calculating world dependencies ...done! |
373 |
>> |
374 |
>> Total size of downloads: 0 kB root@smoker / # |
375 |
> |
376 |
|
377 |
Fine, problem solved. :-D |
378 |
|
379 |
> |
380 |
> |
381 |
> Just for the heck of it, I'm doing a emerge -ev world. Just to be |
382 |
> sure. I'll skip Open Office though. It won't take to long. |
383 |
|
384 |
Famous last words. Good luck. |
385 |
|
386 |
|
387 |
> |
388 |
> My only question is about those broken things in revdep-rebuild. I |
389 |
> guess the emerge -ev world will deal with that though, right? |
390 |
|
391 |
*THERE ARE NO BROKEN THINGS IN REVDEP-REBUILD.* |
392 |
|
393 |
Dynamic linking on your system is consistent... All done. |
394 |
|
395 |
means that nothing is broken! It's just bent, and that could be due to |
396 |
user inexperience; if you ran revdep-rebuild -p after a depclean and/or |
397 |
after an emerge -uaDNv world, you might find even those 'bent' reports |
398 |
disappeared. |
399 |
|
400 |
And even if something did come up as needing re-emerging, it would |
401 |
*only* mean that one or more individual programs were linked against a |
402 |
missing or updated library, so those individual programs would not work |
403 |
until the linkage was fixed. |
404 |
|
405 |
Just because you put a single book away out of alphabetical order when |
406 |
at the public library, and the next person wanting it couldn't find it |
407 |
(because |
408 |
it's in the wrong place) wouldn't mean that the entire library building |
409 |
needed to be demolished and rebuilt! |
410 |
|
411 |
That's essentially what you're doing, with your emerge -e world. |
412 |
|
413 |
As I said, good luck. |
414 |
|
415 |
Holly |
416 |
-- |
417 |
gentoo-user@g.o mailing list |