1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA1 |
3 |
|
4 |
On 10/05/12 09:54, Olivier CrĂȘte wrote: |
5 |
> Hi, |
6 |
> |
7 |
> On Thu, 2012-05-10 at 06:34 +0200, Fabio Erculiani wrote: |
8 |
>> I think expressing my own opinion about Lennart-made software is my |
9 |
>> right, after all. |
10 |
> |
11 |
> I would express my opinion about Fabio made software, but I've never |
12 |
> heard of any. |
13 |
|
14 |
Not his fault, he wrote plenty of interesting stuff though. |
15 |
Fabio attitude still isn't that horrible regarding feedbacks, Rigo got |
16 |
created more or less because the previous UI got a sound "it sucks". |
17 |
|
18 |
His quite short and a bit extreme reaction probably is due having lots |
19 |
of unhappy user complaining at him for some issue with avahi (hangs in |
20 |
bonjour now and then) and pulse (skype freezing randomly anyone). |
21 |
|
22 |
>> Firstly, it's almost impossible nowadays to avoid including avahi, |
23 |
>> systemd and pulseaudio into a desktop distro so, there is no real |
24 |
>> choice. This issue became a sensible matter for those users who for |
25 |
>> instance, wanted to have a silly mp3 player working without going |
26 |
>> through the PA nonsense, really missing the old |
27 |
>> ALSA-oh-it-was-always-working days. |
28 |
> |
29 |
> Maybe the reason every sensible distribution uses Avahi, Pulseaudio, etc |
30 |
> is because they are better than other solutions out there? |
31 |
|
32 |
If there are solutions somebody will use them, if people are aware of |
33 |
them and doesn't get too hard. I did like the concept about pulse and |
34 |
even wrote support for pulse in a certain fringe software you might use. |
35 |
|
36 |
The pulse concept is quite good, some corner cases and some design |
37 |
issues make it annoying at time. The fact some of them are consider |
38 |
"features" or "design" obviously make the whole thing less nice. |
39 |
|
40 |
> Do you think is a fast conspiracy to make your life suck? I believe |
41 |
> engineers in every distribution are looking at what's available and |
42 |
> picking what they think is the best solution, and it turns out Lennart |
43 |
> is pretty damn good at making useful software. |
44 |
|
45 |
No, he is pretty damn good in getting interesting concepts, having |
46 |
people sold on them and then you need 5 years to have the audio seldom |
47 |
crash, bonjour seldom kill pidgin and so on. |
48 |
|
49 |
Till it is some minor annoyance that is comparable to not having the |
50 |
feature or the same to other feature provider (dmix isn't exactly great |
51 |
as well) you surely can live with it. |
52 |
|
53 |
> Was alsa always working? I remember spending hours trying to figure out |
54 |
> the right control in alsamixer and fighting with alsa's arcane |
55 |
> configuration languages (it has 3 different ones). And how do you deal |
56 |
> with modern technologies like Bluetooth audio without Pulseaudio |
57 |
> exactly? |
58 |
|
59 |
I used to do that and it was working sort of fine even if it was |
60 |
crashing in dbus... |
61 |
|
62 |
>> Of course, I am not only bringing my personal opinion here, but the |
63 |
>> one of the majority of users I've been talking with. |
64 |
> |
65 |
> I think you only hear from users who like to complain, others are just |
66 |
> happy that everything works for them thanks to Pulseaudio, systemd, etc. |
67 |
|
68 |
As said, if they are minor annoyances most people would just cope with them. |
69 |
|
70 |
A - "Skype hangs because pulse? oh well, let's reload it no biggie" |
71 |
B - "AAaargh I missed the important confcall because #%$#@ skype hang |
72 |
due pulse, I hate YOU Lennart!" |
73 |
|
74 |
A and B are different reactions from the same small issue. |
75 |
|
76 |
> If you think that Lennart does not solve problems, maybe it's because |
77 |
> you don't even understand what the problems were? For example, I |
78 |
> encourage you to read about how the dynamic latency in PA allows for |
79 |
> lower power usage or how modern audio hardware is designed to use a |
80 |
> userspace sound server, etc. |
81 |
|
82 |
I recall when the whole thing got initially reported and it was "pulse |
83 |
eats my batter" and if you consider that the stock pulse on ubuntu |
84 |
oneric eats about a *least* 10% cpu on imx51 due funny resampling loops |
85 |
you know something needed some more attention. I guess I'm digressing. |
86 |
|
87 |
The main issue is that udev best replacement so far is mdev plus some |
88 |
additional helpers to let applications using libudev or the dbus |
89 |
interface still get compatibility. |
90 |
|
91 |
So having udev merge with systemd is quite in the shovel meet throat side. |
92 |
|
93 |
People that had and have some bad experience with pulse and avahi or |
94 |
directly with Lennart stubborn and abrasive personality can be *quite* |
95 |
concerned about this "vertical" and linux-only approach. |
96 |
|
97 |
If you consider that in 2 weeks the whole thing went from "udev moves to |
98 |
systemd since is easier for us, but not be concerned udev can build |
99 |
stand alone" to "udev stand alone is unsupported" you can see that isn't |
100 |
that simple and lots of people might start to get angry. |
101 |
|
102 |
lu |
103 |
|
104 |
- -- |
105 |
|
106 |
Luca Barbato |
107 |
Gentoo/linux |
108 |
http://dev.gentoo.org/~lu_zero |
109 |
|
110 |
-----BEGIN PGP SIGNATURE----- |
111 |
Version: GnuPG v2.0.18 (GNU/Linux) |
112 |
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ |
113 |
|
114 |
iEYEARECAAYFAk+xU3sACgkQ6Ex4woTpDjTNewCfU5cahmNPbgKQJt/2GkbVBh4o |
115 |
F1gAnjheSaIVRF55g1//9wu5dFe8ga3w |
116 |
=FlU7 |
117 |
-----END PGP SIGNATURE----- |