Recently in Puppy Category

I've been thinking about building my own very small machine around the dual-core Intel Atom processor with Nvidia graphics. Yes, I know that Nvidia is freedom-hating and all, but I think that for the small form factors such as Mini-ITX, Intel and Nvidia are heading in the right direction when it comes to compactness, power consumption and graphical sophistication.

I usually begin my search with my favorite Mini-ITX vendor, Logic Supply, but I have also begun looking at pre-assembled systems that ship with Linux. Both ZaReason and System 76 are building small boxes around the Intel Atom/Nvidia platform, some single core, others dual core — and I do recommend the latter.

The one stopping point for me, other than money, is that I'm not sure whether or not these pre-built boxes have CPU fans or use passive cooling from massive heatsinks. For years now I've been leaning toward machines with no spinning fans either in the box itself (on the CPU or elsewhere) or the power supply. With Logic Supply I can easily make this happen.

At ZaReason, the Ion Breeze 4220, starting at $399 for single-core, offers a variety of options, including the above-mentioned dual-core Ion CPU. I don't know if Earl, the ultra-accommodating chief technology officer at ZaReason, is offering the option of a fanless motherboard — I'll ask him.

System 76 offers its Meerkat Ion NetTop with dual-core Ion starting at $359.

One thing that ZaReason offers in the Ion Breeze that I like is an optional external fanless power supply.

I've been running my converted Maxspeed Maxterm thin client as a standalone Linux/BSD box almost since the beginning of my foray into open-source operating systems, with only a single fan blowing across the Mini-ITX motherboard and its heat-pipe-cooled CPU. The fan doesn't work when the box is upright, so for all intents and purposes this is a fanless computer, and I've never had a problem with thermal issues — in fact, it runs quite cool, if not quickly with its VIA C3 Samuel processor (that's supposed to be a 1 GHz model but for some reason only runs at 500 MHz), maximum of 256 MB RAM and woeful sound and video chips.

Right now the Maxspeed is running Debian Lenny from an 8 GB CF card inserted in the thin client's built-in CF-to-IDE interface. Yep, no spinning hard drives either.

System 76 does offer solid-state drives on the Meerkat Ion, starting at $110 extra for a 40 GB Intel drive.

If the Intel Atom Ion processor isn't what you're looking for, both System 76 and ZaReason have plenty of other desktop, laptop and server machines to look at.

The best thing about buying a computer from a shop that ships with Linux (in the case of these two retailers, Ubuntu) is that your hardware is pretty much guaranteed to work. You'll have audio, video, suspend/resume, all that stuff that sometimes is hard to get straight on the box that shipped to you with Windows.

In the times I've spoken with ZaReason's Earl, and the company will build, test and ship pretty much anything you want. They specialize in Ubuntu, but you can ask for a box to be loaded with Debian or CentOS, and I believe they'll do it.

Do ZaReason and System 76 charge more than your standard computer seller? Probably. You can't get the kind of bottom-of-the-barrel deals that are offered on the cover of the Office Depot circular, but those machines often do have bits of hardware that you'll tear your virtual hair out to get working properly.

When you get a machine from a company that specializes in Linux, not only will everything work, but you'll get support that will help you clear up any issues.

And for many people — and I'm getting more like this myself with less time available for banging-my-head-against-the-wall tinkering — it's worth a little extra money for somebody else to have figured out all the issues, or in the case of these companies, to choose hardware components that work well with free, open-source operating systems from the start.

And even if you are a tinkerer, chances are it ZaReason or System 76 have built you a machine, it won't just work well in Ubuntu but will be a great platform for other Linux distros you might want to run.

Not wanting to leave out BSD, you can get a pre-built and -loaded PC-BSD (based on FreeBSD) laptop as well as two workstations (prices unknown) from IXsystems, the company behind PC-BSD. They seem to specialize in selling servers running FreeBSD and ask that interested buyers request a quote to receive pricing info. They're also offering CD and DVD sets of FreeBSD 8.0 if you don't want to bother downloading the ISOs and burning your own discs.

Not to go off on a tangent or anything, I've been giving FreeBSD a lot more thought lately. I've run OpenBSD on the desktop as my primary system for about six months, and I'm considering FreeBSD instead for a future test for the following reasons:

• Easier upgrades and much longer cycle
• More focus on desktop users with hopefully better (and more meta-style) packages for things like GNOME
• Flash 9 and possibly Flash 10 support through the Linux compatibility layer
• Better performance
• I really don't need it for architectures other than Intel/AMD (although PowerPC and SPARC 64 are available; side note — on the various pages emanating from its platforms page, FreeBSD offers not only official manuals from the makers of the hardware in question but also links to other BSDs that run on the architecture. A very nice touch, I think)
• Community that actually cares about end users who aren't developers

I need to try some live images of recent FreeBSD/PC-BSD releases. (Is PC-BSD a live CD yet? I haven't kept up, but I did utilize the live environment of DesktopBSD back when I was testing it).

I never did the full review I promised of Dru Lavigne's excellent "The Best of FreeBSD Basics" book, but I find it to be an excellent reference for the FreeBSD and PC-BSD user. Dru is one of the best writers around in the Unix community, and even if you don't run BSD you can learn a lot about using Unix/Linux from this book. I got a whole lot about the shell, file permissions and other Unix sys-admin tasks, from "Basics," just as Michael Lucas' discussion of sudo in "Absolute OpenBSD" makes that now-way-out-of-date book extremely relevant and useful for anybody running any kind of Unix/Linux today who wants to make the most of sudo in their own environment (and especially on the server).

On the same tangentially arrived-at topic, Dru Lavigne's latest book, "Beginning PC-BSD: Frugal Unix for Power Users," is slated to be released three days from now. If past work is any indication, this will be an excellent book for anybody contemplating the use of PC-BSD.

I'd rather Dru write a book on using FreeBSD on the desktop — not necessarily PC-BSD but building out a FreeBSD-based desktop through ports or packages — but I can understand her focusing on PC-BSD given that the iXSystems-led project is a lot closer to what Linux users are used to.

I've been meaning to do this for ages, and I finally installed Puppy Linux on a bootable USB drive.

I went whole hog and used a 128 MB stick. Yep, that's it. I have a huge 20 MB left for storage. Now that I know this works (at least on my Dell, the only box to which I have access that also allows booting via USB) I'll get a bigger stick and actually have some room to, as they say, maneuver.

Doing the install was easy. I booted Puppy 4.1.2 from a CD I had previously burned (I know Puppy is up to 4.2 ... I'll have to try it). Then I used the menu to install to USB. The only thing I did that wasn't a default was selecting mbr.bin as the boot method. It works.

Things I was pleased about in Puppy 4.1.2, which blazes on a 3 GHz Pentium 4 with 512 MB of RAM, include Abiword with working spell-check (never did get that together in OpenBSD; they should package it to work right ... but I digress), and the inclusion of apps that make this a great working environment.

I already loaded a couple of IMAP accounts into Seamonkey's mail client, and if I did have the disk space, I could use gFTP to load all my stuff onto the USB stick.

Considering that these sticks are pretty much laying around and can be had for free, this is a great way to put together a cloud-computing environment if you have all of your mail and files in something like Google Docs and Gmail. Who needs to wait for Chrome OS?

As an experiment, I decided to bring my Evolutionary Computing presentation on making the journey into free, open-source software — a slide show originally created in OpenOffice Impress 2.4 — into Google Docs, which happens to have a presentation app in addition to the better-known Docs and Spreadsheets components.

I revised the presentation — taking some things out, adding others and providing some updates on what I'm doing — and output it as a PDF.

Download that PDF for your reading pleasure by clicking on the image above or the link below:

Evolutionary Computing (revised July 2009)

Interesting note: I believe that no previous entry on this blog has been filed under so many categories. (And I've been considering dumping Categories entirely and just using tags ...)

It's been many months since I last used Puppy Linux. I bet more than a year has passed since I seriously ran Puppy, still one of the best Unix-like distributions/projects for older, underpowered computers.

I decided tonight to break out the 1999 Compaq Armada 7770dmt (233 MHz Pentium II MMX processor, 144 MB RAM), which has OpenBSD 4.2 on the 3 GB hard drive (yes, I know 4.5 is out, and yes I do have the CD set, and yes, I'll probably reinstall) and two pup_save files in its 0.5 GB Linux partition.

During my extensive tests of operating systems on this platform, I ended up running the aforementioned OpenBSD 4.2 and Puppy 2.13, the latter from live CD.

I'm in Puppy 2.13 right now. I know it's old. I know Puppy 4.something is out now and that the project is in some sort of turmoil.

I decided to get deeper into Puppy 4.1.2 on my Toshiba Satellite 1100-S101 laptop.

I'm always looking for platforms on which I can do all my Daily News-related work, which means I need the Java runtime and Flash video.

Well, there is a Java package for Puppy. I'm surprised Java isn't part of the base install, but it appears not. I installed the package, and I even brought in the Opera Web browser to augment Seamonkey.

Both browsers are performing well, but for some reason Flash doesn't work in either. I distinctly remember Flash working in all of the Puppy 2 and 3 releases I've used previously, and now I'm left wondering what happened.

Also, Java did NOT work in either browser, so easy use of the LogMeIn remote-desktop service is not something happening in Puppy. I'm getting to the point where I'll need to bit the proverbial bullet and install Java from source in OpenBSD on this laptop so I can get that functionality. I can live without Flash (and the Flash I do have in i386 OpenBSD via Opera is marginal at best; it works in YouTube but not in Brightcove). I can sort of live without Java.

But it's better for the work that I do to have both of these things working well.

Also, I was surprised to see not Pidgin or Gaim as the IM client in Puppy but something I'd never heard of. Pidgin is available as a package, so that's not such a problem.

The end result is that while Puppy 4.1.2. runs quite well at first blush, I need to look closer at why I was so unsuccessful at getting Flash and Java to work. It should be easier than this.

And while Flash remains somewhat of a problem in OpenBSD (I probably need to be running an up-to-date Linux such as Ubuntu, Debian, Fedora, Slackware, Zenwalk ... take your pick) I'll probably stick with it for the time being as my primary OS.

As of today, here are all the machines I use and what they run:

At the office:

Work box:
Dell Optiplex GX520
Pentium 4 (3 GHz)
512 MB RAM
Windows XP SP2

The Debian Mac:
Power Macintosh G4
466MHz single PowerPC processor
384 MB RAM
Debian Etch

The Self-Reliant Thin Client:
Maxspeed Maxterm 5300(??) thin client
VIA C3 Samuel (1 GHz, running at 500 MHz for some reason)
256 MB RAM
8 GB Transcend Compact Flash module as boot drive
1 GB USB flash drive for backup
Debian Etch

At home:

iBook G4
1 GHz CPU
384 MB RAM
120 GB Fujitsu hard drive (replaced by me in a 3-hour odyssey)
OS X 10.3

This Old PC:
Pentium II MMX (333 MHz)
256 MB RAM
10 GB hard drive
Windows 2000 (I haven't booted this or connected it to the Internet in over a year)

The $0 Laptop:
Gateway Solo 1450
Mobile Celeron (1.3 GHz)
1 GB RAM
30 GB Toshiba hard drive
Ubuntu 8.04 LTS, Debian Lenny, Puppy 3.01

The $15 Laptop:
Compaq Armada 7770dmt
Pentium II MMX (233 MHz)
144 MB RAM
3 GB IBM hard drive
OpenBSD 4.2

I have quite a few machines in various states of repair that I might resurrect over the next year if and when I get the time, but this is what I have right now. With the exception of the white-box This Old PC, all of these get fairly regular use.

I have my Self-Reliant Thin Client running Debian Etch turned on all of the time. I haven't been able to find power-usage specs for the Maxspeed Maxterm (it could be a 5300, but there are no model numbers on the box), but with no moving parts, a Mini-ITX-size motherboard, Mini-ITX-type fanless power supply and fanless VIA C3 Samuel CPU, as well as non-working case fan (except when tilting said case at a 45-degree angle) and a Compact Flash chip instead of a spinning hard drive and no optical drive, the thing is totally silent and must be fairly sparing on electricity use.

I don't think I even moved the mouse yesterday, but today when I brought it out of screen-saver mode, there were three updates to Debian Etch:

dbus
dbus-1-utils
libdbus-1-3

Thus far, the 8 GB Transcend Ultra Speed 133x Compact Flash is performing quite well, meaning it hasn't died.

The last time I killed a CF chip, a 1 GB Transcend, I think the premature death occurred due to inserting or removing the module while it was mounted.

Since in this case I have the Self-Reliant Thin Client sealed, that CF chip is staying in there and won't be plugged and unplugged all that often.

That might stay true, but I want to get more CF chips and load different OSes on them. Then I could remove the cover to the CF-to-IDE board in the thin client and pop in and out different CF cards with totally different configurations.

Some of the CFs I'd want to do:

• Puppy Linux (could be a much smaller CF due to the nature of the Puppy distro and its "frugal" install)
• OpenBSD (I'm anxious to see how easy/difficult it would be to install to CF)
• Wolvix (which also offers a "frugal" install, though I'd chose a "traditional" hard drive install so I could use slapt-get/Gslapt to update the box)

Not having an optical drive hooked up makes the "preparation" of CF cards on the Self-Reliant Thin Client difficult. To install a new OS, I'd have to:

• Remove eight screws to open the case
• Remove the CF card cover
• Remove current CF card and plug in new one
• Unplug the CF board's IDE cable from both the CF board and the motherboard
• Plug in a standard IDE hard-drive cable into the CF board on one end, the motherboard on the other
• Plug CD-ROM drive into "middle" of IDE cable
• Plug hard-drive-style power cable (the thin client has one, even though it doesn't need it for its intended purpose)
• Install new distro (and probably do more than one so I don't have to repeat this procedure)
• Test new distro
• Remove IDE hard drive cable
• Plug CF board's IDE cable into CF board and motherboard
• Replace case cover

I could leave the CF board/adapter's cover off if I wanted to do a lot of swapping of CF cards. It would be a very easy plug-and-play way to swap distros, that's for sure.

And I could keep the current 1 GB USB flash drive plugged in for backups of the various systems. That would also facilitate file-sharing between the OSes on the multiple CF cards.

Not that anything approaching brain-surgery-level thinking was in any way involved here, but I figured out why and how it's easy to get paragraphs to automatically indent when writing in the Geany text editor.

First of all, it's not called automatic tabbing or paragraph inentation. The correct term for what I'm enjoying so much is auto-indentation and it can be turned on and off under the Document menu in Geany. The defaults for auto-indentation can also be set in the Edit menu under Preferences--Editor.

When writing for print, where I don't need — and can't stand — having two returns between paragraphs. After transferring the file from this laptop to my newspaper's print publishing system, those double-returns demand that I delete one of them. That's because in most non-Web publishing, indented first lines make paragraphs distinct from one another, not extra linefeeds.

So having the indents on the first line of every paragraph helps me seen where each paragraph begins.

I know that programmers use indents to help structure their code. But when something so right for coding in C also helps hacks like me, making traditional word processing applications less needed, everybody wins.

(This post was originally written on May 22, 2008; since that time, I've added the RAM, and it does indeed make a difference. It's still not easy to live with 144 MB of RAM and 233 MHz of CPU, but it's easier than having less than half of that M. What I can say is that 500 MHz of CPU and 256 MB of RAM is positively picnic-ish. Also, I finally did the OpenBSD 4.2-to-4.3 upgrade on the VIA box. It wasn't easy, but I did get it done.)

If the question is "how low can you go" in terms of computer memory, it's all about applications.

If you stayed in the Linux console and never ran X, just about anybody could be happy with 32 MB of RAM. It might be hard to actually run Linux or a BSD in 16 MB, but I've heard of Linux distributions that will do it, Damn Small Linux, Tom's RtBt (is that the right spelling?) and DeLi Linux among them.

But as much as the hard-core users talk about how they stay at the command line all the time, it's hard to get much done strictly in a console when you're a regular person. Sure you can use Lynx for text-only Web browsing, you can set up Mutt (and Postfix/Sendmail/msmtp/esmtp, Procmail and whatever other helper apps are needed) with highly customized configuration files designed to handle and filter multiple mail accounts, use Vi or Emacs for text editing and all that.

But the bottom line for me is that I need a Web browser. A "real" Web browser, something that works with Movable Type and Google Docs, and that pretty much means Firefox or some Iceweaselish derivative.

I don't tend to use OpenOffice very much (although it runs better in Debian with 64 MB that you'd think), I barely even use AbiWord these days. I'm not saying that I won't need OpenOffice in the future, but at present I'm most comfortable using various X text editors, including Geany in most Linuxes and BSDs, Gedit when I'm in GNOME, and Google Docs half the time just for the easy portability of my copy.

And while Geany doesn't load super quickly from a "traditionally" installed distribution (but is quite quick when loaded into memory as it is in Puppy Linux, once it's loaded it runs very well indeed.

And the Dillo Web browser -- which looks better in its OpenBSD incarnation than it does anywhere else -- performs quite well in 64 MB of RAM. The only problem is that Dillo can't do everything I need to do on the Web. At least the Dillo in Puppy and DSL has https support. That's not turned on in OpenBSD, and the app needs to be recompiled to add it. I can manage to turn on cookies in OpenBSD, which helps me with some sites, but for anything remotely complicated, Firefox is essential.

And while Firefox will run in 64 MB of RAM, it does so very poorly. There just isn't enough memory to keep the program from swapping to the drive incessantly whenever doing just about anything.

In this very 64 MB, I've run just about everything that will load on this Compaq laptop: Puppy, DSL, Debian (the Xfce install, plus a "standard" install with Fluxbox), Slackware (without KDE) and OpenBSD.

Truth be told, Almost all of these OSes run just about the same. Damn Small Linux has a bit of an edge, and if DSL 4.3 ran as well as 4.0, its inclusion of Firefox 2 would put it over the top. As it is, I've lost my desktop wallpaper, and I can't figure out how to display the menu in Fluxbox (even though I prefer to run JWM).

Puppy definitely needs more memory, especially to run the Mozilla-derived Seamonkey Web suite.

Debian Etch was OK. While the Xfce install is odd in many ways, as I say, I was surprised to see OpenOffice run at all -- and not too badly at that. Iceweasel was, again, an exercise in frustration. But Debian remains a distinct possibility for this machine.

It's main OS for awhile has been OpenBSD, with a partition set aside for the Linux files generated by the Puppy and DSL live CDs.

OpenBSD runs pretty well, but as I said, Firefox remains an issue.

The question: Will things improve with the boost of RAM from 64 MB to the Compaq Armada 7770dmt's maximum 144 MB? From my past experience, I know that Puppy can run in 128 MB if you have swap space, and DSL is certainly comfortable with 128 MB.

To answer the question, I could reduce the memory in my Via test box from 256 MB to 128 MB and see how OpenBSD (now version 4.3) runs in that configuration. But I'd have to pull the cover from my converted thin client and find a 128 MB SIMM. I've probably got one ... somewhere.

Better to just wait for my Compaq memory to come in the mail (luckily it's cheap).

I've know for awhile that 256 MB is a significant sweet spot for Linux, but I'd love for 144 MB to be just sweet enough to give this laptop a new lease on open-source life.

And while I managed to upgrade my VIA box from OpenBSD 4.2 to 4.3, it takes a lot more work than a simple apt-get, and I'm reluctant to do it

I've been able to have OpenBSD's /etc/fstab automatically mount the ext2 filesystem on my Compaq Armada 7770dmt's hard drive with no difficulty lately, but every couple of days or so I get a message while booting OpenBSD that says the Linux filesystem is not clean and that I should run fsck on it.

I then boot Puppy Linux 2.13, run e2fsck on the partition, the errors are cleared up, and all is well until a few more days pass.

I haven't lost any data, but I'm going to do a few experiments.

First, I added noauto to the /etc/fstab line so the Linux filesystem will not be automatically mounted. Then I'm going to run Puppy for a few days and check the filesystem with e2fsck.

It could be that the errors are coming from Puppy alone. I think that's unlikely, but it is a possibility.

Then I'll experiment with manually mounting (with mount) and unmounting (with umount) the Linux filesystem while in OpenBSD.

That way I can see whether or not automounting and unmounting the ext2 filesystem in OpenBSD is what's causing the problem.

Hours later: Looks like OpenBSD is NOT responsible. I ran Puppy totally in RAM (using the puppy pfix=ram boot parameter), than ran e2fsck to clean up the filesystem on my ext2 partition. Then I ran Puppy the "normal" way, in which the system mounts the partition to access the pup_save file. I then rebooted and once again ran Puppy without mounting the partition. At no time did I boot OpenBSD or mount the filesystem in that OS.

Once I was back in Puppy, running pfix=ram to keep the partition unmounted, I ran e2fsck and got this message:

/dev/hda3 was not cleanly unmounted, check forced.

I had one more test to do.

Now that I had run e2fsck on the ext2 filesystem, I needed to boot OpenBSD, mount the filesystem, write a file to it, then unmount it. After that, it would be time to boot Puppy Linux again, using the pfix=ram boot parameter again so as not to mount the filesystem in Linux, and then run e2fsck again to check the filesystem and see if mounting, writing to and then unmounting it caused any errors.

So I booted into OpenBSD 4.2, mounted the ext2 filesystem, modified a few files, added a few, then unmounted it. I rebooted and did the same thing again.

Then I booted into Puppy, again with the pfix=ram boot parameter so as not to mount the Linux partition.

I ran e2fsck. After two boots of OpenBSD, during which I modified files in the Linux filesystem both times, there were no errors in the ext2 filesystem.

I said it was "unlikely," but in fact it's Puppy Linux, NOT OpenBSD that is not "cleanly" unmounting the Linux filesystem. I truly expected it to be the other way around.

I'll have to test this with Damn Small Linux, Wolvix and maybe even Slitaz to see if this is a Linux problem, or just a Puppy (or Puppy 2.13, to be more specific) problem. But right now, OpenBSD has absolutely nothing to do with it.

Mounting the filesystem in:

Damn Small Linux 4.3 caused no errors