Notices

As a backup, of course it may be a useful tool. I've used both MPlayer and VLC for such purpose for a long time on various OSes. I have for long time seen the mess of multimedia on Linux with 587925 audio and video players, each one with its own UI or written in its own language. At least when they use the same backends there is not duplicate work in that regard!

So don't read as if I'm saying VLC is useless. I just say on long term it is not a viable solution. Especially not in USA where the legality of some plugins is shady; which is why its better to buy a Fluendo Bundle instead. In that regard, it may not even be wise to distribute ffmpeg/libavcodec/xine/vlc/libdecss in Extras-*

Also, Windows users are familiar with VLC having a GUI. On Maemo, it doesn't have one.

I quote from Wikipedia Maemo entry:

Nokia provides these frameworks to be used; not mimicked.

Ofcourse, there are situations where you're OK with an alternative. Qt besides GTK, WebKit besides Gecko. Not ideal, but OK. But if you have for example some applications using WebKit and other Gecko you're not sharing memory from shared libraries where you could, and if you use a Qt application in a GNOME desktop you must load a lot of extra libraries which also uses more resources besides the fact it doesn't look good (yes, even with qgtkstyle).

So if you look at VLC from this angle (duplicating work, different expectation/outcome from desktop port) it is not a very good addition. Instead, we want the right plugins in GStreamer. So if you cannot play content in a multimedia player using GStreamer multimedia framework talk about it here so we can instead get it supported. GStreamer supports tons of plugins as well on Linux/x86-{32|64} hence solution could be as simple as compiling it for Maemo because it already works on Linux/ARM, or we could team up and ask Fluendo to port to Linux/ARM, either by paying then individually for licenses or getting Nokia moved to pay for it with a discount license.

So, it's okay to have modular software, but it's not okay to expect modular cell transceivers in a handheld device.

See what I did there?

Most people don't care about legality of plugins. The only problem with the legality of the plugins would be its inclusion into Maemo Extras as you pointed out. And for most people who are new to the platform they shouldn't be sent out of extras. For everyone else they could just easily add a VLC repository (if they were to offer one) while waiting for whatever alternate solution there is.

Yeah, I see what you did there: you applied something I said about software on embedded hardware.

What I said about software applies even more on embedded software than desktop software because of limited resources.

In hardware is much more limited. Especially embedded hardware. Modularity comes at price of more power usage and more size used. Same for features. I'd also love to have dual SIM, for example, or be able to replace the WLAN later with 802.11n, but that is not possible either.

Question is also which interface you'd have that modular cell transceiver. I think we discussed the options before: there is no suitable one for handhelds for end users while on x86 hardware from netbooks and up you can have MiniPCI Express.

That is why Apple solders their stuff, and does not allow user to replace. The side effect is ofcourse you are dependent on them and have to upgrade whole hardware. Well, Nokia benefits from such side effect too. Else we'd just upgrade our phones ourselves the whole time.

So, you didn't see what I did there. And who said anything about Apple?

I could see VLC being a big problem on Android- due to byte code translation (Coreplayer team tried and quit), but hardware access is not restricted in Maemo- is it?

If truly optimized to the N900's drivers and OS, would VLC not perform as good as Mplayer?

Please explain your riddle.

Me, for example of embedded hardware modularity design choice.

It's called "MPlayer".

You were asking developers to be modular because of limited resources, but you argue that it's not reasonable for manufacturers to do the same. I consider that a contradiction, and your reasons seem to be just rationalizations.

This comment has more to do with planned obsolescence and less to do with modularity. More to my point about contradiction. With your reasoning, wouldn't it be more in a software developer's interest to share fewer libraries with other software? They could then force the user, because of limited resources, to have fewer choices, which appears to be your agenda anyway.

I think I'm gonna fork gstreamer to daperlstreamer...

Here's my pleasant experience with the Almighty VLC that can play Every Format Known To Mankind:








Excuse the cartoons.