DSL-N :: Move from DSL-N rc4 to DSL-N 1.0

I know that DSL is somewhat "developer constrained" and really appreciate the hard work and commitment that goes into it.

I can't help thinking that all DSL-N needs for it to take a leap forwards in terms of popularity is for dpkg/apt-get functionality to be added (as opposed to the current work-around) and for some tidying up to allow applications to be compiled under DSL-N from tar.gz downloads.

If the above were done, people would be able to build their own extensions easily and take advantage of the additional 2.6.x functionality.

I have a need for both flavours - DSL (frugal hd install on an ancient desktop) and DSL-N (usb boot on a relatively modern laptop).

Would it be possible to take a break from the hectic pace of development of DSL at 3.3 and move DSL-N from rc4 to 1.0?

Quote

I have a need for both flavours - DSL (frugal hd install on an ancient desktop) and DSL-N (usb boot on a relatively modern laptop).

Me, too! I have great respect for the tremendous technology that goes into DSL, but it may become a victim of its own success.

Moore's Law mandates that the natural install base for a 50 MB distro is gradually shrinking, while the base for a smallish distro like DSL-n, Puppy, Arch, eLive, Zenwalk, Mint, and Vector is growing. If you read the reader comments at Distro Watch Weekly, "small system" people are looking for friendliness and configurability.

Quote

I can't help thinking that all DSL-N needs for it to take a leap forwards in terms of popularity is for dpkg/apt-get functionality

DSL is technically brilliant, but the other small distros are far easier for *nix newbies. Most users don't want to have to deal with configuring backup/restore and persistent directories, let alone .uci & .unc. Great technology of speed & space efficiency, but not newbie-friendly.

Further, the DSL app repository seriously lags the other distros. While this may (occasionally) be for size reasons, it makes DSL somewhat less competitive when open source apps go through relatively rapid function & security updates. E.g., DSL 3.2 has Firefox 1.06 while current is 2.02. How to install that on a (recommended) Frugal system, should the user have the resources & desire to do so?

DSL has been very successful in the ultra small system area and the system rescue area. Now that Gparted, SystemRescue, and BackTrack are moving into the rescue area, one has to wonder if there will continue to be enough interest in the ultra small distro or whether there needs to be something a bit bigger & friendlier for newbies.

Let me emphasize that this is said with the utmost respect for the authors and the knowledge that I'm in the minority in the DSL user base. I'm concerned that the user base won't grow if the emphasis continues to be primarily on the 50 MB user. Moores law may start passing DSL by.

Quote (dougz @ Mar. 12 2007,17:56)

I have great respect for the tremendous technology that goes into DSL, but it may become a victim of its own success.

I disagree. DSL fits a particular niche that other small distros don't. And never will.

Quote

Moore's Law

...has nothing to do with minimal Linux distributions. It has to do with the pace of bleeding edge technology, specifically the rate at which transistors can be -- and are -- added to circuits over specific periods of time.

Quote

I disagree. Everything has a learning curve. DSL has plenty of documentation available for anyone to understand how to get the most out of it. DSL is also a lot more versatile than Puppy (since you mentioned it) in that DSL can make use of Debian's apt-get, and it also gives the user the option of loading to RAM (Puppy does it by default, doesn't it?). And "newbie-friendly" is pretty subjective. If that includes auto-detection of hardware, DSL is on par or even ahead of Puppy and Vector (I have hardware DSL has detected right off the bat that the latter two took some manual configuring).

Quote

Further, the DSL app repository seriously lags the other distros. While this may (occasionally) be for size reasons, it makes DSL somewhat less competitive when open source apps go through relatively rapid function & security updates. E.g., DSL 3.2 has Firefox 1.06 while current is 2.02. How to install that on a (recommended) Frugal system, should the user have the resources & desire to do so?

They do. Let them develop it themselves if they seek to be on the bleeding edge. This is an area where I wonder what people are smoking. DSL can be extended to include any update any user wants. No? Take a look at which versions of SeaMonkey (almost the most recent) and Opera (latest) are available. Concerned about security? Take a look at the following link and note two things. First, the greatest number of vulnerabilities are with x.0 -- major -- upgrade versions. I don't want 2.0 -- which was the latest available at the time of the last DSL release -- and its bugs on a STABLE (old stable, heh) system. Second, note that the severity of the vulnerabilities increases proportionally by release -- 1.5+ has many more vulnerabilities to patch than 1.0.6, etc.
http://www.mozilla.org/projects/security/known-vulnerabilities.html

Quote

DSL has been very successful in the ultra small system area and the system rescue area. Now that Gparted, SystemRescue, and BackTrack are moving into the rescue area, one has to wonder if there will continue to be enough interest in the ultra small distro or whether there needs to be something a bit bigger & friendlier for newbies.

"Bigger" is a problem for those who want an embedded system to run in Windows or Linux, or who want to boot and run Linux completely with a USB thumbdrive.

Quote

Let me emphasize that this is said with the utmost respect for the authors and the knowledge that I'm in the minority in the DSL user base. I'm concerned that the user base won't grow if the emphasis continues to be primarily on the 50 MB user. Moores law may start passing DSL by.

Moore's Law has nothing to do with this issue.

The number of computers that will reach obsolence or a dead end with respect to XP over the next few years will make them as much candidates for "bloated" distros like Ubuntu, Mandriva, etc., as for smaller ones. Anything capable of running XP comfortably will just as comfortably run Ubuntu.

What you expect is DSL to try to keep up with the Fluxbuntus, UbuntuLites, Vectors, and Puppies. I think DSL has its own niche that the others can't compete in, and that DSL should stick to its guns by cramming as much function into as little space as possible. The user base for DSL isn't necessarily the same drawn to the antialiased eyecandy of something like Fluxbuntu, which is quite similar to DSL-N in its approach to things. DSL is a choice among many -- changing it into just-another-distro with more bloat than function would be a bad thing.

I believe you will see movement after I complete and release DSL v3.3.

To lucky13: I don't think it would be profitable to do a point-by-point rebuttal, particularly where your comments deal with preferences or slight misunderstandings of my points.

Remember: I'm talking about DSL-N, not DSL.

I was trying to point out that the "sweet spot" for any piece of software is influenced by the available hardware. Yes, DSL will fit on smaller boxes than anything else. However, most user's "sweet spot" will inevitably move from smaller to larger hardware as people are willing to seek more ease of use because they have access to larger hardware. RAM == ease of use. As those discarded 256+ MB Win2K and WinXP boxes become available, people expect more. A distro optimized for 50 MB will increasingly be a "niche" distro. A bigger sibling can accomodate additional users who would not be attracted to a tiny distro.

I highly recommend the DWW comments cited in my previous post. Most commenters expect Linux to "just work," like they do an Apple. I'd also highly recommend reading esr's "World Domination" paper. Linux is still a niche OS. DSL, while a superb piece of technology, is a niche Linux. Bigger user base == more donations == more development.

I understand embedded systems. I've programmed them -- 8051 (4K ROM, 128 bytes RAM) cross-assembler and Forth-based tiny systems. Very rare now. The "sweet spot" moved on to the point that people program embedded systems in C and other HLLs now. I programmed to the "bare metal" (no OS, just my "real time" code) while today's developers expect to have an OS and toolkit to work with on modern embedded systems. Different "sweet spot."

Thumbrives are 4GB now! Different "sweet spot!"

Quote

I think DSL has its own niche that the others can't compete in

We agree! My point is that I think that niche will shrink as people's definition of a "small system" changes. My first box was a TRS 80 Model 1, Z80@ 1.77 MHz 4K ROM, 4K RAM, cassette storage. I expect more now. Different "sweet spot."

I still think Juantio's point is correct:

Quote

I can't help thinking that all DSL-N needs for it to take a leap forwards in terms of popularity is for dpkg/apt-get functionality to be added

That allows a user of small-to-moderate skill to extend his/her DSL-N in any way they see fit.

DSL's user base is both an enormous strength (skilled, motivated, self learners) and a weakness (skilled, focused on very small systems - by today's standards). However, I really don't think it is incorrect to say that DSL has achieved its extraordinary efficiency at the price of "ease of use." That is not a tradeoff that most users will make. As Juanito said, DSL-N would be more appealing if it was easier.

Next Page...
original here.