IS wrote:
>> This is why I have resisted mISDN so much... I have NO WAY to really
>>test it. Our particular scenario is not a good example (I can at least
>>try to load the module), but beyond that I don't even have hardware -
>>let alone something to connect it to.
>
>
> First of all, I did not want to start a huge discussion now. I just wanted
> to point out that Astlinux becomes over time much more complex since it
> supports more and more tools and applications - what is really fantastic.
> But the level of complexity plus the dependancy on updates on all the
> packages from the original developers, makes it hard for a small team to
> keep the pace with high quality. You are doing a fantastic job, everybody
> admires your work and I really like astlinux a lot!
>
>
>> I would love to test this more, but at this moment in time I am
>>reluctant to test anything related to mISDN because it is literally the
>>only part of the system that I can't test.
>
>
> I fully understand you. But do you really want to test all possible
> hardware components yourself? How can you manage this as more and more
> hardware you want/will support? Wouldn't it make sense to delegate some of
> this work to other people? Sure, you have to trust them and they have to
> be capable in doing this job.
>
>
>>This combined with the clear
>>fact that mISDN is inherently more flaky than other parts of the system.
>
>
> What I don't like about mISDN is how it gets build. It's all done in a
> home-grown makefile. If there is a change in the makefile and we want to
> upgrade, we have to patch the patches. That's awfull! The overall quality
> of the software itself is quite good. Maybe some features are missing, but
> it works. And it works reliable.
>
>
>>Your work with it has been excellent, but it is obvious that
>>mISDN/chan_misdn still isn't quite "stable stable".
>
>
> I agree, that it could be more stable. However, I have a few systems up
> and running now in production environment and I have no issues. It just
> runs.
>
>
>>I have done
>>hundreds of builds over the months and mISDN breaks more than any other
>>single component - no doubt about it.
>
>
> What can I say...
>
>
>>I got so sick of it that I flat
>>out disabled it in the standard config.
>
>
> That's ok since not everybody is using it. Nevertheless, it can be added
> to the release builds. People probably had/have issues if they don't have
> ISDN hardware, asterisk doesn't start because chan_misdn stops the whole
> startup process if it doesn't find its subsystem. That's annoying. By
> having chan_misdn excluded from loading by default, would make sense.
>
>
>>Even though only a very small
>>percentage of users actually use it, it seems to be the single most
>>often complained about "feature". Now that you have switched to
>>snapshots it should be better, but still...
>
>
> Again, what can I say... If there is a suitable alternative, we may want
> to switch. There is bristuff (never tried it, I only read that it probably
> has some more features, but again some code patching is necessary. I
> personally don't like that.) and there is visdn. I tried visdn, but at the
> time I tested it it had too many bugs.
>
> So what now? There are several options... I don't know. It's frustrating,
> I know.
>
>
>> The builds from the last few days were made solely for people to test
>>them. The announcement was only sent to astlinux-users (not even
>>announce) and they have only been provided in a directory on
>>mirror.astlinux.org - no mention on any website (except for DAN). I
>>purposefully wanted to limit their distribution to only people on this
>>mailing list that were interested in testing them.
>
>
> Good, unterstood. I only had the feeling that people here join the list,
> try the stuff and get unhappy (mostly about misdn as you said) because
> they assume that it just runs fine.
>
> It is good that you explained this fact here and I think it was helpful
> not only for me.
>
>
>> Anyways - it appears that this is (hopefully) a simple linking problem.
>> I'll try to get the module to load when I work on it later tonight.
>
>
> What acutally caused this linking problem? Something outside mISDN must
> have changed since we are using frozen snapshots...
>
> So again, sorry if I made people upset here, especially you. But in my
> opinion it gets very difficult to manage this project if there is only a
> single person is managing the entire project and if more and more packages
> get added.
>
> BTW, I am NOT saying that I should be the one who should maintain the
> mISDN stuff. I did this so far because it was fun for me and I could at
> least contribute a bit to the project. There are lots of smarter people
> out there and everybody should feel free to jump in and contribute!! If
> someone wnats to add bristuff or visdn -> do it!
>
> cheers, Ingmar
Ingmar,
I didn't want to sound defensive, but I was concerned that others
missed the point on these last couple of releases - they were for
testing because I *knew* that things like this would be broken. I have
learned by now that I am not the best person to find problems -
especially in things that I don't use :). 0.4.4 is supposed to have
working mISDN support again.
Anyways, moving forward - when I merged your branch with trunk/0.4
there were conflicts with some of the patches (good 'ol Makefile
patching), and I had to edit some of them. It is possible that I broke
something. I'll grab your branch and try to get to the bottom of this.
And as far as BRI/ISDN goes, quite frankly, it all sucks (under
Linux/Asterisk) :). I have looked at the others, and mISDN is the best
overall solution. :(
--
Kristian Kielhofner
_______________________________________________
Astlinux-users mailing list
[email protected]
http://lists.kriscompanies.com/mailman/listinfo/astlinux-users
Donations to support AstLinux are graciously accepted via PayPal to [EMAIL
PROTECTED]
