| DistroWatch Weekly
1 • Friggin *buntu and their stupid ideas (by Heavensrevenge on 2013-09-09 09:26:00 GMT from Canada) |
I respect you all, I love linux, but what the HELL is Cononical thinking??? Their as bad as Windows 8 at this point, siloing themselves in and screwing all the other linux distros "because they can". Sure, if your a *buntu addict and your already running *buntu sure, you are incompetent enough to not care.
Needing this "Ubuntu SDK" to build a program which is self-contained with 0 dependency checking for absolutely 0 reason while theres many perfectly good package systems is ludicrous.
Why do the idiots at Cononical think they need to just branch out and do EVERYTHING differently and in their own little world SEPARATE from the environment they were born in causing even more fragmentation in our beloved linux world?
They use a different binary and proprietary patched kernel, their own display server(Mir), their own desktop environment(Unity) and now an entirely and INCOMPATIBLE new packaging format that no other system will ever care about and can't use because it doesn't tie in to anything on the system AT ALL and is only a hashmap of software -> version? They need to get pimp-slapped or go bankrupt to get the darn clue that they are harming our ecosystem as a whole for their actions.
These packages seem like BSD PBI's. But PBI's exist for a good reason, in BSD things in general are built using the ports tree from source and binary distribution wasn't the norm so PBI's stepped up and did the job quite nicely. But trying to just be douchebags and do something because you can with 0 technical reason other than their too stupid to use a .deb or .rpm format specification is just beyond me.
Sticking a fork in an electrical socket "because you can" is not very reasonable, is not because it's a good or bad idea, but because there's nothing stopping you(them) from doing so.
Thank god for Intel telling them to screw off with their XMir code for inclusion in the Intel GPU driver included in the Kernel.
2 • Lubuntu (by ange on 2013-09-09 09:31:43 GMT from Hungary)
Another "light" distro, with installer iso above 1GB size.
Lubuntu in itself is better as base for building lightweight desktop IMHO. No need for unnecessary softwares. Anybody knows a Lubuntu-based REALLY lightweight distro?...
3 • #1 (by zykoda on 2013-09-09 09:43:53 GMT from United Kingdom)
Give them a rope and let them play! After all, */Linux is is far from perfect with ageing paradigms. Maybe something useful will surface...... eventually.
4 • Mir support (by Ian on 2013-09-09 10:00:38 GMT from Ireland)
I read (somewhere) yesterday that AMD/ATI aren't going to support Wayland, so there is going to be a split developing in the hardware world. Maybe they are plotting the demise of all Linux or open source operating systems between them.....I love a good conspiracy theory.
5 • LXLE (by kc1di on 2013-09-09 10:24:46 GMT from United States)
Thank Jesse for reviewing this great little project. LXLE has become my standard install on all my older equipment and as you say for the most part it has work flawlessly out of the box.
I like Lubuntu too but they never offer the Long term support version so when I found that LXLE was I had to try it and been hooked on it ever since. The developer/s have done a great job of putting together a solid distro and hope lots of other find it too.
6 • Xorg (by bam on 2013-09-09 10:51:22 GMT from United States)
No senseless rant. Click packages=click 'n' run.
Xorg's days are numbered., X11 acts largely as communications protocol between the client and the window manager. This changed role works, but can be awkward, and frequently results “in visual glitches, such as the old juddering [shaking] resizes or flashes of a gray-and-white window before it fills itself in.” Although these glitches can be patched, the patches further complicate the structure and the necessary maintenance. Hence you have Mir on one hand and Wayland on the other.
Intel will not support Mir at this time. And AMD/ATI will not support Wayland. No need to rage against Canonical.
7 • @2 - Lightweight distro based on Lubuntu (by Uncle Slacky on 2013-09-09 10:54:22 GMT from France)
Try Peppermint OS ( http://peppermintos.com ) - the ISOs are less than 600Mb in size.
It's touted as a "cloud" OS, but has access to the usual Ubuntu repos, so you could install as much as you want locally after OS installation.
8 • lightweight doesn't necessarily mean file size (by Ronnie on 2013-09-09 11:00:43 GMT from United States)
Lightweight is a term that gets thrown around way too much. It can mean so many things. A lightweight system to me isn't about the ISO file size as much as it is about an OS that uses few resources to perform its duty. Many will have varying definitions of what lightweight means, some think that everything from top to bottom, all programs, iso size , everything. I tend to lean on the side of the OS.
9 • Simplice-5 (by Joe on 2013-09-09 11:02:14 GMT from Mexico)
What a big surprise! . . . this five version has a marvelous combination: Very fast by openbox, very solid by its debian base, Software and hardware support updated by the SID version of debian; the italian design in all its power, and th e inspiration given by the beer! . .u very much, . Congrats and Thanks, I allready instaled in my vaio netbook and in my dsktops in home and office...
10 • Simplice-5 addendum (by Joe on 2013-09-09 11:30:21 GMT from Mexico)
in 2o years (yes i had 20 years using linux) i had used hundrets (yes hundreds ) of distros.... Simplice-5 is the FASTEST distro that I was used. in the last months I used crunchbang-11 whith 9 seconds at booting in my Vaio core i5, Now I obtain a booting of 7 secs with Simplice-5... I am so surprised..!
11 • Lxle (by UUUnicorn on 2013-09-09 12:01:27 GMT from United States)
A question I did have about Lxle is, aren't there two different 32-bit ISOs of this distribution? I thought that there was one that was PAE, and one that was not PAE.
I do already know that there is a 64-bit ISO, but I am not addressing this one here now.
12 • Good ole Monday morning. (by LinuxMan on 2013-09-09 12:16:51 GMT from United States)
A lot of good info this Monday on Distrowatch. Well it seems that Intel's folks aren't going to support Mir and AMD/ATI's folks are not going to support Wayland. I don't really believe that is going to change things much. X is on the way out, as it should be, and so it seems that only nVidia remains to tell which way they are going to go. Of course I don't believe Intel or AMD/ATI or their claims. Intel drivers have never been that big a deal and patching will be done. When everyone else goes to Wayland they will have to deal with AMD/ATI. It really surprises me that so many people still do not know what Canonical is trying to do with their display server. I can tell that by the first comment today. The new display server is needed. The new package system that Canonical is developing along with Mir are necessary steps for a unified system. I don't understand why people get so upset unless they just don't understand. No ecosystems are being harmed. As with all open source projects and developers, ones not dependent on Ubuntu will go on their merry way and life goes on. Others will evolve.
Nice review on LXLE. It makes a person want to download an iso to see what goodies it has. Semplice Linux 5 really looks interesting. I will try it out. Does anyone know what is the most difficult system to disable secure boot? I've heard some really good horror stories about HP but not many more.
13 • @2 Trisquel Mini (by Frank on 2013-09-09 12:41:12 GMT from Germany)
You might give Trisquel 6.0 Mini a try.
Free, very beautiful, lightweight, small 500mb iso and very fast.
14 • LXLE - lite/Light/Lightweight/WHY (by Sondar on 2013-09-09 13:32:01 GMT from United Kingdom)
Very comprehensive and useful review, Jesse. Sadly I cannot agree with many of your findings. On the other hand, I am in complete agreement with the above comments regarding size. This is a bloated offering on every count, not just size. It is painfully slow to boot even on respectable machinery, despite it's claims. Furthermore, I was underwhelmed by the modest offerings in such a grossly over-endowed distro. Geoff, rg and the gang achieve massively more power, speed and productivity in their ~3650Mb Carolina Xfce offering, for example. There are other Xfce ports, too, offering a lot more for less.
15 • 14 (by Sondar on 2013-09-09 13:42:58 GMT from United Kingdom)
365Mb - bloat abounds, allegedly!
16 • LXLE (by Hollandhook on 2013-09-09 14:01:34 GMT from Mexico)
I like it. Comfortable, quick and smooth.
17 • @ #2 (by Robert on 2013-09-09 14:02:40 GMT from United States)
I'm running Zorin Lite 6.2. It is based on Lubuntu and the ISO is under 700MB. It doesn't come with a lot of "junk" and runs better than Lubuntu. Artyom Zorin knows what he's doning!
(The latest Zorin Lite is 7.0, and weighs in at 786MB.)
18 • slack-based (by EH on 2013-09-09 15:12:33 GMT from Philippines)
Try porteus...its fast and easy..
19 • Also @2 (by James on 2013-09-09 15:28:05 GMT from United States)
If you want a system that's light on applications and RAM usage, the question is how much time and work you're willing to invest. Distros like Arch (or its fully free cousin Parabola) let you start out with "just barely enough" and add only what you want, which is fantastic if the setup time isn't an issue. I'm also fond of the Debian netinstall image.
A warning, though, a taste of minimalism does change your expectations. I really don't like running distros that use more than 300 MB of RAM when sitting idle, these days. That means no big desktop environments, no flashy effects, and certainly no system services that I'm not using. ;)
20 • @5 (by jaws222 on 2013-09-09 15:46:11 GMT from United States)
I agree. I have LXLE in a virtualbox with 2GB of RAM and it runs really well. As for the different desktop environments they do pretty much look the same, but they are simple and they work.
21 • LFS 7.4 (by jaws222 on 2013-09-09 15:48:40 GMT from United States)
Anyone familiar with LFS? I'm curious as to how long it would take to build the OS for someone with basic to intermediate linux skills.
22 • @10 (by jaws222 on 2013-09-09 15:55:06 GMT from United States)
If you like quick boots try Manjaro. Boots in 5 secs on my laptop.
23 • Semplice Linux 5 32 in Virtual Box Win 7 32 (by capricornus on 2013-09-09 16:30:16 GMT from Belgium)
installation crashes early in the process of installing the Live failsafe versio. A pitty.
24 • nOS (by Niki Kovacs on 2013-09-09 17:37:40 GMT from France)
I took a peek at the nOS project page. Correct me if I'm wrong, but as far as I understand, nOS is basically a stock Kubuntu republished by a 13-year-old who wants to earn money with ads.
25 • @21 (by Niki Kovacs on 2013-09-09 17:41:07 GMT from France)
Basic skills : forget LFS and come back in a couple of years.
Intermediate skills : give it about a week until you get a working minimal system with a kernel, some basic libraries and a shell.
26 • Lightweight means.. (by Chanath on 2013-09-09 18:06:06 GMT from Sri Lanka)
Lightweight means Puppy Linux, what else? Everything you want and bit more.
27 • A tale of two operating systems (by Carlos on 2013-09-09 18:51:33 GMT from Guatemala)
Another option is that you buy a System76 laptop, they use Ubuntu as their main distro but you could install any other Linux you want, and Windows if you want.
28 • LXLE (by Bob on 2013-09-09 20:03:43 GMT from Austria)
Really liked LXLE at the beginning, because a few things looked smarter than in the majority of distros. But it eventually hung and forced me to press the reset button. Now I'm almost done wasting my time by checking out the smaller distros. Just waiting to see if openSUSE does it right this time (13.1) ...
29 • @21 & @25 (by SAL-e on 2013-09-09 20:46:00 GMT from United States)
It depends. I consider myself to be between basic and intermediate Linux skills and managed to compile LFS in 2 working 10h days in side VM running on white-box VMware ESXi server. I learn quite a bit about boot process and building blocks. But before you begin with LFS you should first understand compile/link process and read a bit about MAKE tools. You should read about the kernel - something like: "Linux Kernel in a Nutshell" by Greg Kroah-Hartman Read carefully and if you don't quite understand the instruction stop and ask or read more about it. Don't rush because error in early step can 'kill you' in much later and waste a lot of time.
30 • Lightweight LXDE (by MikeF on 2013-09-09 22:48:27 GMT from United States)
@2 and others, if installing from the command line is acceptable you can start with the Ubuntu 'mini' iso (~35 MB) of your choice. After the base system is ready just install the packages lxde-core and lubuntu-default-settings. It will take a little longer since all dependent packages have to be downloaded. You Will have to know the package names of all other apps that you wish to add - I recommend synaptic for browsing and installing from the repos.
31 • Another lightweight LXDE distro (by MikeF on 2013-09-09 23:00:03 GMT from United States)
Brain-cramp follow up post.......
I've come to really appreciate TinyCore 'Plus' (a whopping ~66 MB iso) as a starter or rescue system. Unlike Puppy, the default install includes very few apps and a pretty clean desktop.
It doesn't have near the large package repo as Ubuntu but it Does have LXDE and GRUB2 available. Definitely worth a spin if your needs are modest.
32 • Click (by Serge on 2013-09-10 00:17:56 GMT from United States)
Two problems that software vendors face when attempting to distribute software for Linux are that 1) different distributions package different versions of libraries and that 2) libraries frequently break binary compatibility while maintaining API compatibility. This makes it very difficult for software vendors to release binaries of their software that will work across multiple distributions.
In the Linux world, we are used to dealing with open source software. Binary compatibility is not a priority as long as the APIs remain compatible because all that's needed to get a program to work with different versions of its dependent libraries is to recompile that program against these different versions. Distributions' packagers / maintainers do just this and give us ready-to-use binaries that work with whatever versions of the dependencies the distribution carries. This approach works well, as removing the need for maintaining backwards binary compatibility allows library developers and distribution packagers more flexibility.
However, this approach is not an option with proprietary software. Packagers cannot recompile the programs because they do not have access to the source code. Upstream software vendors are stuck trying to figure out a way to make their programs run on multiple distributions and across multiple revisions of the dependencies themselves. Due to the closed nature of such software, the distribution packagers are unable to help. This in turn increases the cost of developing proprietary software for Linux and somewhat hinders the growth of proprietary software.
If I recall correctly, Click solves this problem by packaging private copies of the dependencies within the package itself. Therefore, the end user ends up with larger packages, and many copies of the same libraries, but the vendor doesn't have to worry about whether or not future revisions to library XYZ will break their package.
I'm not a big fan of this solution, but then again, I'm not a big fan of proprietary software in the first place. Personally, I'd advocate for open source solutions to be developed to make the proprietary software in question irrelevant. But if you desire more proprietary software on Linux, Click has a purpose.
Click has a number of other technical innovations, but I can't recall all of the specifics at the moment. I'd advise anyone interested to take the time to read up on it themselves.
33 • @4@12Relax, AMD Catalyst Will Not Support Wayland (by Bbig on 2013-09-10 00:45:51 GMT from Germany)
I say it again: AMD Catalyst Will Not Support Wayland - Catalyst!
Once you are on 3.12 Kernel, you a good - AMDs Open Source Driver is in very good shape!
34 • About Ubuntu's Click package format... (by Explorer09 on 2013-09-10 01:29:26 GMT from Taiwan)
Isn't this something that has been tried before? Autopackage shares the similar goal as Click packages. Canonical seems to be "reinventing the wheel" this time.
35 • nOS, a beautiful start... (by Tim on 2013-09-10 05:31:51 GMT from United States)
I didn't see it as "13-year-old who wants to earn money with ads". Instead, I read that she's offering personalized support @$24.95/yr ~~ quite a bargain. Betcha she'll go all out, spending a lot of time carefully / diligently / thoroughly handling the support requests.
36 • Configuring LXLE's weather application (by eco2geek on 2013-09-10 05:37:53 GMT from United States)
An explanation of how to configure Fast Forecast to show weather for your area (as well as other configuration tweaks) is available on the LXLE forums here --
(Note that you have to change the zip code in two places.)
37 • CLICK packages, reinventing the wheel (by Tim on 2013-09-10 05:43:37 GMT from United States)
Yes, reinventing the wheel... in order to get their "play store" populated. We'll soon see 20 variations of Tetris, 30 fingerpainting apps... the same crop of "apps" which already exist in other (Google, Apple)(Kindle?) ecosystems. Ubu wins when eyeballs browse to find new apps (opportunity to display ads with each "page" viewed)... and devs win when their app is purchased... and user wins with "install once and sync across all (?) my Ubu devices". What's not to like?
php-gtk died on the vine. pyGtk hasn't fared much better. wxWidgets is "soooo last week". Qt-gtk is a maybe someday, eventually proposition... so what's an aspiring "app" developer to do (other than code android "apps")? I'm guessing the ClikUtu SDK will be well-received.
The only other upcoming prospect seems to be google nacl (native client), so I wish Ubu well in their SDK endeavor.
38 • Semplice 5 ~ close, but no cigar (by Hylas on 2013-09-10 07:17:41 GMT from United Kingdom)
Runs nicely in live mode, but installer crashed trying to install to logical partitions.
The partitioner is less intuitive than the native Debian (partman?) IMO, and much less awesome than GParted - sorry Semplice folk :( , I ended up setting up my drive with GParted!
That said, I'd had great times with Semplice 3 in the past, so I persevered, and flawlessly installed it to primary partitions.
I like this release of Semplice. It has great system configuring tools, the ability to turn off/uninstall bluetooth, cups and more with the flick of a switch, newly installed programs immediately appear on the menus, a great gui for controlling graphic effects and compositing too, and a very respectable and well thought out selection of pre-installed software.
I'll keep this on my netbook i think, alongside Crunchbang 11 ~ which is still the fastest and most user-configurable Debian based Openbox distro I have tried yet.
39 • Two operating systems (by Adam Williamson on 2013-09-10 07:41:37 GMT from Hong Kong)
I'm afraid 'just disable Secure Boot and follow any multiboot guide' isn't really the whole story. As I've been saying like a broken record for months, Secure Boot and UEFI are not the same thing. If you disable Secure Boot you still have a UEFI system; if you do UEFI native installs of your desired distributions you're now dealing with UEFI multiboot, which is not the same as BIOS multiboot (could be better or worse, depending on your firmware implementation).
If you want everything to work 'old skool' and be able to follow any old guide you find lying around, you want to make sure you do BIOS-compatibility mode installs of your OSes, not UEFI-native. If you decide to go for UEFI native installs you'll need to figure out the quirks of your hardware's UEFI implementation, just as the article suggests. I would not, in general, advise mixing UEFI-native and BIOS-compatible OS installs, and _definitely_ not on the same disk.
40 • Semplice-5 addendum (by Joe on 2013-09-10 10:57:32 GMT from Mexico)
in @9 & @10 I apologyze some Semplice characteristhics, but definitively I totally am accordig to @38: Crunchbang is of course more stable and recomendable for productive uses, among other important things because Crunchbang 11 is based in The stable branch of Debian, it is more configurable, is a solid rock Debian OS, it has Dropbox by default, its welcome with a marvelous script to add software, etc. but the beauty of the italian first impact vs the default gray desktop of CB the autoconfiguratioa and distribution of its main menu are incredible. I also have both OS instaled...CB to serious work, Semplice five to relax browsing the web ¡
41 • Lightweight Distro (by CrazyDog on 2013-09-10 11:16:40 GMT from United States)
If you want light and quick, don't forget Vector linux. They are working on their next release and if you can judge by their beta releases - it will be good.
42 • Click and Business (by LinuxMan on 2013-09-10 12:43:45 GMT from United States)
#32, and #37 pretty much has the correct info on what Click is about. Problems are solved using Click that has affected Linux adoption for a very long time. Problems for developers. Developers want their applications to be used on multiple platforms. The more sales, the more bread on the table. The donation system does not work for people's livelihood. I prefer open source over proprietary software always, but not in the business sense of the word. Redhat is not a good example of a successful open source software business because they deal in support and the implementation and administration of open source server software. They do their own patch work and some coding but that is not how they make their money. Money can be made with Linux. Look at Android and don't start whining that it's not Linux. There are tons of proprietary systems using Linux, just not on the desktop. If there cannot be found a way for good developers to make good money in the Linux ecosystem, (apart from android), then you will have people going the way of Warren Woodford (MEPIS), and Jean-Michel (K9Copy). We are reaching a time where the operating system can be irrelevant. People will be happy as long as they can use their electronics the way they want. This is the big chance for Linux to shine on devices besides the desktop. There will always be a place for open source software and that is my preferred format, but it would be hard to make a living on just that and the ones who do are the exception.
43 • click and Kubuntu (by greg on 2013-09-10 12:46:08 GMT from Slovenia)
Click packages make sense. it's about time we have something like that. phones often have limited data plans. with this you can download on computer stick it on phone click and isntall. also for places where connection is low or hard to get this offers a good solution. not everyone is constantly online with 10Mbit+ connection. debs are ok but they sometimes still can make issues. not sure is it because devs don't follow rules or what. in windows it is easy to have two versions of same software one next to the other. in linux not so much.
anyway Kubuntu support. it's great to see it. but as i went to check the prices - i don't think it's so great anymore. especially for home users. i mean for 1 hours of their support you get like 5+ years customer support available in Windows (they too claim experts will support you). i think the right price would be more about 20-30 EUR /year and support can be via email, forums, phone, video conference or whatever. 80GBP/h is just ridicuous. or maybe they should have various type of support with various price.
44 • Kubuntu Paid Support. (by LinuxMan on 2013-09-10 14:29:35 GMT from United States)
It seems that I don't know the value of support service for a Linux distribution. I had no idea that $785.90 was a good rate for one day's worth of support. I guess it's better than $125.74 per hour. Someone tell me. Are these the going rates for computer/IT support?
45 • #1 • Friggin *buntu and their stupid ideas (by Heavensrevenge) (by Pierre on 2013-09-10 17:34:57 GMT from Germany)
On one hand side I can understand your anger about the way Ubuntu is heading. But one basic value of open source is, that you are free to use or not use existing code and standards.
And I can't see anything that is bad or stupid about creating a new packaging format or to build a new display server.
What would be bad is forcing others to do the same.
But Canonical is not forcing anyone to follow. Sure, they are trying to entrench their own new technologies, but who doesn't?
And maybe it's good, even if or because they fail. They are not doing things only because they can but because they have slightly other visions about how things should work and there is nothing wrong to build completely new tools and programs. Maybe not everything is of any use for others but there maybe are some of the ideas that are to be adopted in other projects, making them a little better.
So calm down. You don't have to share their opinions or visions, but you should respect them because they only take their liberties to do what they can and want to do and what they think is the best for them and their project.
46 • nOS (by Dave Postles on 2013-09-10 18:54:30 GMT from United Kingdom)
I hope that this doesn't sound patronizing, but I think he should be commended.
47 • LXLE is NOT bloated (by tzontag on 2013-09-10 19:48:03 GMT from United States)
I disagree with the person who says LXLE is bloated.
LXLE boots in under a minute and is clean, fast and responsive. It is beautiful and as the reviewer says "just works".
I have installed this on many computers both fairly new and old with no issues. I am glad that it has gotten reviewed here and hope it gets the recognition it deserves. KUDOS to the developers. And thank you for reviewing my now favourite distro.
48 • LXLE @47 (by kc1di on 2013-09-10 20:23:53 GMT from United States)
#47 your experience mirrors mine. I don't see it as that bloated at all.
it's elegant and just works :)
49 • re: debs are ok but they sometimes still can make issues (by Tim on 2013-09-10 23:47:23 GMT from United States)
my takeaway from reading about Click was:
install THIS set of files to a local dir on your ubukin device, click play/run, and it WILL run.
That brought back fond memories of the Win95 scenario (hmm, or was it Win3.1?) where you could reasonably expect that every non-beta program downloaded from SimTel (or BBS telnet) would "just work". Back then, storage drive space was expensive; now that storage is cheap/plentiful... yeahboy, "the self-contained (and sandboxed) apps, each with its own working directory" really seems appealing. Again.
"debs are ok but they sometimes still can make issues. not sure is it because devs don't follow rules or what."
application devs? No, I would seldom blame them. Instead, the finger wag goes to:
package maintainers ~~ who too often set unnecessarily loooong, or outright asinine, recommends/suggests lists for a given package
the various "DiSTRo" maintainers ~~ who choose to (or neglect to) setup sensible defaults and
KDE, Gnome, Unity developers (along with their puppeteers) ~~ who devicively promulgate dependency conflicts, in their endeavors to create "lock-in" scenarios (devicively, claiming "we just want to provide the BEST user experience" when their actual intent is toward "providing the ENTIRE user experience")
# apt-get install --no-recommends
( ref /etc/apt/apt.conf.d/99synaptic )
( ref: /etc/apt/apt.conf.d/00recommends )
When reviewing a "distro" (I mean both professional reviewers, and our individual liveCD tyouts), if it includes a pre-installed "synaptic package manager" and the l'il "TREAT RECOMMENDED PACKAGES AS DEPENDENCIES" preference box is tickmarked by default... isn't that a big red flag, indicating the "distro maintainer" doesn't know (or care) which end is UP?!?
Distros which ship with pre-installed appstream/packagekit "application installers" ~~ Mint SoftwareManager, Ubuntu SoftwareCenter, et al ~~ arguably represent "a whole new level of stupid" (or evil). They ripoff content (appstream database, shared by gtk-apps.org gnome-look.org qtapps.org etc) and serve it to their userbase through a crippled browser interface. Users can't (and probably don't know to) find/clear the accumulated cached screenshot images... which accumulate and bloat the users' backup filesets. Further, these "Centre" -ish installers are configured to remove (vs purge) each trialed-then-uninstalled "app" (webphone mentality? not enough storage on the device to hold allz muh games but i wants ta keep muh hiGh sCoReS and stuffs).
Of the above, based on the DW popularity rankings, the latter (distro + Centre-ish app installer + userbase) probably represents the bulk of the current "desktop linux user experience". Ah, how curiously Hegelian dialectic-ish ~~ Canonical's "Click" is coming to the "rescue" to solve certain "problems"...
Among the recent bloat/dependency gripes I've found this past year, several cases amounted to an "otherwise DE-agnostic" program pulling a sole icon from "gnome-illustrious-icon-theme" (or whichever yet another 10Mb+ icon pack). I can't, or won't, blame the app developer in those cases ~~ instead, I'm adamant in expecting that "shite" shouldn't have gotten past the package maintainer. In fact, more than once when I've downloaded package sources from an author's site, the source has no reference to said icon imagefile (indicating that the "package maintainer" must have stoooopidly introduced the dependency by choosing it.)
50 • Light distros (by Basil15 on 2013-09-10 23:54:01 GMT from South Africa)
To me a "light" distro essentially comes with a light desktop environment and not too many resident utilities (it can have an enormous potential set of downloadable apps via the repositories, no problem. The point is, how "light" is its use of resources for normal operations?) My rule of thumb is, if the DE is "heavier" than LXDE, it's on the wrong side.
Right now I'm running Lubuntu 12.04.2, waiting for the .3 upgrade. In RAM are Opera with about 40 tabs going, LO Writer with an 84-page document in preparation, a CD-full of family pix open in EasyShare, and a couple more modest apps. CPU usage on the twin-core AMD is generally under 10%, RAM usage is just under 700MB (it's a LOT of pictures). Response is brilliant. Light enough, really.
"Lighter" than that is #! (Crunchbang) which boots into its "ready-for-action" state in 99MB. I'm transferring more and more production over to it. No LXDE currently, may play with adding it for familiarity's sake. Very solid Debian Stable base, enormous number of .deb binaries available.
Most of the Puppies are brilliant and boot to about half that amount of RAM, using "interesting" desktop environments. Precise Puppy is based on Ubuntu 12.04, if you want a very small footprint but very easy access to all the apps in the Ubuntu repositories.
51 • the point ---^ (by Time on 2013-09-11 00:00:44 GMT from United States)
(clicked too soon)
my unstated points:
Currently, for the bulk of "desktop linux users", app PLUS GOBS OF "RECOMMENDS" PACKAGES are continually getting installed. This severly increases likelihood of package conflict / breakage. It further encourages user supplication ("better stick with what's provided with, and by, the distro's magic installer box thingie") AND it promotes the notion that the "latest, greatest, alliterative -titled version" is SOOOOO much better/faster (compared to a user's cumulatively-bloated, by design, dated install).
52 • LXLE (by Zhymm on 2013-09-11 00:43:19 GMT from United States)
I really like LXLE. I prefer openbox/LXDE. I've been using #!Crunchbang (with a few LX* packages added) on my main boxes for quite a while. LXLE may be 'the one' that will knock #!Crunchbang off of them. As for 'liteness', LXLE is using 80.9 MB of RAM (via conky) upon startup on my Dell Inspiron 6400 (and less than #!Crunchbang at 92 MB on the same machine). Launch firefox and that jumps to 142 MB. And it 'just works' out-of-the-box.
53 • #32 click (by jack on 2013-09-11 14:39:12 GMT from Canada)
A comparison of:
would be very helpfull
54 • What? (by LinuxMan on 2013-09-11 15:57:42 GMT from United States)
@49-Tim and @51-Time,
Have a lot of application problems? It seems by the tone of your comment that you seem to have a lot of problems with breakage. I remember the days before Win 3.0, and it was an interesting time. With that being said, I've only been using Linux for around 10 years but have never encountered the problems you seem to have. I do use the command line most of the time and Synaptic some also. I seldom use the software centers but they are necessary and for the most part they work without problems. It seems that you believe everyone should compile their own programs. That's great if you have a lot of time to kill and are just a hobbyist or maybe it's a part of your work. It's really not feasible for most people or I feel not even warranted. Sometimes I do but not often. Package maintainers are not evil. Open source development encourages modifications and for the most part its an improvement. Software centers, one click installs, and ease of use are user enhancements that work very well, are necessary, and hopefully are here to stay.
55 • @41 (VectorLinux) (by Jordan on 2013-09-11 19:51:05 GMT from United States)
Yep, agree.. great distro based on Slackware.. been using it for years.
On PCLinuxOS now, though, as their version of Mate has me hooked, various other nonsensical and semisensical reasons. ;)
56 • @49 (by Serge on 2013-09-11 21:31:18 GMT from United States)
Are the dependency conflicts you have encountered recent? My own experience is that dependency conflicts are frequent in beta-quality archives (such as Debian unstable and Debian testing) but are ironed out in release quality archives (such as Debian stable). I typically don't use Ubuntu myself, so I can't comment on the state of dependency conflicts there. I have, however, had generally good experience with Ubuntu LTS releases once a given LTS release has had a chance to mature a bit. I think that if you give an Ubuntu LTS release about three or four months to mature before installing it, the stability and reliability is roughly on par with a new Debian stable release.
Also, you might be interested to know that the .rpm packaging format does not support the concept of recommended and suggested dependencies. .rpms have only one dependency type, which is usually cited as a drawback when .rpm and .deb are compared. I'm with you, though. I always make a 99user file with a APT::Install-Recommends "false"; line in it.
Although Click and similar packaging concepts solve the dependency problems, I don't feel that it removes bloat. On the contrary, by discouraging the sharing of dependencies, I feel that Click will lead to the exact opposite. My other concern regarding Click is that of security updates. I do not know if Canonical has a solution for delivering security updates to components within a Click package. I think that if the entire package needs to be rebuilt in order to update some of its components, then Click is a regression in the delivery of security updates.
57 • yes, an article comparing of other installer pkgs would be great (by Josie Winters on 2013-09-12 04:01:38 GMT from United States)
comparison of: Click vs klik vs autopackage vs zeroinstall
Did zeroinstall find much use beyond the Rox Desktop project?
58 • Discouraging Sharing of Dependencies? (by LinuxMan on 2013-09-12 19:10:19 GMT from United States)
Maybe I'm reading it wrong, but I read it as most of the time an application package will just need to depend on the Ubuntu SDK. If a straggler dependency was needed it would be included in the package. Granted you will need an online connection it seems for this to work. It will be interesting to see how this plays out. One thing for sure is that it won't be a pain for the end user. It could be for the developer tho, or not.
59 • Ubuntu SDK (by Mirix on 2013-09-13 08:34:57 GMT from Belgium)
Not being a Ubuntu user, it seems to me that Ubuntu SDK is a good idea. If I have understood it correctly, it would be something similar to PC-BSD's PBI Module Builder. In fact, creating a bundle that includes the application along with all its dependencies is what people developing programs for Windows and Mac have always done. It is pretty much the same nowadays with iOS and Android apps.
The clear disadvantage of such a system is that it creates large packages and there is plenty of redundancy (for instance, the very same library version may be included in several packages). It is, therefore, not a strategy I would use systematically on my desktop computer but I must admit that it can be useful in certain cases. In any case, if Ubuntu wants to compete in the mobile devices' market, it has not choice for it is the easiest way to ensure the portability of the program along different OS versions.
60 • #59. (by jack on 2013-09-13 13:44:32 GMT from Canada)
I don't have a smart phone so if the following is inapplicable please ignore.
I have heard that there are thousands of apps for these phones but I do not know how many apps can be on a phone and, specifically, how much a fully loaded phone is slowed down.
From a noobie viewpoint I could put up with some slow down in return for ease of use.
>> if Ubuntu wants to compete in the mobile devices' market, it has not choice for it is the easiest way to ensure the portability of the program along different OS versions.<<
61 • Android = Linux (by Jordan on 2013-09-13 14:25:05 GMT from United States)
In the important ways. ;)
I know.. Google owns it. But Chrome is in Ubuntu's repositories... etc.
Ubuntu wants to set itself apart as "the" Linux OS for smart phones? Watch out, Google will buy you, too.
62 • Ubuntu SDK ? (by Kazlu on 2013-09-13 16:03:44 GMT from France)
Not an expert here, but if the goal is to get rid of dependances and pack every program with everything it needs, is a new SDK really needed ? What currently prevents anyone to compile a binary including the potential dependancies thus requiring 0 more dependancy on the user's machine ? Indeed, if everyone did that every software would be much bigger, but this is not necessarily a drawback : it's a matter of compromise between weight and risk of dependancy breakage (yes, quite simplified here). One can chose one side or the other, but why bother with a new SDK ? Maybe just a marketing move to tell devs that they invented a super awesome way to make sure apps work on a user's device without caring about dependancies :-D
63 • Not Google Chrome (by LinuxMan on 2013-09-13 18:27:25 GMT from United States)
Google Chrome is not in the Ubuntu repositories. You have to install the Google Chrome ppa in order to install and update Google Chrome. Furthermore you could just download the .deb file and run it. It will also install the ppa for you and keep Chrome updated. You will have 3 versions of Chrome to choose from.
64 • Maté, Mate? (by Somewhat Reticent on 2013-09-15 01:10:56 GMT from United States)
So instead of Spanglish, this was Englatin? No worries, mate, it's maté, but only for us gringos.
Number of Comments: 64
Display mode: DWW Only • Comments Only • Both DWW and Comments
|• Issue 554 (2014-04-14): Review of FreeNAS, OpenSSL bug, Fedora.next, Robolinux Stealth VM, measuring memory|
|• Issue 553 (2014-04-07): Puppy 5.7 "Slacko", end of Ubuntu One, file encryption with GPG|
|• Issue 552 (2014-03-31): Tanglu 1.0, Ubuntu GNOME LTS, SliTaz for ARM|
|• Issue 551 (2014-03-24): Linux Mint "Debian" 201403, call for end to proprietary firmware, LVM|
|• Issue 550 (2014-03-17): Review of NixOS 13.10, Lubuntu seeking feedback, Android-x86 4.4-rc1 impressions|
|• Issue 549 (2014-03-10): ClearOS 6.5 and UCS 3.2, Gentoo interview, Ubuntu app contest, Into the Core|
|• Issue 548 (2014-03-03): Review of Mageia 4, FreeBSD console driver, filtering web content, Pitivi fundraiser|
|• Issue 547 (2014-02-24): Chakra 2014.02, Ubuntu privacy, preventing unwanted remote logins|
|• Issue 546 (2014-02-17): Review of PC-BSD 10.0, Red Flag closure, Ubuntu and systemd, SlackE18, Fedora book review|
|• Issue 545 (2014-02-10): Impressions of FreeBSD 10.0, Debian votes systemd, Ubuntu file manager, server security|
|• Issue 544 (2014-02-03): Netrunner 13.12, openSUSE future, Ubuntu Touch in emulator, running commands in multiple places|
|• Issue 543 (2014-01-27): Review of Korora 20, FreeBSD 10.0, DNF, ZFS rescue CD, Bridge Linux interview|
|• Issue 542 (2014-01-20): QupZilla, Ubuntu with MATE, Arch on Raspberry Pi, best applications|
|• Issue 541 (2014-01-13): openSUSE 13.1 and Zentyal 3.3, CentOS joins Red Hat, Bodhi on Chromebooks|
|• Issue 540 (2014-01-06): SMS 2.0.6 and SME Server 8.0, Hawaii desktop, PHR statistics 2013, more on multi-part archives|
|• Issue 539 (2013-12-23): Centrych 12.04.3, Fedora 20 and its spins, dividing archives across multiple discs|
|• Issue 538 (2013-12-16): Mint 16 review, RHEL and CentOS 7 plans, SteamOS, Windows XP replacement suggestions|
|• Issue 537 (2013-12-09): OpenMandriva 2013.0, Gentoo developer interview, project Neon, Linux Mint and security|
|• Issue 536 (2013-12-02): Impressions of openSUSE 13.1, Ubuntu Touch, FreeBSD 10 delay, troubleshooting OS lock-ups|
|• Issue 535 (2013-11-25): GhostBSD 3.5, Debian and MATE, Ubuntu 14.04 features, security updates|
|• Issue 534 (2013-11-18): Review of OpenBSD 5.4, Fedora on ARM, menu names vs command-line names|
|• Issue 533 (2013-11-11): Point Linux 2.2, Pisi update, Debian and Xfce, Bruno Cornec interview|
|• Issue 532 (2013-11-04): Ubuntu and Kubuntu 13.10, Debian's init, FreeBSD's PKG-NG, Linux on ARM|
|• Issue 531 (2013-10-28): PC-BSD 9.2, openSUSE testing, nftables, upgrade pros and cons|
|• Issue 530 (2013-10-21): Kwheezy 1.2, DPL interview, Zenwalk's future, keeping up with vulnerabilities|
|• Issue 529 (2013-10-14): Ubuntu's Mir, dmesg and photorec tips, Tiny Tiny RSS|
|• Issue 528 (2013-10-07): Semplice 5, Haiku package management, Klaus Knopper interview, making custom distro|
|• Issue 527 (2013-09-30): Tiny Core Linux 5.0, SteamOS, moving operating system to new computer|
|• Issue 526 (2013-09-23): Look at ArchBang 2013.09.01, BSD Now, kernel stats, command-line tips|
|• Issue 525 (2013-09-16): The Official Ubuntu Server Book, FreeBSD 10 and OpenBSD 5.4, Skype alternatives|
|• Issue 524 (2013-09-09): Look at LXLE 12.04.3, Ubuntu's new package format, Secure Boot and dual-booting|
|• Issue 523 (2013-09-02): OpenIndiana 151a8, openSUSE "Evergreen", GNOME and DuckDuckGo, running apps from RAM|
|• Issue 522 (2013-08-26): Look at gNewSense 3.0, Ubuntu Edge fundraising failure, exploring GPL|
|• Issue 521 (2013-08-19): Review of Korora 19, Fedora considers return to "Core", Haiku package management|
|• Issue 520 (2013-08-12): Salix OS 14.0.1 "KDE", Xubuntu experiments with XMir, managing passwords with KeePass|
|• Issue 519 (2013-08-05): Review of Porteus 2.0, Kubuntu lays out plans for Wayland adoption, adjusting system swappiness|
|• Issue 518 (2013-07-29): MidnightBSD 0.4, Razor-qt, Ubuntu Edge, mounting infected drives|
|• Issue 517 (2013-07-22): Zorin OS 7 "Lite", Slackware turns 20, UbuntuForums compromise, Raspbian as home server, Tor|
|• Issue 516 (2013-07-15): Review of Fedora 19 "KDE", Shuttleworth on Mir, Seth Vidal, Kingsoft Office for Linux|
|• Issue 515 (2013-07-08): Whonix 0.5.6 and Deepin 12.12, MintBox, processor capabilities, distros for Raspberry Pi|
|• Issue 514 (2013-07-01): Peppermint Four, Mir, Mandriva forks, ThinkPenguin on libre hardware|
|• Issue 513 (2013-06-24): Look at ROSA, PC-BSD updates, Xen4CentOS6, Slacko vs Precise, Mageia interview, shells|
|• Issue 512 (2013-06-17): Trisquel 6.0, RHEL 7 with GNOME Classic, from Linux to FreeBSD, first look at Wayland|
|• Issue 511 (2013-06-10): Mint 15 impressions, GNOME Classic, Ubuntu Community portal, Absolute OpenBSD|
|• Issue 510 (2013-06-03): Impressions of aptosid 2013-01, Wayland comes to Raspberry Pi, maintaining DNS settings|
|• Issue 509 (2013-05-27): Mageia 3, Debian GNU/Hurd, RebeccaBlackOS with Wayland, ports|
|• Issue 508 (2013-05-20): Review of Debian 7.0, interviews with Clement Lefebvre and Gaël Duval, scripting with xdotool|
|• Issue 507 (2013-05-13): Impressions of Calculate Linux, 13.4, Ubuntu's portable packages, mintDrivers|
|• Issue 506 (2013-05-06): Ubuntu and Kubuntu 13.04, Debian "Wheezy", Slackware on systemd, distros for Raspberry Pi|
|• Full list of all issues|
|Free Tech Guides
Markdown is a useful tool for producing professional research papers, memos or blog posts without the hassle of remembering HTML tags.
FREE 28-page guide