Crazy idea / suggestion

[Joe1962]

BTW, could the XFCE Start menu / Quit bug be memory-related? It has a 100% failure rate in a 64 MB vm, and AFAICT a 0% failure rate in a 512 MB vm.

[Witek Mozga]

Can we produce a "really light" version of VL that would be aimed @ even older boxes than STD is?

This is wrong way in my opinion. Generally  speaking I dislike the fragmentation of linux distros. Instead of spliting forces on several hundreds distros people should rather unite to make fewer of them but more polished and less buggy. The same IMHO holds true for Vector: instead of dividing power among Standard, Live and SOHO maybe it would be better to concentrate? Don`t try to make a version for every possible purpose.

Why is Vector based on Slackware? Because Slackware is good, I suppose. Why is it good? Because there is only ONE Slackware that shines.

BTW: Vecor is rather addressed not for newbies. Thus people using it probably should have enough experience to removepkg and feedle in /etc to make it work on older machines.

Anyway, the developers must decide if they have enough time to take care of all those Vector clones.

[easuter]

Distro proliferation is Linux's weakness and strength at the same time. Weakness, because as you pointed out, the development effort is somewhat spread around instead of concentrated in one single distro.
BUT: there is NEVER "one size thats fits all", there are always niches that can't be satisfied by using one distro, so a new one is made.
And competition is always healthy 

[Witek Mozga]

BUT: there is NEVER "one size thats fits all", there are always niches that can't be satisfied by using one distro, so a new one is made.
And competition is always healthy 

Competition is healthy because there is natural selection. And current data for selection of the desktop use (and Vector is aimed for the desktop, I guess) according to latest Slashdot is as follows:
"OS X has seen continual growth, from 4.21% in Jan 2006 to 7.31% in December 2007 at the same time, Linux's percentage has risen from only 0.29% to 0.63%"
Which means linux on desktop is within statistical error. I think that distro proliferation is one of the causes. Competition is very healthy within linux world only, but when compared to outside, well...
Maybe Windows and Apple are succesfull on the desktop because for them there are no niches that can't be satisfied by using one distro?

[easuter]

Maybe Windows and Apple are succesfull because for them there are no niches that can't be satisfied by using one distro?

I really wonder how accurate those numbers really are. Either way, MS and Apple are in it for the money, nothing else. Their users are led to believe that their problems have to be adapted to the OS, when it should be the other way around. and most people don't question that way of thinking. The few that do are having this conversation right now 

Most of FOSS devs couldn't care less if their app/OS dominates the world, its the technical challenge and ego that drives them 
Linus Torvalds wrote the kernel "Just For Fun" (tm).

And thats why companies like MS and Apple are like so many others that can just die out, because their entire premise for existence is money, nothing else.

[bigpaws]

You have to love numbers. You can take any
numbers and  through languafe make look how
you want.

So look at the facts if these numbers are correct
then:

Mac had increase in use of  76% impressive numbers.

Linux on the other hand had 117% increase, maybe
smaller numbers but larger growth.

Bigpaws

[Witek Mozga]

Most of FOSS devs couldn't care less if their app/OS dominates the world, its the technical challenge and ego that drives them 

I realize that Same for me: it`s the technical challenge and ego that drives me to use linux
But wouldn`t it be most honestly to emphasize when installing a linux distro: "hey guys we make this distro because of technical chalenge and ego and don`t care much if it dominates desktop usage so don`t expect we understand the needs of average-joe"?

[Witek Mozga]

You have to love numbers.

Slashdot does. In a way I do, too.
Emmanuel Kant said: there is as much truth in science as there is mathematics in this science. And I agree. Every computer users loves numbers in a way, even if he does not realize that 

Mac had increase in use of  76% impressive numbers.
Linux on the other hand had 117% increase, maybe
smaller numbers but larger growth.

I know that there is lie, bigger lie and statistics
Anyway 117% grow in values as low as within statistical error can be safely ignored.

[rbistolfi]

If I understand Moe-LNX proposition correctly, the difference between the VL flavors is mostly in the packages, Moe stated would be not much work to develope another vl flavor. Saulgoode said this could attract more people to the vl project. Those arguments are still there. This is being proposed with the assumption VL will have more resources by a depreciable amount of work. We should focus on those arguments. In the other hand, as Joe pointed, STD already performs very well if you do careful choices and heavy apps can be replaced easily.
I just wanted to do a resume and kick the thread back to the original topic 

[exeterdad]

Agreed rbistolfi.

My thoughts.  I've seen many requests lately, and many people stating that VL has become bloated compared to past releases.   I personally think that a slimmed version of VL would be great. A minimalistic install.  Don't try to make the masses happy.  Just enough to get a usable desktop.  I feel Fluxbox would be good choice.  It might not be a preference to many, but it's usable enough for any first timer to comfortably get productive, or enough get to Gslapt to start customizing.  I am hopeful when the iso is pruned down, it isn't pruned too deeply as the majority of the core packages  have no slack-depends file.  So if the user in the future needs to replace these packages, they are going to have a tough time as there will be no dependency tracking.

Also, as pointed out, it would be more visual activity for us on DW.  And maybe even attract an audience that may of passed us over in the past.  And for much less effort from a development standpoint.

This is all fine and dandy except for one thing...  Who will step up to the plate and take this on?  I'm not a dev, but I've gotten a small taste of the "behind the scenes" here at VL.  There is a ton of work going on until the wee hours of the morning, on a daily basis.  And it's sick how few dev's are doing this work.  I have a tiny little office, and there is no doubt in my mind that all the dev's could fit in here and have a beer with me.  I've no idea how many VL users there are, many thousands at least, but the VL user, VL dev ratio is way out of balance.

I think M0E has a great idea, he's done his time as with the rest of the devs.  But we really shouldn't consider this project unless people step up to the plate and are willing to pitch in.  There is plenty to be done that doesn't involve coding and such.

[nubcnubdo]

DeLi, DSL, Puppy, NimbleX, NimbleXsub100, Custom NimbleX 2, Wolvix Cub

NimbleX is a pretty outstanding effort so if you're interested in Vector Lite, you might wanna check out NimbleX. If we do a lite version of VL intended for older machines, maybe we could improve dialup support/modem detection.

[saulgoode]

Correct me if my line of thinking is impractical or if I am missing something.

Instead of having a core "vector-essential" set of packages based on Standard, this core base would need to be dictated by the more minimal demands of the light version. Conceptually, the Standard installation would first extract the vector-essential bulk of the light version and then extract a bulk which brings that base up to the level of a "Standard-essential" -- though the installer itself could use a merged version of those two bulks.

Packages which might be needed for Standard but excluded from the light version would need slack-depends files created if they don't already exist. It would not be acceptable to have different packages specific to the different versions of VL, so any packages available for the light version would be the same package (if) included in the Standard version's bulk.

Thus releasing a light version would have a developmental impact on Standard; it could not be developed completely independent of the other Vector versions. However, for the most part this impact would consist of regenerating (adding dependencies) the packages in "Standard-essential" which are not in the core light version. The lion's share of the burden of regenerating these new packages should rightly fall on the light version developers, rather than on the standard developers, even though these packages may not be included in the light version. Nonetheless, usage, support, and maintenance of these packages would eventually fall on the entire development team.

As stated earlier, I would be willing to contribute towards such a release -- and perhaps MOE-lnx and Lagnanon would be interested -- but the ultimate decision lies with the Vector management and developers.

[rbistolfi]

I think you are right, a packaging station must be always the smallest VL.

[uelsk8s]

We think that this is a good idea.

To move foreward with this we will need to have a package list to start from.
The list of packages on the ISO for 5.9 is here: http://numsum.com/spreadsheet/show/57314
Saulgoode, if you (or anyone else) want to make an account at numsum.com I will make a new copy of this spreadsheet and you can start marking out packages to form a new smaller list.
Once this is done we can move on from there.

Uelsk8s

[Dweeberkitty]

Cool spreadsheet. I could use something like that for the MMBD. How did you make it? Do you have scripts?