Stability Release (2 Viewers)

[elliottmc]

Here is my opinon:

1. Mediaportal is still in beta phase. I don't see how anybody can expect beta software to be fully stable

Very true.

2. You guys are requesting a stable version, but at the same time
i see >1000 feature requests in the forums. This conflicts and it is not possible to please everybody. We cannot stop adding new features to make MP stable because users demand new features

Now here I respectfully disagree with you. This is the difference between open source and commercial software. A commercial development would decide on the features which they wanted and implement them, hopefully in a stable fashion.

I think you need to focus on what people are using Media Portal for, and which features the vast majority of your users will actually use.

Let's take a case in point, the Tetris plugin. Sure, it's cool, but it's a novelty (and I mean no disrespect to the programmer when I say this). Tetris is not a "required" feature of a HTPC. Now, a MAME plugin (which ShowShifter has) would be of more general use, but even so, I tried this for a couple of hours and then got bored.

The core functionality (IMO) is watching, recording and timeshifting TV, playing video and audio files, and generally being user friendly. Linking in with CDDB and IMDB, ripping CD/DVD and burning CD/DVD are nice extras, and "almost essential", but only once the real basic featureset is done.

As a personal opinion, I think that you need to put these 1000 user requests on ice. No-one (myself included - hypocrite that I am!) is in a position to make demands on you. You are doing this because you want to, and ultimately you will do it however you see fit. I'm just looking from the point of view of an end user, and Media Portal will not be replacing ShowShifter on my HTPC. Meedio has a chance once they release DVB-T support later this week, and that has a very extensible architecture, but is still snappy and responsive.

3. i think 80% of all problems are related with the My TV plugin. All the other modules are pretty stable

I would certainly agree with this. If myTV was as stable as the rest of Media Portal then I would switch.

4. Reason TV causes all the problems is because :
MediapPortal Team:
- bugs (offcourse)
- the fact we're volunteers and dont have 40hrs/week to work on My TV
- the fact that there are just 2 developers working on My TV
- the fact that we dont got enough serious/dedicated and motivated testers

Environment:
- the vast amount of different tv capture cards
- the hardware differences between every PC
- the audio/video codecs people use

3th parties:
- some tv capture cards have very buggy drivers
- the lack of good documentation on directshow (blame M$ for this)
- 3th parties not releasing documentation to opensource projects

End Users:
- not reading any documentation
- using pc's which dont follow the system requirements for MP
- wrong setup/configuration of MP
- using latest CVS versions and expect it to be bugfree
- using non-supported tv cards

I think everyone realises what a difficult job you have. myTV runs like a dog on my system. This could be my fault. I would be more than happy to help with the testing, but you would have to let me know what you need from me. I'm happy to install software for testing. My programming skills are rusty, but I have written code, compiled and profiled.

Now what have we done to improve this:

1. we introduced the pre-releases and final release
The +/- 2 weeks between a pre-release and final release is used for testing. Its forbidden for developers to add any new code and the entire 2 weeks are used for testing, debugging and fixing bugs. However since we cannot test everything (due to lack of time and available hardware) we have to rely on testers

Based on my own experience, 2 weeks is barely enough time to get a response from people about bugs, never mind fixing them. Of course, as this is an ongoing project, any bugs not caught in the 2 week period will get fixed in CVS versions and subsequent releases.

2. Attracting new developers and testers
We tried in every way to get good testers and good developers.
However in the end this resulted in nothing. Developers seem not interested which is weird because SageTV/Meedio and others have no problem attracting developers.

I'm amazed that Meedio has a much more active community than ShowShifter, despite the current lack of DVB-T support. Maybe the underlying architecture is better. From a user point of view, if I wanted to add a plug-in to HTPC software, I would choose software which I was using - again, stability is the key.

- Making MP less dependent on 3rth party code
Part of the problem is that MP now uses codecs, filters etc written by 3th parties like microsoft, intervideo,... We're currently in the process of writing our own codecs/filters. This means we will have full control and can fix bugs because we wrote the software ourselves.
At the moment when there's a bug in a microsoft filter we cannot fix them. Even worse, microsoft is fixing those bugs, but is only making them available for microsoft media center edition and not for windows xp users

I have no experience here, but from a user point of view, other software seems to be able to use the codecs fine. ShowShifter can record from my card and use my divx install to convert between SSF and divx avi files. Maybe you need to restrict your MPEG2 and avi codec support a bit in the short term in order to narrow down the problems. Maybe some things which you think are codec-related actually aren't. I know from experience that a change to one section of code can have unforseen implications.

I'm certainly not intending to tell you how to do your "job", as ultimately I wouldn't have a clue how to write a true multimedia front-end. I guess I am a little more informed than many potential end users of Media Portal (as are most people here) and would like to make helpful comments. I have the utmost respect for the team's skills, as you wouldn't have got this far without being superb programmers.

If there is any way that I can help, please let me know. I'm more than happy to help with testing.

Best wishes,

Mark

 

[Smirnuff]

My time is better served coding than expressing my opinions here but on this occasion I feel obliged to make a few comments and would request that each of the advocates for a freeze put themselves in the shoes of the team members for a few moments, ignoring the fact that we don't get paid to do this and have day jobs and families and so on, thats not so important as we actually choose to spend some of our free time coding for MP...

Through the course of a day a skilled developer can make a huge dent in a bug list when presented with the appropriate information. We don't have enough serious and dedicated testers to have that information to hand so usually spend more time trying to recreate a bug than we do fixing it. It would be very nice to have a small army of developers, it would be nice to get back to a point where we can do fun things like code Tetris plugins and so on, but the real need IMO is for more testers. If you want the stability of MP to improve then its time you all started doing something for yourselves!!!

We are all doing our level best to develop an application that we hope will go on to be the undisputed leader in its class. We have the skills and the motivation to achieve this goal, however we aren't miracle workers and for me it comes back to the sad fact that the community don't do enough to help themselves, you're not helping us here, we can take care of most of those ourselves the following day, you are helping the 16 people that you communicate with via ICQ and your families' trouble free evening of TV and movie viewing via MP.

My closing comments would be that we simply cannot win, we freeze and you'll ask for features, we give you more features and you want us to freeze. You disagree? The thread starter, posted a feature request just 3 days ago yet today he's suggesting a feature freeze?!?!?!

For me such threads serve no real purpose, in fact all they tend to do is lessen my motivation . What would be really useful is if a couple of you could get together and come up with a structured approach to testing and then offer your services to the team, that way you'll have a hand in driving towards a marked improvement in stability.

@kamrat: I'm not sure what it is that you're doing but it would be good if you could look into it, I've deleted your duplicate posts at least 6 times already today

 

[infinite.loop]

We don't have enough serious and dedicated testers to have that information to hand so usually spend more time trying to recreate a bug than we do fixing it. It would be very nice to have a small army of developers, it would be nice to get back to a point where we can do fun things like code Tetris plugins and so on, but the real need IMO is for more testers. If you want the stability of MP to improve then its time you all started doing something for yourselves!!!

if i would have a PC where i can just test mediaportal with different codecs/filters,... to find bugs i would do that. but like many of us i only have one PC i can use Mediportal on.
this means i am realy glad that it is working on this PC with only a few issues i have allready posted in the bugs section. i cant risk to skrew it up with using cvs or changing codecs/filters to see if they work or not.

the only thing i can do is report you the bugs i noticed when configuring and using mediportal and if i could solve them with the help of the community. if you tell me install this, try that, is the error gone now i will do that. but when its working i wont change anything just to see if its working or not.
and i have done that quite often as you can see in the forum.

My closing comments would be that we simply cannot win, we freeze and you'll ask for features, we give you more features and you want us to freeze. Don't believe me? The thread starter, posted a feature request just 3 days ago yet today he's calling out for a feature freeze?!?!?!

my request was an warning message when exiting mp "attention a recording is in progress" not more.

i my opinion the feature request forum is there to give you an idea what is wanted by the users.
But when you tell me:"sorry we have no time to add new features at the moment because is more importend to get independent of 3rd party codecs/filters (or something else". i say: " no problem". you know whats more important to do. just wanted to tell you my request. when you think its useful please integrate it. thats it.

 

[Marcusb]

Something seems to be up with the forums tonight, as two of my posts have vanished from this thread (and I promise I was clean and on-topic)

@Smirnoff: I appreciate where you are coming from, but most of the regulars do try and help as much as possible. I have had two problems with the past two major releases, the memory leak in v0.1.2 and the crashes to desk top in v.1.3. In both of those times most of us posted as much information as we could, using the support template. We also asked if there was any further information we could provide that may help the devs, but we never got any answers to that request.

I wish that I was a .net programmer, so that I would be able to provide some assitance to a very overworked and dedicated team of volunteers.
As it is, the best that I can do to help is to lurk around the forums to answer noob questions to try and take some of the strain off you guys.

We respect your work immensely and every person here is grateful for all your efforts. We try to help as much as we can, but sometimes we need directions

Please don't take this thread as a negative, even though some posts may seem so. We are all being as tactful as possible as the last thing we would want to do is make you think we don't appreciate you all.

Thanks,

Marcus.

 

[Marcusb]

ok, it seems posts are returning this error message:

General Error

Could not connect to smtp host : 0 : Success

DEBUG MODE

Line : 111
File : smtp.php

The post is still being accepted but if you hit refresh it results in a double post. Just a warning to the others [/b]

 

[Smirnuff]

Fortunately one of the benefits brought about by the .NET framework is that it supports side by side installations, MP continues this support by allowing you to have as many varied copies of MP on a single machine as that machine's hard drive will allow so for those that are interested in helping out further with testing but are also concerned with breaking a working install...

Open Windows Explorer and navigate to your Program Files folder and simply copy the existing MP folder then apply whichever CVS build you wish to work with over the copy, that way you can do as much damage as you like in the 30 minutes and not have have to worry that you're going to have to get MP up and running again, full of channels and guide information in the 10 minutes you have left before your partner's favorite soap starts; which could potentially lead to reduced nookie for the rest of the week.

Its best to do the copying while MP isn't running, some files may be locked exclusively so the copy would fail.

 

[infinite.loop]

i didnt start this thread to attack the developers.

i just want you to consider my idea of an Stability Release with focus on bugfixing and stability improvements (i.e. 3rd party software problem).

if you say no, thats not the right time now. its okay.

 

[Smirnuff]

i didnt start this thread to attack the developers.

Nobody is suggest that you did or have, my comments reflect my own personal views on how we can get to the point where we have a more stable application and I guess could also be seen as another attempt at recruiting some serious testers.

Marcus: I don't see your post as negative in the least and we do appreciate the help that yourself and others provide around the forums, on IRC and through the submitting of bug reports, something we'd like to see much more of.

 

[CHli]

I love MP, of course it's not stable at the moment but that's the matter of a beta software.

The only think that's sure is that no one will answer : "No I don't want a stable version of MP" because that's all what we want. But that's won't be achieved in one day so be patient.

I think frodo is right, the only big problems at the moment are with the MyTV plugin. The main concern is that DirectShow is so badly documented that only MS can use it without trouble. But that would be really cool if this plugin could be a little more stable cause it's the only important one for me. (If I can't watch, record, timeshift TV properly why should I use MP ?)

So thanks to the developpers, thanks to everyone who helps improvings this great software and stability will be achieved with the help of everyone. (e.g. by sumbmitting well documented bugs !)

 

[elliottmc]

Through the course of a day a skilled developer can make a huge dent in a bug list when presented with the appropriate information. We don't have enough serious and dedicated testers to have that information to hand so usually spend more time trying to recreate a bug than we do fixing it. It would be very nice to have a small army of developers, it would be nice to get back to a point where we can do fun things like code Tetris plugins and so on, but the real need IMO is for more testers. If you want the stability of MP to improve then its time you all started doing something for yourselves!!!

I'd love to. All I really know about Media Portal is based on my own experience.

Would it be fair to say that on the developer's machines, myTV works perfectly, switches channels quickly, records and timeshifts without maxing out the CPU?

If this is the case, then I understand why it is so difficult to optimise, and someone who knows what they are doing needs to tell me what information you want from me in order to work out why my experience of myTV is so different.

Based on what I've read on here, it would appear that I am in the majority. However, people are far more likely to post with problems than with success stories, so I may be seeing a very distorted picture.

For me such threads serve no real purpose, in fact all they tend to do is lessen my motivation . What would be really useful is if a couple of you could get together and come up with a structured approach to testing and then offer your services to the team, that way you'll have a hand in driving towards a marked improvement in stability.

I'm not sure I could come up with a structured approach to testing. What you maybe need is a group of up to 10 people who are experience computer users, possibly even programmers, who are having problems with Media Portal. You need to define the problem for them. No sense us spending time trying to pin down bugs if you don't consider those bugs to be a priority (i.e. if you are happy with the general performance).

If you just ask the whole community for an opinion, you will either get no response, or so many varied opinions that they make no sense at all. You need a small group with realistic expectations and goals.

If I could profile the code and work out where it is spending most of its time, I would. My previous programming experience is all on the Mac (pre-OS X), so that although I am familiar with the concepts, I don't have any PC programming experience which will help.

Nevertheless, I am volunteering if you can tell me what you need me to do.

Mark