Why Knoppix 3.4 is delayed again ...

[TyphoonMentat]

I'm definitely against removing development tools, as for one they're invaluable when it comes to remastering (I remaster with the LiveCD), and general testing purposes - if your program crashes, it can't wreck the system.
It's amazing how many people ask on forums of mini-Knoppix remasters (i.e. Damn Small or Feather) for programs like gcc, so a LiveCD with these tools is obviously in high demand.
Also, it helps attract Windows techies and programmers - "Look, you can even compile something from this CD in C, C++, Perl, Python..."

I'll take an in-depth look at the Knoppix package list later and try to give exact packages that in my opinion are merely ancillary.

[TyphoonMentat]

Right, here's my list of things to remove:

aalib1 and aalib-bin - these are pretty much useless
acroread - this is non-free anyway, and there's xpdf and gv
bb - an ASCII art demo, no real use
bochs - some might want this, but in general people don't use it (this means all bochs packages and bximage)
chromium - On most systems it runs far too slow anyway
dietlibc-dev - This is only needed for those who are perhaps creating bootfloppies with dietlibc
fb-music-high - Does Frozen Bubble really need high-quality music? Can we save space by reducing the quality?
freeciv - This one's debatable, but it's a game
icewm - KDE and Fluxbox are enough IMO
ISDN utils - you could keep these, but I don't know of anyone who uses ISDN nowadays
jpilot - use kpilot
kasteroids, kbattleship
Koffice - reasons have already been mentioned
larswm
lynx - elinks is better, and it's already there
Mozilla - Firebird and Thunderbird are smaller, unless you use the Debian packages...
Nedit - large, and we have just about every other editor anyway
Perl - remove the threads support, it's not often used and makes Perl much bigger
selflinux - if you really had to, you could save around 20Mb doing this
TCT - replace with the Sleuth Kit
vim-gtk - Most people only use the console version (I think...)
vlc-aa - More extraneous ASCII art stuff
xscat
zile - we have Emacs, so we don't need a tiny version of it

I have some things I'd like included as well, but we're trying to save space here, so I'll refrain, except for one thing - I really really think amsn should be included, as lots of Windows users use MSN Messenger, and it has a familiar interface and practically all the features that MSN Messenger does.

[Del]

I am probably way too late to make any difference here. Anyway, here ti goes:
Those of you who still need the 2.4 kernel have the 3.3 version, and I see no reasons (might be my ignorance) why you would need the 3.4 version. Remove the 2.4 kernel in 3.4.
Some of you don't realize that language is critical, European languages use different keyboards (which is what I am familiar with), meaning that without language support I won't even be able to get the keys on my keyboard right. I don't think Klaus would like an abundance of different ISO's to handle, and local remasters will always be remasters. Leave language support in!

The most sensible thing is to cut applications that serve the same purpose. Konqueror and Mozilla are not such applications, they are both needed. Many text editors however are, but don't cut Emacs. I know many of you love to hate it, but we who belong to the majority use it. PDF viewers were mentioned and in my opinion is a good example, likewise with image viewers, lightweight desktops (those who uses them can easily make the transition, e.g. from window maker to fluxbox). Office suits however is a tough one. MS users get the chills when they see how slow OO (and Mozilla for that matter) loads. Then again OO is more powerful. Keep both!

I think multimedia support on the other hand should be strengthned, cutting out TV card support is stupid. If we want to convince MS users we'd better show them that their fancy hardware is supported, and for todays desktops multimedia support is extremely imortant. For that reason I have been forced to ditch knoppix and rather use overclockix when I want to convince MS users that Linux is a viable alternative (nvidia/ati driver and mplayer included).

[zenok]

At my side - a idea is coming up - I will create maybe an own Live Linux and this will parted in two Live CDs: Zenoppix User and Zenoppix Server

I think from time to time Knoppix has to say this is our target and maybe Knoppix is soon only a base for other specific remasters ... mh .... as other says kde-network suite VS mozilla and koffice VS openoffice and many other duells .... but when I create my own Zenoppix LiveLinux the user will vote/post for the packages and it bases of the next 3.4 DL Edition

[proto]

Dont get too excited abt 3.4 how many of us have mastered 3.3

[Hasse]

... European languages use different keyboards (which is what I am familiar with), meaning that without language support I won't even be able to get the keys on my keyboard right.

The internationalization (i18n) support contains translations of kde apps and menus to different languages, at the cost of 100 mb. Keyboard layouts and locals (currencies, time formats etc) are not part of the i18n packages and you will not loose any keybord layout by removing any of these.

Remove all languages but a few, in order: english, german, spanish, french, russian and japanese.

[agent_smith]

Remove all languages but a few, in order: english, german, spanish, french, russian and japanese.

True, true (Budweiser style)

[MildBill]

when reading the comments, alot ppl favor removing Koffice and Mozilla, and only having the 2.6.x kernel. guess thats ok, although I prefer Koffice for its speed - but you can always install it after the hdinstall.

The whole point is that with the HD install, speed is a lesser problem, it is on the live CD that speed is an issue. Leave KOffice, there must be many other ways to regain a mere 65 MB.

The many ideas of a second CD with lesser used development items, other window managers, etc is probably a very good one. This is where an HD install shines, as you can then add these items if you want/need them.
However, the very notion of the live CD version of linux is what makes Knoppix so very attractive. Trim off the fat, keep the main muscle items, and put the rest on the second CD for those who want an HD install.

[OErjan]

eeeh, if the ISO is 65Mb tooo large that is almost 250Mb to be removed in apps.
LOTS of trimming.

[MildBill]

eeeh, if the ISO is 65Mb tooo large that is almost 250Mb to be removed in apps.
LOTS of trimming.

A good point, but still...
Seems to me that that large of a size increase could not have simply snuck up on the developers, it has to also be the inclusion of more apps, more bells and whistles, etc. So, again, trim the fat, make the live CD the best it can be, but move lesser used, slower apps to a second CD, for installation on the HD, and possibly for inclusion on a custom live CD of the users design.

OR, as has been suggested in a backhanded way already, do some redesign (probably for the 3.5 release, as this will take time) such that you have one CD which is the trimmed down live CD, and a second CD which allows for install of it's packages to the HD. The key change is to make it simple for people to create a third CD, which is a custom live CD containing whatever packages they really want/need.
They would start with the original live CD to create a 'barebones' version, then add what they wanted from the other CD, or from the Internet, until they were done, or ran out of room. Here you could build in the option to allow for 650 MB CDs, 700MB, 800MB, or DVD, as the users option. And, as much as possible, include necessary scripting or whatever to optimize the remaster as much as is possible, to reduce speed issues.

Just my $0.02

MildBill

NOTE: As of this time, I have to say that I am NOT a user of Knoppix yet. Since I am a laptop user, I am waiting for the 3.4 release to try it. I have, however, used linux before, so I am not a total newbie. but I am anxious for the 2.6 kernel, and the updated version of KDE before I commit to it.