Her name is Leela, she got a special award. Cutest pet in show.
Just to prove how crazy IRC can get sometimes.
Laptopmag just published a story about how the ‘window of opportunity’ for Lin… it’s GNU/Linux darnit, has ‘closed’. Something attributed to a few small changes in the market. Several bloggers have already taken the article apart for the degree to which it over-blows both a few small victories and losses, I felt that somebody should take it apart for it’s sheer ridiculousness.
How do you take something ridiculous apart ? Easy, you make fun of it.
Well how do you make fun of somebody who says GNU/Linux is dead ? Well you tell him it’s not dead, it’s undead. Hell I’ve seen so many ‘GNU is dead’, ‘Linux is dead’,'GNU/Linux is dead’,'It’s dead, we really mean it this time’ and ‘Linux is so dead it was never even there to start with’ articles over the years that by now the only reasonable assumption is that GNU/Linux in fact passed away sometime in the 1980′s and have returned from the grave.
Ever since then it has been eating the brains of all the FOSS developers out there, using their smarts to keep it’s accursed form alive and acquiring their skills as new powers…erm i mean features.
First Balmer called us ‘a non-issue’ and didn’t really ‘see it as a threat at all’. Then we were the enemy.
Now, they are at least admitting the truth: we scare the shite out of them. Which is appropriate, zombies are scary and just like our horror-movie kin, the problem is, the no matter how many you kill, three more are waiting to take his place.
So far, we’re winning, actually we’ve always been winning. It’s just that journalists like stories that happen really fast. I’ve said it before, changing the desktop status-quo will take at least a decade to complete, we’re getting there. As Neil Armstrong taught us, sometimes a small step is a big leap. Every small step for GNU/Linux is a big leap for software freedom and mainstream acceptance (even dominance though I shudder to use that word in anything but an absolutely joking sense: the whole point is to not dominate the users).
So the window for Linux to suddenly replace all the windows desktops (sheez try saying that ten times fast) hasn’t closed, it never existed. The windows platform was basically the first graphical interface most people ever knew. It spread fast because the only other one was much more expensive – and it had almost 20 years to build up inertia. You do not replace something like that in a few months.
You replace it, one computer at a time, even if it takes ten years. We’ve had some really epic wins lately, and a few epic fails along the way. This is not new. When egcs split up from gcc that was an epic fail too. Not egcs – the very fact that a split was required to light a fire under gcc’s ass. When they remerged, that was an epic win. Basically, you win some, you lose some. But something with the momentum of GNU/Linux, the drive and passion behind it. Something that ultimately really is better for most people must ultimately win.
I see the GNU/Linux revolution in the fact that: every single day percentage of people for whom it is better is growing. I think it may already be a majority, if not it’s very close. The bigger it gets, the louder the voices of that small part that can overcome their inertia becomes, which drives more people to switch yet again. Ultimately, GNU/Linux will win the desktop because we’re getting better – every day, and nobody else can really say that. But it will take a few years.
PS. I actually started writing this post a few days ago, but a power failure cut me off halfway and I hadn’t even realized there was a draft until today, when I decided it was worth resurrecting it.
So, wanna guess what has occupied the majority of my time lately ? Yep, reconciling the company’s books with the bank statements to make sure everything adds up right and all the numbers fit so that I can hand over to the auditor to do my tax filing.
I hate accounting. I mean, a few bits are complex at first- but it doesn’t even have challenge to keep it interesting. It is just a continuous repeat of the exact same steps… yet requiring utmost concentration every single time. Basically it seems to be almost exactly designed to stress the human brain to the ultimate extent.
Sure software engineers have been trying to write easier and easier accounting packages in recent years, but the reality is that accountants are clearly not engineers. We’ve had more than 4000 years o accounting history, during which it was frequently refined – not the least of which would be Pacioli’s revolutionary treatise on double-entry accounting – now the standard method for all but the most trivial of businesses.
But at no point, and this is radically clear, was the process of accounting ever engineered. From ‘writing down where the money went’ – which seems to date back at least as far as the Egyptians (and there is evidence the Sumerians were doing it even earlier than that) it was just always refined. One revolution in 4000 years – which was designed to make errors findable. Double entry accounting makes it easier to audit the books (you can pick up when the balances don’t zero out if every transaction has both a debit and a credit) – but it does not make it easier to DO the books.
In fact, as business has become ever more complex, accounting seems to have become a classic example of freeping creaturism. Today an accountant has to keep track of thousands of types of transactions, each with their own quirks and tricks, always subtly modified by local laws. Computers may make it a bit easier to enter -but frankly just from my books I now doubt very much that even the best accountants in the world could possibly keep track of any large corporation’s books. So it all gets fed into the computer and blindly trusted because any kind of understanding of whats in there is completely unfeasible. The auditors then check for accuracy to the best they can (assuming there is nothing Enron-esque behind the scenes) – and finally a bunch of financial statements are produced which bare about as much resemblance to the books they are a summary off as a hallmark card does to the KJV Bible.
No wonder we have massive corporate corruption – it’s practically impossible to actually do accounting that makes sense – and what could be easier than to cheat with numbers nobody understands anyway.
Personally I would be in favor of a radical grounds-up reengineering of the entire principle of accounting – with no accountants involved. Imagine if a group of really good programmers, with some very good businessmen teamed up – what they are tasked with is ‘design a way for a business or individual to keep track of all financial information, from the 5c thrown into the gumball machine to the total bank- balance at any moment. The process should require as little human intervention as possible and yet allow data from the highest to the lowest level to be referenced instantaneously and in real time – including by authorized outside parties (e.g. the taxman). It should be design to capture and correct errors as far as possible and use whatever technological means it can to ensure the capture of the data and the reporting there-on is as efficient, logical and easy as possible.
Then let them… ENGINEER a way to do accounting. Not just 4000 years of adding on, one major breakthrough and another 800 years of just more and more and more added to try and keep up with ever greater demand. When a piece of code reaches excessive clutter and complexity, we rewrite it from scratch, sometimes we redo an entire project from nothing, using only what we LEARNED from the past, none of what we actually DID – to DESIGN it better.
From my day to do day experiences doing the books of my company, as compared to my day to day experiences as a software engineer* – one thing is clear to me, accounting needs to be redesigned, or at least have the worlds largest ever code-cleanup done.
True there are complexities like getting every law system out there aligned to a new accounting system – since they all demand classic double-entry with strict rules about how it should be stored and presented, but those rules are not entirely inflexible (after all we have more than one accounting program GUI even now) and laws can change – especially if the result is to achieve the very reason they were made much better. I see more likely opposition from current billionaires, every single one of them had at least a little help from ‘creative accounting’ (a term that means – ‘we know nobody can read the books so we can write anything we want in them’).
So politics aside, and getting past all the accountants out there going: EEEK our beautiful paychecks ! I still think it’s a good idea. Who is with me ? And here comes the next revolution: imagine adding the development power of open-source and the ethical enforcement of free software to the single most corrupt profession there is ? (Yes, accountants are more corrupt than politicians – politicians cannot use accountants to be corrupt with sure some accountants are honest, some lawyers too – it’s just that they are few and far between – and it’s virtually impossible to tell the difference.) ?
Yep, I think we should re-engineer accounting using social development. WikiGPLify our money !
Now that is a revolution that may actually change the world more than the internet did (though it would inherently depend on the internet of course)… or maybe I just really hated the last few weeks and I need a laydown and a beer ?
*I have my viewpoint on those who believe programmers are not engineers, and the debate is very long with much that must be considered in fairness. I have no desire to get into it now – I merely use the term here to explain the difference between ‘applying scientific knowledge to the solution of a practical problem’ as opposed to ‘trying to do the best we can with the system we inherited’.
I am 28 now. I feel old. I don’t feel too old. This blog post sucks. I, on the other hand, do not. What more can I say ? It’s my birthday and I don’t want to sound conceited. Please send gifties.
I have been wanting to try out Fedora 9 ever since I read in an interview with the project leaders about 2 months ago that it was going with a KDE4 default desktop (as far as I can tell, no other distribution has taken the plunge). Now Fedora is an interesting beast… it gets little press these days yet it has been consistently in the top ten at distrowatch for years, and high in the top ten too.
So who is using Fedora these days ? Back in the 90′s all us linux’ers pretty much used redhat… who uses Fedora today except a few die-hards and Eric Raymond ? Actually I would say Fedora has become the Haute Couture of GNU/Linux. It is constantly creating and showcasing new technologies that will be fundamental in other distro’s a year from now (remember who actually wrote NetworkManager ?).
So when the beta came out, I couldn’t wait any longer and decided to take the plunge and try it out. I had been getting tired of increasing issues with gentoo (most of them more community than technical… it’s kina falling appart), and I wanted to see what a KDE4 default desktop would behave like. I had to wait two days for the month-end so that my bandwith could roll-over but I downloaded the KDE-live ISO and booted it up. Here is my review of the experience. Note please: this is a beta, and part of my motivation is to test said beta and report bugs – so while I will mention any and every bug or annoyance I found – this is because I would love the stable release to be amazing, not because I don’t think the beta is pretty damn good. I am however more critical than normal, since that is what beta’s are for.
Live CD and Installer:
1) It was slow as hell, at least on the liveCD. Now I should add that I ofound myself out of CD’s this morning and burned it to a DVD disk so that may have had something to do with it, also this box has only 512mb of ram which is rather low for a LIVEcd running KDE4. On every single LiveCD startup, plasma would crash before finishing the skipping, skipping intro sound. It would then recover (much better than KDE4 on gentoo did) and only then be usable. Everything ran fast, loaded painfully slow – this is classic of a LIVEcd with too little ram however so the only complaint is the fedora seems to have put itself outside low-end machines ram-wise. Of course, just using the install-only media would be a solid work-around.
2) It is interesting that the live media has gone full circle for a proper 3rd generation distro. It installs in one go without needing any package selections or anything to start.
3) My network and most of my hardware was detected and working automatically, my screen ran at full resolution (despite being a 22″ widescreen LCD) – the hardware detection is certainly at least as good as the best out there.
4) The first real bug I found was in the installer. I did not want it to wipe out all the linux partitions on any drive and replace it with it’s own scheme – I wanted to do a custom one. This proved impossible, several attempts were made but it kept failing to initialise swap. At first I just used the swap that was there already, later I tried to delete and re-add it… ultimately the only way I could install fedora was by deleting the swap and going without any swap… which it then warned me about. I added swap again manually after installation but this is a critical bug which will affect most people coming from another distro (e.g. who have data on some partitions that they want to preserve) – and most people without a few years on a GNU/Linux system will not know how to do this, especially the fstab entries which are rather unlike any other filesystem.
5) The installer strangely didn’t ask to create a new username, though it did set the root password (and was the first one ever to consider my normal root password not strong. I actually chose to go ahead with it though my systems are very well firewalled, if somebody actually gets to them – no password will stop him).
6) Very nice feature for the security concerned is the ability now to encrypt filesystems right in the installer and install to them (I didn’t test this however).
7) I tried to test the new add/remove programs tool but on the LIVEcd at least it was either not working at all, or just so slow as makes no difference.
Installed system
8)I then proceeded to boot into the new installation. As I rather expected I was greeted with a first-time setup wizard. It began with a brief indication on the fact that the code is licensed, without being very specific (of course I know the details but how many noobs do ?)
9) The next screen did include creating a username, I was wondering there whether or not use my normal username and absorb my old desktop settings, ultimately I chose to do so, if only since there no longer is a KDE3 that I could preserve them for anyway. Just for safety though I wanted to quickly Alt-F2 and mv my .kde directory out of the way into a backup. Sadly, it refused to allow me to do this.
10) Alone among all distro’s I’ve ever tried, it picked up that my home directory already existed and offered the option to reuse or create a new one. I chose to reuse.
11) Then it created a hardware profile, requesting it be shared. Since I plan to post a bug report about the swap thing, I decided to do so, just in case it helps.
12) Finally, a real login screen… WTF !??!?! It’s running GDM ! Why is Fedora KDE edition with it’s spanking new KDE4 desktop not using KDM ? Just to prevent duplication of effort ? Did they find a work-around for the KDE features that depend on KDM ?
13) About 30 seconds after logging in, the desktop died. SELinux popped up an AVC denial message, and I was dumped to a terminal… not cool.
14) I ran init 3 and then init 5, which got me back to a login screen. I then tried to log in again. Once more the desktop loaded… I waited… again there was a message about yakuake being missing (well it isn’t there but it was obviously still in my autostart menu). As soon as I clicked OK, it died.
15) I started thinking about possible fixes. For example I may have had an old KDE4 directory… so let’s go see… I mv’d the kde directory and rm -fr’d the .kde4 one… this all I did by hitting Alt-F1 and logging in as root on the terminal.
16) I logged in again… I waited… it seemed to work…yes… I could stay logged on !
17) So far, I still haven’t go the add/remove software thing to work… I’ll figure it out I’m sure but since it’s alone on the menu… shouldn’t it be set up out of the box ?
Conclusions:
I really like it so far. It’s a nice system and could become the easiest and most powerful desktop I’ve used in the GNU/Linux world in a long time. However it has serious problems with people who already have installations, several of them which took some serious expertise to work around. This definitely needs to be resolved before stable release if Fedora is to get the success that I truly believe this version deserves.
If you use Linux, the command line cannot be entirely escaped. New users may hardly ever need to see it, but to take full advantage of the system, you ought to learn a little at least, seasoned users cannot live without it. This post however is not about how cool the command line can be, but about my top ten commands and tricks that I use to boost my command line productivity. The list ranges from shell applications to trick codes but it is what makes the shell so powerful to me – and I hope you benefit from them.
This list is more than just tips, it also shows off some of the true power of the bash shell, especially on the GNU/Linux operating system base. Have fun with it, and if you have some great tips – please post them in the comments.
Keep an eye out for part 2: My top ten KDE tricks.