Introduction About Site Map

XML
RSS 2 Feed RSS 2 Feed
Navigation

Main Page | Blog Index

Archive for the ‘Linux’ Category

The Old Sirius ‘Open Source’ Was a Patron (Sponsor) of KDE and FSF

Summary: The company my wife and I joined was (at the time) still Free software-centric and reasonably friendly towards staff; today we examine Sirius of a decade ago

IN THE previous part we showed some preliminary statements about this report regarding Sirius ‘Open Source’, a company known very well from the inside for nearly 12 years. Today we can introduce the softer side of Sirius or what Sirius used to be.


The Open Source Era

At the Beginning

Sirius is early Patron (sponsor but a more modern term for sponsorship) of KDE, a prominent European project for GNU/Linux- and BSD-centric desktops and laptops. Sirius is also an early Patron of the FSF, which stands for the Free Software Foundation (listed and thanked by the FSF for several consecutive years, as The Internet Archive’s Wayback Machine confirms). The FSF was established in the mid 1980s, i.e. almost 1.5 decades before Sirius was even founded. Many early employees of the company were GNU/Linux users, KDE users, even Debian Developers. They were highly technical people who shared the philosophy reflected upon by these (aforementioned) generous donation.

The Wayback Machine shows the Sirius site (old snapshots). Wayback Machine screenshot of the front page:

Wayback Machine screenshot: Sirius and GNU

Wayback Machine screenshot from the FSF:

Wayback Machine screenshot: Sirius and FSF

The company was able to attract high-calibre staff based on these credentials and hard-earned track record. Roy too was attracted to the company based on these publicly- and readily-visible credentials.

People worked overtime to please Sirius clients, some of which were very high-profile. Sadly, as we shall show later in this document, that’s no longer the case and hasn’t been so for several years. The company is living off or leeching off its (distant) past reputation and is extremely paranoid about people finding out about a rapid pivot across numerous dimension, e.g. in-house technology, levels of relevant skill mastered by staff, overt nepotism, and promotion of technologies not compatible with the company’s original mission statement. A lot of the work produced by the company — and it is no longer so much in-house work — leverages Open Source/Free software (libre, or free as in freedom) but does not share back the contributions (or mere code changes), even when initially there’s intention to do so, even if not for licence compliance purposes but status (companies that share back code and don’t just use or exploit code have better karma, averting the image of becoming parasitic to the community).

An Exercise in Optics

The company’s Web site is intentionally outdated. It projects outwards an image of a company that may or may not existed about half a decade ago. Some of the clients being bragged about are well over a decade old. The intention there is to use past clients, no matter how old, to present a credible, potent, highly-experienced firm with high competency. A lot of the actual work gets done by associates (external contractors), not dedicated staff, and those associates have their own firms, which aren’t connected to Sirius at all, except maybe loosely. We’re left to assume that Sirius quietly transforms into a sort of middleman or reseller across a number of domains. For instance, there are a number of things Sirius claims to be supporting, but managers inside the company have no actual staff familiar with ways to maintain such things, so Sirius would typically contract outwards or outsource. This is a crucial point. This isn’t how the company presents itself to the public.

The company, at least in the past, not only had legitimate credibility in the Free/Open Source software world; it goes beyond that. This is well documented and it’s not too hard to find the company’s founder cited extensively in the technology-centric media, especially over a decade ago (Roy used to cite him a lot, including in his site, Techrights). The founder is very visible in national and international press.

The company made a name for itself by attending international events and even hosted an event promoting the use of OpenDocument Format (ODF) in the UK. That’s vital advocacy of Open Standards at a very crucial point in time (format wars and struggles against vendor lock-in). As we shall see later on, these laudable outreach efforts have played a considerable role in attracting Roy’s interest in the company. As an aside, the company first sought to recruit Roy, way back around 2006. The founder of Sirius phoned Roy when Roy was completing his Ph.D. degree in Victoria University of Manchester.

Social Control Media Isn’t Social Because You Don’t Lose Real Friends (or Social Life) Just Because Some Site/Instance/Pod Shuts Down

I will soon make changes to my microblogging routine. Probably some time next month. Here’s the alta vista of it.

“Social Control Media” is a term I coined many years ago; many people, even Wikileaks, have adopted the term since then. When I say “Social Control Media” I don’t limit myself to Twitter and Facebook; it also applies to LinkedIn (Microsoft claiming to ‘own’ your identity), GitHub (Microsoft claiming to ‘own’ your work/code), YouTube (video), and TikTok (crap). More importantly, as I’ve repeatedly pointed out in Techrights, Free software- and freedom-based sites aren’t robust to many of the same issues (volatility, misinformation) and even if they’re self-hosted, decentralised, federated etc. their existence is transient. Some sites or software will cease to be maintained within 5 years or less (in the case of self-hosting, a new version of PHP, for instance, can break the software you self-host).

Static sites with simple files are generally a good idea if you intend to keep your data, not only through the Web but whatever protocols will exist and get popularised in the future. More importantly, never rely on making “connections” online; make them “in real life” as real friends don’t need the Internet to keep in touch. I’ve hardly used the Internet at all to keep in touch with real friends.

Next Friday JoinDiaspora will go offline after more than a decade. When it comes back online it’ll be “read-only”, available temporarily only for users to be able to export their data and move it elsewhere (to another pod). All the connections will be lost, even if posts and comments are going to be preserved, according to the promises from the project’s core team.

Maybe I’ll fondly remember those 3,000+ “followers” I gained in JoinDiaspora.

schestowitz joindiaspora connections

What about the 2,000+ “followers” of Linux (TuxMachines)? I hardly ever knew them.

linux joindiaspora connections

After thinking about it for over a month and having spoken to another person who has posted heavily to JoinDiaspora for over a decade (with many “followers”, too) I am pretty certain I’ll migrate both accounts to another domain, another pod. I’ve not decided which one yet, but it needs to be something that can last and keep alive for at least another decade. As I explained here a few days ago, I’ve already lost more accounts than I can remember. I don’t want to jump from one dying pod to another soon-to-die pod. As for self-hosting, it’s out of the question due to complexity (Diaspora became bloated; I tried installing it almost 8 years ago and it was already very heavy and complicated to manage).

For me, IRC provides a more reliable means of communication and it’s vastly easier — not to mention a lot cheaper — to maintain.

2 Decades Without Windows

“Windows 98 should have been released for free on Jan. 1, 1996 and titled Windows 95.1. If this were Hollywood, then Windows 98 would be the equivalent of ‘Heaven’s Gate’, ‘Waterworld’ and ‘Godzilla’ rolled into one. A huge, overhyped, bloated, embarrassment.”

Jesse Berst, ZDNet editor

MY GENERATION (I’m 37) grew up on DOS. Not necessarily Microsoft DOS, either. Just DOS. As a kid I used to work from the command line. We, as kids, taught one another new tricks; sometimes an adult would visit to teach us things and copy some programs for us (floppy disks with compressed archives). Various utilities like RAR were useful. Sometimes an infection (malicious program) needed to be removed. That was before the days of Windows 3.x — the days we used ncurses-type interfaces to type documents and send these to printers. Later on I did some programming with batch files and at around age 15 I started with Pascal (quite popular at the time owing to simplicity and relative elegance).

I mostly missed the BBS generation (some friends of mine used it; they’d copy for us files they got from there). When bulletin board systems were still popular many computers did not even have modems (few of my classmates had them, usually because of lack of a technical parent, and only one of them was a GNU/Linux user in the mid/late nineties). I think I got my first modem when I was 14 and IRC was probably the first thing I used “on-line”. After Windows 95, which many people used at that time, I bought my last Windows laptop. Actually, my father bought it. He used it and then passed it to me. It had only 32 MB of RAM and Windows 98. I carried it around and used it in university as an undergraduate student (at the faculty I used GNU/Linux at the time). It retired years later and I’ve not bothered with Windows since then. I wrote a great deal about it in USENET at the time. Memories from these days are mostly gone by now; I barely ever touch Windows and when I do it’s over Remote Desktop, typically to access a client’s network, e.g. to run PuTTY from a remote system. That happens about once a month (patching Debian GNU/Linux servers).

Was Windows 98 a decent operating system? No, it was unreliable, but at least it ran on modest hardware without much RAM. I ran Firefox on it, with a total system capacity of something from the mid-nineties (~400MegaHertZ CPU, 32 MB of RAM). That was before Microsoft added back doors to Windows (this was reportedly done in 1999), before the bloat of NT and before DRM (Vista).

With 3 weeks left before the end of this year (and this decade) I remember not so fondly the 90s, back when I used Windows. In 2000 I moved to GNU/Linux, helped by a Finnish friend, an exchange student at the university. In a sense, next year I become a 20-year GNU/Linux user.

I Never Had a Machine With More Than 2GB of RAM. But Phoronix Portrays GNU/Linux as Sucking on Memory Management.

The main problem is bloated software, not Linux

LAST night I saw a somewhat ‘trollish’ bunch of reports. I saw Slashdot [1], linking to Phoronix [2] with a grammatical mistake in the headline (“Yes, Linux Does Bad In Low RAM / Memory Pressure Situations On The Desktop”).

Let me first clarify that I’m no kernel guru. Far from it. I’m a programmer, but not an OS programmer or kernel developer.

“Is this accurate?”

That’s what I asked people who may know better. They know kernel developers (and development) better than me.

“Is it true or is Phoronix taking the piss?”

I saw comments on it (almost 100 in Phoronix and 400 in Slashdot), but they’re short and vague. How is Linux doing compared to other OSes?

“ZSWAP makes a huge difference (RAM compression),” one person told me. “GRUB_CMDLINE_LINUX_DEFAULT="zswap.enabled=1 zswap.compressor=lz4 zswap.zpool=z3fold"

“Windows 10 & MacOS & Ubuntu have RAM compression on by default,” he added.

“I’m not sure exactly what they have,” he continued, “but I remember reading about it before. RAM compression isn’t new at all. But it is relatively new (few years) to be on by default.”

Linux kernel space is typically ahead of the curve (compared to the competition); Con Kolivas comes to mind when it comes to claims that it’s optimised for servers but not for desktops (the scheduler, not RAM/swap management).

To check compression status on one’s system:

grep -R . /sys/module/zswap/parameters;sudo dmesg|grep zswap; sudo grep -R . /sys/kernel/debug/zswap;f;sudo sh -c 'cd /sys/kernel/debug/zswap;perl -E "say $(cat stored_pages) * 4096 / $(cat pool_total_size)"' # to check if loaded ; used ; ZswapCompressionRatio

How can one argue that GNU/Linux does worse than counterparts? Slashdot promoted this story with about 400 comments and Phoronix even has a grammatical mistake in its headline (Slashdot corrected it, Michael of Phoronix has not.) It’s an eyesore in a sense; both the message and the English. The headline also states that as fact even though it’s to be attributed to just one developer, Artem S Tashkinov. To an outsider (to kernel development) it may smack of clickbait. It’s stigmatising “Linux” as not successful on “desktop” because of “technical” “issues” (not OEM bribes, ISVs etc.), but if it’s based on purely factual bits, then let it be, I’m fine with it.

My gut feeling was, there’s likely more to the story; can Apple and Microsoft handle compression of RAM for instance? If so, how well? I don’t know people running Apple-branded systems and PCs with Windows on just 2 gigabytes of RAM (which is the most I ever had on any of my systems; same with my wife, who is a GNU/Linux user).

Another person, who is proficient at kernel matters, told me: “I have caused Linux to stall in swap hell many times and there are long list of particular causes of it. facebook made oomd to attempt to deal with it in userspace.”

The first person weighed in again: “they should compare to other distros, and to MacOS/Windows. But the complaint is valid, IMO: that Linux (defaults) *should* be ‘smarter’ when OOM; check what is your vm.vfs_cache_pressure. $ cat /proc/sys/vm/vfs_cache_pressure [...] I have mine set to 50, because this article makes sense to. (but I’m not sure what default is nowadays?)”

A third person wrote: “my Orange Pi with 2 GB of RAM running everything mainline is my main desktop, which runs generally fine with a swapfile of also 2 GB…”

That was in the #techrights IRC channel this morning.

The person added, “the bigger question for me is why applications these days are so heavy and slow…”

That last point is what I too have raised many times before. GNU/Linux is handling reasonably well a complete system with 2GB of RAM (or less). Super-bloated applications is where things start getting trickier.

Related/contextual items from the news:

  1. Linux Performs Poorly In Low RAM / Memory Pressure Situations On The Desktop

    It’s been a gripe for many running Linux on low RAM systems especially is that when the Linux desktop is under memory pressure the performance can be quite brutal with the system barely being responsive. The discussion over that behavior has been reignited this week.

  2. Yes, Linux Does Bad In Low RAM / Memory Pressure Situations On The Desktop

    It’s been a gripe for many running Linux on low RAM systems especially is that when the Linux desktop is under memory pressure the performance can be quite brutal with the system barely being responsive. The discussion over that behavior has been reignited this week.

    Developer Artem S Tashkinov took to the kernel mailing list over the weekend to express his frustration with the kernel’s inability to handle low memory pressure in a graceful manner. If booting a system with just 4GB of RAM available, disabling SWAP to accelerate the impact/behavior, and launching a web browser and opening new web pages / tabs can in a matter of minutes bring the system down to its knees.

Linux.com is Nowadays Too Much Microsoft Propaganda and Openwashing

This is what the site looks like this morning

Linux.com

Welcome to Linux.com!

Microsoft… is “OPEN SOURCE”!

Sadly, this has become far too common in that site under its new (and sole) editor. Whose interests are served now? Linux Foundation sponsors.

The Linux Foundation Killed Linux.com After Two Decades of News Service

Linux.com

Remember NewsForge? Or Geeknet? Yes, once upon a time Linux.com was a site for geeks; under the Linux Foundation’s ownership it gradually became a site for large corporations, where they could buy puff pieces and interviews (agreed upon in advance, as covered in Techrights). But among those puff pieces and ads one could at least find some journalism.

I’m sad to see, 2 months down the line, that Linux.com is practically dead. Nothing new is being added to it. It’s now a fossil. Eternally? The Foundation has said not a single word about it. So much for openness and transparency, eh?

Argos Uses GNU/Linux, Windows Leads Only to Malware

200px-Argos.svg

ARGOS, the British retailer which according to Wikipedia employs over 50,000 people, seems to have moved many of its well-integrated systems (online and in-store) to a new platform not too long ago. We’re frequent customers, so sometimes we see downtimes and issues which reveal details about the back end. There is something we can report today.

Yesterday at Argos there was a malfunction at the terminal (thin client with IBM touchscreens as the only user-facing part) and it uses GNU/Linux, as one can easily tell. Argos seems to be using Fedora (an old version likely) based on the window decorations, probably with GNOME/GTK. For 50,000+ staff plus millions of customers that would be a lot terminals running GNU/Linux.

Argos recently changed the terminal systems/front end (not to my liking, as I liked the old interface better). The crash yesterday resulted in an error message showing Opera (proprietary Web browser), so presumably that is what they are using on top of GNU/Linux. Odd choice of browser, but that’s what they want….

Last year we reported that another British giant, Ryman (smaller than Argos, but still a highstreet chain), had moved to GNU/Linux. They told me that had dumped it due to Windows malware.

Incidentally, a friend who goes by the name iophk (for anonymity), shared this link with me yesterday. “Windows should not be on PoS or anything else mission-critical,” iophk wrote. Well, based on Ryman and Argos turning to GNU/Linux, many large stores gradually learn this. My current employer moved Specsavers (British giant) to Free software in the server room a long time ago, about a decade ago.

Retrieval statistics: 21 queries taking a total of 0.117 seconds • Please report low bandwidth using the feedback form
Original styles created by Ian Main (all acknowledgements) • PHP scripts and styles later modified by Roy Schestowitz • Help yourself to a GPL'd copy
|— Proudly powered by W o r d P r e s s — based on a heavily-hacked version 1.2.1 (Mingus) installation —|