The Pengpod

I bought the Pengpod. Or, more precisely, funded its Kickstarter campaign. I wanted to have a Vivaldi tablet, but since Aaron Seigo is being such a slowpoke, i jumped on the Pengpod instead (although i'm sure i'll get me a Vivaldi as well, for reasons that will become clearer as a result of this post).

And let me tell you. Linux, not optimized for tablets, sucks cock. It really does. The default Linaro image is an atrocious experience. The main reason i wanted the tablet was for the KDE Plasma Active 3, but by the time i got my tablet, the OpenSUSE image link from the PengPod wiki was giving out error 404, and to this day still hasn't appeared. Moreover, the Plasma Active from Linaro repositories is an old version 2, and is extremely sluggish (well, Linaro is Ubuntu, so what the fuck did i expect). In other words, Linux on this tablet sucks balls. Even the XBMC, which looks cool, didn't work - or, more precisely, it did, but the touch screen doesn't work in XBMC.

So i finally gave up and installed an Android image. Now we're talking. At least it works on a tablet.

But why Vivaldi if PengPod sucks? Well, the PengPod is clearly a product for tinkerers. I'm not shy to hack away but sometimes i just want a working piece of technology. As in, i'm fine with having to do some hacking to make it work, but i don't want to do everything from scratch! I mean, there was a video of Plasma Active 3 running on exactly my tablet (PengPod 1000), and here i am, getting a 404 on OpenSUSE image for fucking three months now! I at least expected a somewhat working Plasma Active 3 when i was buying it.

The Vivaldi, however, is positioned as much more user friendly and, more importantly, it will be backed by the KDE community and official Mer images. That means i won't have to compile half of KDE repository myself just to get latest KDE Active stuff. I'm still waiting, Aaron. Please don't let me down.

Calligra lost for Words

I am an avid KDE advocate. I love KDE. I applauded when they started developing Calligra Suite. However, i've just tried Calligra words to do some basic editing. Sorry, KDE guys, but Calligra Words isn't even remotely ready for general population.

Two major gripes. First. I created a template. Then i tried to create a new document based on that template. Calligra Words has entered what it seems like an infinite loop. It stopped working and responding.

Second. I was trying to edit a document. Styles are unusable. Period. They are unusable, uneditable, behave really weird and are bug-ridden. Sorry, guys, no go. I'm installing LibreOffice. Calligra Words isn't ready. I hope it will be, some day.

Linux sucks!

First of all, sorry for not annoying you with much posts lately, been a little busy moving to a different place and settling in. I hate moving, i hate all this anxiety, i hate when a steady rhythm of my life is anally raped like this, but i guess that's life.

Now, Linux sucks. Yes, this is going to be a yet another post on how Linux sucks, with me bringing arguments about Linux while still assuring that i'm a fucking Linux zealot, advocate and whatnot. Yep, just like that, let's get it out of the way. I love Linux. I use it as my everyday OS, i occasionally do some development on Linux, i happen to feel comfortable in the command-line, et cetera, et cetera. That's all great.

But there's a big butt! For starters, i'm tired of and frustrated with Kubuntu. I've been using Kubuntu since, well, practically the day it showed up on the radar. I used it through 7.x series, through 8.x, through 9.x series, through 10.x series, through 11.x series, but i will not use 12.04. At least not as my main system. I'm moving to Chakra Linux. Why? Because Kubuntu is so god damn sluggish it's unbelievable. It literally crawls like a snail. Plus, i'm a bit of an adventure guy, i like adding PPA's and testing semi-stable packages. The problem with that of course is that every once in a while everything breaks. Yes, it's my own fault, but i don't care. Chakra manages to run testing branches of KDE4 without depriving me of graphical desktop every so often, so it's not like it's impossible.

The only successful upgrade to a new release i've had ever was from 10.10 to 11.04. This was Kubuntu at its best - not only it didn't break anything, it actually made things better! Not a single upgrade before or since was clean and easy - every time i had to shoot adrenaline into the kernel to bring the system up again. During big updates package manager often hangs and does nothing. Lately it's been hanging on the very last step of updates (processing triggers or something). Whether it's Muon or KPackageKit it doesn't matter - it still sucks.

Fonts. Fonts fucking suck. After - what, twenty? thirty? - years of this phrase first appearing on a public internet, it's still true. Fonts suck in Linux. I don't care that i can make them not suck, that i can edit zillion config files and install some shit just to display fonts properly - i don't give a fuck. When i install Windows, i don't fiddle with fonts. Why should i do so here?

Open Source projects. Some of them are pretty awesome, most of them suck. LibreOffice? Big fat sluggish piece of shit. Firefox? Don't get me started. The GNOME software camp? Fucking abomination. GNOME is a disease, a cancer on a free software movement. GCC? If this is the best compiler the community is capable of producing... Well, i rest my case. KDE software and KDE-related stuff (e.g. ownCloud) is probably the only worthwhile FLOSS software around, and it's not without its problems either.

Sound. SOUND!!!! How many god damn sound frameworks is enough?! OK, things have been slowly moving towards PulseAudio lately, and with version 4 my Skype is finally able to work properly without taking over my sound hardware or closing YouTube just to make a fucking call. But no pro sound so far. JACK is great as a framework but the drivers for anything remotely professional suck big time. Yes, i know it's not the fault of Linux community, but who the fuck cares? I still can't use neither FFADO nor any of the USB drivers. I keep a Windows installation just so i can record stuff properly.

Graphics. Yep, nouveau is making progress (dunno what's up with ATI, i haven't used a single Radeon in years), but gaming and GPU acceleration on Linux still sucks balls. You know what? I've decided to stick with Intel integrated graphics for the foreseeable future. Yes, i got myself the newest ballsiest Ivy Bridge Core i7, an equally ballsy Z77 motherboard, and that's it. No discrete graphics, fuck em all.

Pro software. Suffice to say, the ones that are worth something either are/were commercial (think Blender), ugly, messy and useless (yo GIMP, how's your CMYK support?), or simply crashy or amateurish (pretty much everything else). And we still wonder why, oh why didn't Linux take over the world...

Rule 34

OMG Ubuntu reports on an Ubuntu Unity lens that allows searching for porn. Big deal, eh? Well, apparently, the world is in such a sad state that people have to post this stuff with a bunch of warnings and unnecessary remarks that are really stating the obvious (such as "whether it offends you or not", in bold), and people are still whining in the comments about how freedom somehow has to be magically equated to "not offending".

If porn offends you - it's your problem. Freedom is exactly that - freedom. Including freedom to do and say what others might not like or find offending.

KDE is the most Linux desktop

This article has caught my attention, mainly because it appeared both on Planet KDE and on Slashdot. The comment section of Slashdot was particularly fun to read, because it sort of reminded me what i have seen back in the day of KDE 4.0 release.

Basically, we have loads of users whining about how GNOME changes things in a way that they aren't where they used to be any more. And then there's Unity. Unity is another topic altogether, and while it may seem that its arrival was completely uncalled for (what? another Linux DE choice?!), that is actually not true. The real reason for Unity to appear is that GNOME2 sucked and the GNOME dev team is a bunch of assholes. Mark Shuttleworth embraced the coming of netbooks and understood that the touch-enabled desktop devices' arrival is inevitable, and he simply didn't want to wait for GNOME developers to pull their head out of their asses and look into the future, not into the past. So while Unity may or may not suck (earlier versions of Unity definitely sucked, didn't try later ones), it may or may not look like Mac OS X (latest versions definitely do share a lot of similarities), it may or may not be necessary after the GNOME Shell arrival, it was necessary in order for those GNOME dumb fucks to realize that backwards compatibility is not the only thing that matters.

And finally - thank God! - GNOME guys started thinking, and they pulled their heads out of each others' asses, and they started moving, and they released GNOME 3 and GNOME Shell. I didn't get my hands on them yet, but from what i see on the screenshots, they look pretty cool. No more Windows 98-ish look, GNOME 3 looks more like a mashup of Mac OS X and MeeGo. Which is good, because these are the interfaces people consider both beautiful and usable.

But what is it? Why GNOME users are whining? Aaaah, the desktop behaves differently from they are used to. Welcome to KDE 4.0 guys, now it's your turn. Pull your heads out of your asses (i like this phrase!), change is necessary. Face it - GNOME 2 has all but stagnated. It wasn't fun, it wasn't configurable. Some people confuse GNOME 2 limitations for minimalism, but whatever. Man up, idiots, if even GNOME dipshits decided that it's time to change - that's for your own good. Remember, GNOME knows exactly what every user needs. That's why they leave out so many configuration options, isn't it?

The problem here is that people are not against GNOME 3. They are against the change. Back in the day when KDE4 was introduced, it changed everything. It was practically rewritten from scratch. Why? So we could enjoy so many great software today. I think the real KDE 4 will be KDE 5. KDE 4, while being a major rewrite, has only laid the foundations. Hell, KDE PIM got Akonadi support only on KDE 4.6/4.7 - that's more than halfway through the development cycle. Despite it being a rewrite, it's still pretty much a long ass transition to the integrated, semantic future. GNOME guys are finally moving that way too with their Zeitgeist thingie. So essentially, things change. And people don't like it.

Look, here's one for ya. You know why GTK and GNOME sucks? Look how much time it took to write Unity. After two years it's finally usable. How long did it take for KDE guys to write two (!) releases of Plasma Active? Three months? Five? Moreover, Plasma Active is practically a project of a couple of weekend hackers. Unity has corporate backing of Canonical. Plasma, KDE and Qt in general are far superior in terms of getting working (and fast!) code as quickly as possible. KDE architecture enabled writing Plasma Active in so little time. KDE architecture enabled Calligra guys to write a backend, on top of which interfaces for both desktop and mobile can be built very quickly. GNOME doesn't have nearly as much power both from user and developer point of view, and because GNOME developers are idiots, GTK will never have that power.

But as much as i despise GNOME, i have to give credit for finally growing some balls and doing GNOME 3. Also, Vala is a terrific idea. Having a real object-oriented language which translates itself into native C code? Brilliant. Now developers can have the speed of native C and experience the full glory of ass-backwards GTK and GNOME architecture. The only thing i'm not sure of is how to debug this thing if it translates to C code? It sorta reminds me of Google Dart, only in case of Vala it actually tries to solve a real world problem of trying to write object-oriented code in a language that is more suitable for low-level system programming than for GUI stuff. Time will tell though.

KGet fucking sucks

Seriously, guys, is this what you call a download manager?

First of all, the why having two plugins to do the same thing? Yes, one of the KIO-based file downloaders is multi-threaded, but why keep the single threaded then?

Second, the behavior is destructive by default. Why does it delete everything instead of leaving the files and deleting just the download job? I've just lost 20 GB worth of downloaded data, thank you very much.

Third gripe is that not only we have two KIO plugins, they also both suck big time. I mean, no matter how much i hate the thing that Firefox has become, there's still a couple of things it does best in class. One of them is download management, with help of DownThemAll extension. This is hands down best download manager available in a cross-platform manner. JDownloader sucks and will always suck 'cos it's Java. Others either don't work on anything but Linux, or don't work on Linux, or even work only on Windows. And they all suck anyway. KGet can't download for shit. It can't parse download.php-type files and rename the files afterwards, so whenever you download practically anything which is not from FTP, you're pretty much fucked, because you'll end up with a bunch of download.php files.

And final complaint is that torrent plugin sucks fucking greasy nigger monkey balls. It can't even scan files if they weren't downloaded by exactly this instance of KGet. Try to continue download with another KGet from another Linux? No luck. Delete everything. Try to transfer your torrent files from KTorrent to KGet? No luck. Delete everything, start from scratch. This isn't even funny.

Thankfully, KTorrent is one of the best torrent clients available. God bless KTorrent developers, KGet developers - burn in hell, you fuckers.

Firefox fucking sucks

I use Linux as my everyday operating system. I love KDE, i love the developer culture of Free Software, i love lots of things about KDE, Linux and its ecosystem. But there are some areas where it not only lacks - it sucks monkey balls. One of these areas are web browsers.

What the fuck?! Web browsing is just about everything most people do with a computer today, and you still can't come up with a decent browser?! Let's see what we have. Internet Explorer? Only old ass versions, only on Wine, and they suck anyway. Safari? No Linux version, and Safari is probably the most inflexible piece of software ever written. Chrome? Hands down fastest rendering and fastest Javascript, no desktop integration whatsoever (looks and feels alien on ANY desktop), extensions suck, and Google is trying to do batshit insane crap with it, like pushing to make it a gaming platform, or allow it to execute native code, or other less than useful stuff. Opera? Bloated proprietary piece of shit (i don't know what do they need extensions for - they already have everything and the kitchen sink in there), sucks at rendering and sucks at Javascript, and hipsters love it. Firefox? Piece of shit with an asshat development team.

Don't get me wrong - there was a time when i was a fan of Firefox. It was a new kid on the block, it had tabs, it was small and fast, it blew IE out of the water, and it was cool to like it because it was all open source and ideologically correct. Yes, Opera had tabs too, but Opera, while being a proprietary piece of shit, sucked then, and it sucks now. Firefox 2 ruled with all its glory.

Things started to go south about the time of release of version 3. Firefox started sucking big time back then. The most unnecessary and idiotic "feature" ever introduced was the "awesome bar". Firefox developers were saying shit like "give it time", "you'll get used to it", blablabla. Well guess what, i didn't. I still despise this fucking "awesome" bar, i still haven't got used to it, and i still miss the old address bar. Sadly, with all the addon flexibility Firefox has, no one is still able to replicate the old address bar.

Firefox also started to be a memory and resource hog. Every time a new version came out, i was pleasantly surprised that it was faster than the previous one, only to discover that it was faster due to some addons being disabled out of compatibility issues. And it got all the bloat back by the time the next release was on the way. The development team was bragging about how they work on memory problems, how they optimize startup time, how they do this, how they do that - but Firefox only seemed to get slower and slower.

Another problem with current Firefox is the development cycle. No, seriously, what were you thinking?! You want to have a rolling release? Fine, but why inflate version numbers? Why not do 4.1, 4.2, 4.3 etc.? Current alpha (Firefox 10? 11? 28?) looks exactly like Firefox 4. So what was the point of all these version numbers? Firefox management didn't just shoot itself in the foot here - it blew out both its legs with a fucking bazooka. Sorry guys, rocket jumps only work in Quake. Not to mention the whole "user is an idiot" mentality that plagues Chrome developers. Admit it, Mozilla, Firefox 4 looks like Chrome.

The main problem with Firefox though is its complete inability to integrate with KDE desktop, which is largely fault of using this piece of shit toolkit GTK+. I still can't understand why those Chrome and Firefox idiots picked GTK+ instead of Qt. Maybe back in the day Qt wasn't a viable alternative, and Firefox being so old ass and having lots of legacy code would find the transition to Qt too expensive, but Google, why?! Why?! Why did you choose a toolkit that looks like ass on my desktop?!

Complete lack of visual integration is not the best part of it, though. The best part is, since i want to have my passwords stored in my KWallet, whenever the new shitty Firefox version comes out, it breaks the KWallet addon. I just can't live without KWallet integration, and this is the only thing that breaks on every fucking update!

It's been a while since i started looking around for better browsers, but they all suck. Chrome lacks DownThemAll, which i can't live without, and is not well integrated into KDE and looks and feels like ass. Konqueror just plain sucks. Doesn't render websites correctly, no extensions, Flash crashes, and due to its swiss army knife nature, whenever you press the wrong button, some lame-ass panel or viewer shows up which i don't know how to turn off.

Rekonq? I have high hopes for Rekonq. It already has KGet integration, which is sorta like having DownThemAll, but it's not nearly there yet. It doesn't store passwords in KWallet, which is really strange, given it's a native KDE project. It also lacks extensions. Some time ago, Chrome extensions were promised, but they didn't materialize. It also promises some interesting ideas like ownCloud sync, but again, it's not ready at the time of this writing. With KGet and KWallet integration though, i can live without most of the Firefox addons (and some of them are available as Chrome extensions as well), but unfortunately some extensions i would strongly prefer having, Ghostery being the prime example (as far too small amount support the Do-Not-Track option). Greasemonkey would be good to have too.

Update: Nope, it does indeed store passwords in KWallet, it just does it differently from Firefox, which is why i couldn't locate them in KWallet Manager - they're stored under Form Data. This makes rekonq barely viable an alternative to Firefox. I wish i could import passwords though. I might look into writing a password import script either from CSV, or from KWallet itself. Probably, CSV/XML would be a lot more useful for the community. I still miss extensions though, and while rekonq has its own flavour of AdBlock, i would really love to see something like Ghostery or Scrapbook.

So, what does it leave us with? Some lame-ass half-assed slow-ass unbranded-ass browsers like Arora, IceWeasel, Namoroka or whatever their names were? Or a semi-usable extension-less rekonq?

It's really simple guys. I just want a browser that integrates with my desktop, and that is reasonably fast and allows at least some degree of flexibility. This is what i'm not getting.

Why GPL doesn't suck

I have recently stumbled upon quite a few blogposts bashing GPL for being "less free" than BSD/MIT/whatever licenses.

The main argument is almost always the following: basically, GPL forces developers to release the source code. Whatever modifications they make, they must release the source "so the whole community benefits". This is somehow presented to be an argument against the "freeness" for GPL and GPL ends up being accused of hypocrisy. I see this as complete bullshit, and here's why.

In GPL, developers don't matter. It doesn't matter who wrote the code, and why did he/she wrote it. It doesn't matter whether they did or did not make money off it, it doesn't matter if they did it as a hobby or professionally, it doesn't matter if they existed at all. What matters is the code. It's not the work of developers that gets licensed, it's the code that gets licensed.

How does that possibly make sense and what implications does it have? The BSD/MIT/Apache/whatever license is aligned to the interests of developers. Ideally, a developer doesn't really care if his code gets seen, used, forked, relicensed, whatever, as long as he is attributed and given credit - in other words, developers often look at their work pragmatically. However, the GPL license is not aimed towards developers, it is aimed towards the code. The code itself, whatever the modifications, should exist and should be freely accessible, freely modified and freely used.

All the people that bash GPL for being "unnecessarily" restrictive, miss the point of freedom. There is no freedom to do whatever you want. There is only freedom to do anything you want as long as you don't infringe on other people's freedom. You have your rights, but you have no right to take away rights of others. Thus, when you close the source code - you take away my right to read, copy and modify the code. And while you as a developer might want to take away that right from me (or allow others to do that), that doesn't mean that you can have that kind of freedom.

Get your own cloud!

I've become involved with an awesome ownCloud project recently. Well, not really "involved", just hang around the IRC, report problems, suggest ideas, help other people to set it up (as i've set it up myself on my home server already) etc. Basically, ownCloud is a project aimed at those of us who doesn't want Google snooping around, who is reluctant to post stuff on facebook and who basically wants to have his/her data in his/her hands.

So? So. I have to admit that ownCloud is awesome. Though it is currently in its early stages of development, it already is pretty functional. I can upload stuff, access my files via WebDAV (natively supported by KDE's Dolphin), i can keep my contacts and calendar there via CardDAV/CalDAV, and listen to music. There are other, more advanced options like LDAP authentication but i haven't yet tested it so i can't comment on that one.

What i can comment on is how easy it is to set it up. The ownCloud setup itself takes a minute - in a sense, you don't really have to "set it up" - it just works. Setting up an environment to run that shit is also as easy as it can get. I admit that for some people setting up an apache/php server doesn't sound like it's easy to do, but trust me - anyone running Ubuntu will be able to set it up, given the right step-by-step guide. Hell, i never had a webserver before, now i'm running one.

Anyway, i've liked the ownCloud so much i actually already ditched Google's contacts and calendar altogether. I have installed a calDAV/cardDAV on my Android, also setting up Akonadi sync on my computers and now i have fully integrated, synchronized and backed up calendar and contacts, without ever giving up my rights to that data. Plus there's file sharing, so i'll start uploading stuff like my photos there at some point. I never liked the idea of music streaming so i didn't even bother to test it. However, i might change my mind later too.

With all that praise i'm giving it, it wouldn't be me if i never said something inappropriate about ownCloud. So here's the gory details. First of all, the settings area is a bit illogical. Why the fuck the admin section only contains LDAP? I'd expect to find all kinds of stuff there, like user management, installed apps, database backend settings and whatnot, but instead i get sections for "Apps", "Users" and a section for "Admin" that contains LDAP. Name it LDAP, for fuck's sakes!

Another gripe is that you have to go to IRC channel to ask for paths to use for CardDAV/CalDAV - there's not a hint in the GUI about that (at least i didn't find it). The "Help" area is bullshit and doesn't really help with anything. The "get more apps" link leads to a dark void. Overall, a lot of good ideas, but unpolished in these little things.

However, my final NERD RAGE dose goes to web browser developers. You're all fucking assholes. Firefox 6 renders like shit (funny thing is, Firefox 3 does a better job). Rekonq has some Javascript problems. Opera is completely and utterly unusable with this website. Google Chrome rules them all. As much as i hate Google for their decision to make their fucking browser use that abomination of a toolkit GTK+, it still blows everything else out of the water when it comes to rendering.

So, to conclude - will i use ownCloud? Definitely. Will your mom use it? How the fuck would i know. It's good enough for me.

In before the lock

Here we go again. Microsoft is known for pushing for dubious technologies to take over the world. One of those is this idea of "trusted computing" (implying that we ought to trust Microsoft). They were once pushing this TPM chip idea, and was planning to include ungodly DRM into Windows Vista (thankfully, they backed up from that). Now they're coming back with another "initiative" - lock down the boot process.

Now, i can understand the need for "trusted computing" in some environments. You might want to have locked down machines in government organizations or secure environments such as military computers. You might want your computer to prevent you from running anything that isn't blessed from high above. But what does it have to do with us, ordinary users? I want to run whatever the fuck i want on my computer! No matter what these people think, i will rip CD's and DVD's, i will share my music and videos with my friends, i will run whatever software i want to run, and i will boot any operating system i choose to boot, regardless of what Microsoft, MPAA/RIAA and other idiots think about that.

But what does it have to do with "secure boot"? Well, you see, the "trusted computing" is legacy entertainment industry's wet dream, allowing whole new level of DRM that is not easy to avoid. Again, i can see DRM being a valid technique in protected environments and generally in some niche markets, but it's got nothing to do with us, consumers. If Microsoft chooses to lock everything down instead of properly securing their OS, it shouldn't be my problem. As one of the commenters on Ars Technica has pointed out, this is Microsoft basically admitting that no, we can't secure the OS enough to prevent userspace from taking over the bootloader, so we will lock everything down instead. Brilliant approach.

I can see the general trend towards locking everything down. We are losing our freedom and control over the hardware for some fancy bells and whistles. You know what? Computer is a tool. It should not tell me what to do.

GMail sucks

Yeah, i said it! GMail sucks balls.

Now don't get me wrong. I am a long time GMail user. I use it since 2008 without any problems. I used it with web interface, i used it with POP and then IMAP accounts (via Thunderbird), and i am now using it with KMail. It always worked perfectly, until i decided that i am tired of having three different "sent" folders and sought to unify my devices.

And so, here we have it. I have a desktop with KMail2 on it, i have an Android phone with K9 Mail on it, and i have a web interface. I want to be able to see the sent mail regardless of whether it was sent using the web interface, KMail or K9 Mail. Previously, it wasn't possible, because webmail was stored in "Sent" folder on GMail, while the SMTP-sent mail was apparently stored in [Imap]/Sent folder.

Why the fuck would i want to have separate "Sent" folders for webmail and SMTP in the first place?! Well, i could understand that it might be some corporate needs or whatever, but why on earth would someone want that behavior by default, on a personal mailbox that is usually accessed from multiple devices by the same person and thus should have everything unified?!

Anyway, i managed to set it up the way it should be, but i encountered a few quirks. First of all, "[Google Mail]/Sent Mail" just looks ugly as fuck. But OK, that might be a fault of client software - after all, nothing stops programmers from giving folders arbitrary names while still retaining the "link" to the original SMTP/IMAP folder.

Then we have these labels. I can get the concept, but i can't get through my head their absolutely idiotic approach to labeling. Why "archiving" means removing all the labels? How can i restore the "Sent" mail from archive? Why is there an option to "archive" sent mail in the first place, because obviously sent mail can't really be archived (it still stays in "Sent Mail" folder)? Frankly, it took me whole evening to set everything up.

And finally, i would like to have an option to actually delete a message. Since my inbox becomes very bloated over time, i decided to archive my old email to my cloud. Naturally, if i archive the mail - i would expect it to disappear from GMail inbox entirely. But no! If i delete it from a folder - it stays in the mailbox! The only way to actually delete the message is to put it into trash first (so that it removes all the labels and excludes itself from GMail archive) and then delete it from there.

Frankly, i wish i didn't pick GMail as my email provider of choice. Yes, it's convenient, fast, reliable and spam-free, but so what? If it makes me jump through hoops just to set up a proper IMAP account - it's not worth it. I understand Google's desire to tie me into its web services, but i'm afraid it's not going to happen any time soon. I hate web apps, i always hated them and probably always will. I use Akregator to read my newsfeeds. I use Blogilo to write my blog posts. I use KMail to check my mail. I use KDE-Telepathy to handle my Google Talk and SIP accounts. Call me an old fag but i'm extremely uncomfortable with using a web browser to do something other than web browsing. The only reason i'm writing this in web-based editor is because apparently Google has changed its API again and Blogilo won't work with Google Blogger until they fix it.

Binary > source

I am currently using Linux (more specifically, Kubuntu 11.04) as my primary OS. I think it's a brilliant distro. Everything works out of the box (for me), wireless surprisingly works even better than it ever did under Windows 7. However, over and over i see this stupid argument every once in a while - (K)Ubuntu is bloated, binary distros suck, apt sucks, Gentoo/Arch/Slackware ftw etc.

Regarding bloat. My KDE/Kubuntu installation with a whole lot of shit on it works faster than my almost virgin Windows 7 ever did. I don't know what happened to it, honestly - at first it was pretty snappy and i was really happy with it after Windows XP (i skipped Vista). But over time it began being slow. I don't know what is the reason, since i usually turn off all that BitLocker, indexing and other useless crap that comes with Windows 7 Ultimate. KDE, on the other hand, even has full-text search enabled (Strigi), not to mention Nepomuk, Akonadi and countless other apps that are constantly running in the background. It still outperforms Windows 7, even with half-baked no-3D-acceleration nouveau driver.

As for apt - i can't really comment on that, since i only seriously used apt. I did try Gentoo at some point and have a slight idea of what emerge is, but i probably have no more experience with emerge than i have with RPM (i started out my Linux way with Mandrake 8). However, i can say that i never ran into any problems with apt. Of course, all the graphical package managers still suck balls when shit gets real and i have to resort to aptitude every once in a while (which isn't Ubuntu's or apt's fault anyway, should have fucked with alpha repos in the first place), but still apt rules. At least i can't say anything bad about apt.

And now to my core argument. Source distributions suck. Compiling from source sucks. Binary is the way to go. Why? Let's see.

I saw many people saying something like "i don't see what's the problem with running configure && make && make install". However, there are several major problems with that statement. First of all, having a compiler on my system is a potential security risk. Even on a desktop. Second, why do these people assume that every Linux user is a software developer? Why do i have to have a compiler, a gigabyte of development libraries and a version control system installed? It's fine with me, i don't have a problem with that, but your average Joe is not a developer, nor he wants to be one. Compiling from source sucks. It really only makes sense for the developers, because they can check out the fresh code, compile it, test it, debug it, whatever. It doesn't make any sense for the average user.

Finally, source-based distributions. They suck. I remember installing Gentoo on my old crappy laptop in hopes that the legendary "performance" and "customizability" will make enough of a difference to breath in a second life in that laptop that used to run Windows 2000. You know what happened next? I started installing software. It's really fun, you know. Open Office was compiling for 10 hours. 10 fucking hours!!! After i installed Open Office, XFCE, bunch of other stuff (another two days of compiling shit), you know what happened? The distribution just worked. It worked no better and no worse than, say, Xubuntu. Same shit. And believe me, i took my time and learned a lot about optimization flags, compiler flags and all that shit that Gentoo users fap to. It made no difference.

However, the worst thing was still ahead. I received an update notification. For OpenOffice. Another 10 hours of compilation? Thanks, but no thanks. And you know what i did? I said "fuck it". No more Gentoo. No more compiling. Binary distributions rule. You know why? Because they tend to work without spending endless hours compiling and optimizing to squeeze those extra 2% of performance. Of course, someone might find appealing to live on the bleeding edge of FLOSS, but i just want my computer to work without hickups. I am somewhat of a software developer, i have no problem reading source code or even compiling from source, but seriously, if i wasn't a developer - why would i have to have a compiler and development libraries on my machine?!

Skype, open up, or i'll kick your ass!

Some russian guy has claimed that he sort of reverse-engineered the skype protocol. Sort of - because he didn't completely reverse it, and it is an old version of skype, which isn't compatible with modern skype versions. He says the goal is to make skype open-source.

Now, i understand the motives. I think the guy is doing the right thing. All the marketing bullshit Microsoft is trying to sell us (IP infringement, spamming/phishing/scamming concerns etc.) is worth nothing. Actually, it makes me wonder - why are we having these news in the first place? What's the problem with open-sourcing the protocol?

Everyone knows skype on Linux sucks balls and has been in 2.5 beta for the last two years. Hardly anyone believes that it will improve much because of Microsoft's aquisition of Skype, given Microsoft's stance towards free and open source software. Everyone knows that skype is doing their business not on the protocol itself, but on third-party agreements (skype-enabled phones) and Skype-out services. If anything, open-sourcing the protocol would enable it to spread even more.

A pecuiliar detail is that the guy doesn't seem to disclose how exactly did he get the de-obfuscated Skype binary and all that code from VEST corporation. From what he writes on his blog it almost looks like he stole the code and the binary files, wrote a couple of lines of code and published it as if he did the majority of it. Of course, he doesn't claim that he did all this, in fact, quite the opposite - he repeatedly makes it clear that he actually has not done much himself. This raises another question, particularly that of his moral right to call his findings a "research" when in fact 90% (or even more) of the work was not done by him, and supposedly was not acquired through legitimate ways. However, let's leave it be for now. Microsoft will most certainly try to combat these efforts, but if Samba guys managed to pull off the reverse-engineering of MS protocol - there's no reason why anyone else can't pull of a similar stunt with Skype.


I think the problem is not the skype protocol itself. The first possible answer is purely business-based. If the Open-Source community will reverse-engineer the Skype protocol and open-source it - what will be the point for third-party phone manufacturers to make an agreement with Skype when they can take an open-source client and incorporate it into their phones without paying Skype? That will surely hinder a chunk of Skype's revenue stream.

Another possible answer is that they fear that the guys would uncover something in the protocol that is not very pleasant for the public to know. I don't remember exactly where and when i read this, but i remember reading about some US laws actually requiring the telecom providers to leave the backdoor open for government agencies to eavesdrop on conversations.

Now, i was looking for an alternative to Skype for a long time. Since i use multiple OS's, something cross-platform would suit me best. And i remember why i stayed on Skype in the end. I couldn't care less about HD video calls, conferencing and skype-out - all of this is available in the Open-Source domain as well - Ekiga being a good example. The reason why i still use Skype despite all its deficiencies, despite it not being open and all, despite Linux version sucking big cock, despite recent Windows version being bloatware, despite its Android version being incomplete (no video calls, no SMS), despite Skype not being the cheapest VoIP service, is simple - its superior firewall and NAT penetration capabilities. I simply can't connect to SIP network using Ekiga, while Skype always works flawlessly.

However, if the Skype reverse engineering project achieves its goal, switching to Ekiga or any other VoIP/SIP provider might not be necessary.

GNOME

I was thinking of a suitable topic to start my whiny blog, and i thought that i would slowly introduce my area of interests.

Now, i am a Linux user. I have been a Linux user for quite some time (with varying degree of actually using Linux as my everyday OS). Moreover, i can consider myself a somewhat advanced Linux user. Although i don't do anything k00l and l33t anymore (well, aside from using my USB drive as a login and sudo device (e.g. without it no login, no sudo), if that's considered l33t enough) but i used to do all sorts of things a casual user wouldn't think of doing - setting up a server, using remote access, recompiling kernel, writing complex scripts to automate some tasks...

In my trials and tribulations i have gone through many distributions, ranging from Mandrake 8.1+, (K/X)Ubuntu 6.06+, Gentoo, Debian Etch et al and upwards to things like FreeNAS and DSL. To put it another way - i've seen them all. Well, not all, but a fair share of what Linux world has to offer. I've used countless WM's ranging from ominpresent Metacity and KDM and down to Enlightment, *boxes and practically any other WM you can imagine. The search for perfect desktop was long, but in the end i came back to where i have begun (well, not exactly, but still).

However, i wasn't really serious about using Linux until a couple of years ago. Back when i was fiddling with it, it was fun - i was a fearless explorer diving into SSH, command-line, trying different things and bending Linux backwards just to see if it breaks. It didn't, and that's what i like about Linux to this day, it hasn't changed. The core system never fails. Ever.

On the GUI side, things were not so bright. Although it was quite k00l to send commands over SSH and doing stuff i never dreamed of doing, i wasn't really trying it out in a regular, casual use department until a few years back. You know, every day stuff like browsing, checking emails, using IM and listening to music. When i did, i quickly found out that the Desktop side of Linux is far from ideal. The majority of software was written with GTK (the GIMP toolkit), with KDE lagging behind and other stuff (wxWidgets or just directly X) being in the ass. Of course, with the introduction of KDE4 the situation is finally changing.

Which brings me to my main point. GNOME. GNOME SUCKS. Literally and metaphorically. GNOME sucks, GTK sucks, GIMP sucks, and every bloody piece of software that uses GTK wil forever SUCK until they switch to something more sensible. And for the sake of simplicity, i will use GNOME and GTK terms interchangeably since they really are part of one big crappy framework.

Why, you might ask? First of all, it's ugly as hell. Windows 98 with rounded edges. Yes, i'm not afraid of CLI, but i value great looking sleek interface. I am one of the guys that doesn't care about system performance since i usually use up to date hardware (as opposed to running Linux on a typewriter) and don't jack off on benchmarking results - as long as it's responsive enough it's fine with me. I never liked Beryl/Compiz and i hardly ever use ANY desktop effects (aside from window animations), so you can stick all your GTK+Compiz arguments up your ass. GTK is ugly, get over it.

GTK is crap. It's written in C and contains whole fucking load of legacy code and features that go back as far as 1998 (i mentioned Windows 98 for a reason). KDE4 API proved that good looking, feature-rich object-oriented C++ API that is easy to use for a developer can exist without too much C crap. It's not that C is a bad language, it's just that IMO it is not suitable for userspace tasks like GUI API. Drivers, kernel - anytime, GUI library - fuck you. WinAPI is a nightmare, GTK is not any better.

Have you actually noticed how crappy looking GNOME icons are? Again, Windows 98. For fuck's sake, Windows XP had a similar style "plastic" icons, but they looked WAY better back then than GNOME icons look NOW, 9 years later! What the fuck is wrong with GNOME people?! They don't care about innovation, all they care about is backwards-compatibility. Innovations always come at a price, and that is a price i am willing to pay - if something is broke during the process i'll just wait for an update, i don't want legacy code holding me down. But apparently, GNOME guys think otherwise.

There's lots of great software written with GTK and it's really a shame. Before you start saying that it doesn't matter which toolkit they use, let me give an example of truly HORRID GNOME crap - the file selection dialog. Even Windows 98 got it right! Hell, Windows 95 had a better one! Mac OS always had it! KDE got it right from the start! And GTK is now at version 2 and they STILL haven't figured out that this file selection dialog just plainly SUCKS BIG FUCKING MONKEY BALLS?!!! Firefox! OpenOffice! Chromium! Thunderbird! Eclipse! They are all GREAT pieces of software, and they are spoilt with this GTK crap - this is just plainly stupid! I would resign from using ever GTK just for this very reason! But what amazes me even more is WHY do all these people use it for development?! I understand that OpenOffice, Firefox et al have a lot of legacy code so it would probably be a lot of rewriting just to get rid of this GTK crap, but things like Chromium?! I thought Google were smarter guys...

Thank God for KDE4. The guys weren't afraid to break everything in favor of laying the foundation on which they can build a desktop better than it is now. Plasma workspace is a brilliant idea! Plasmoids are a brilliant idea! Everything is done RIGHT - everything is flexible enough to fit in anywhere - regardless of display resolution and size of the toolbars, and this flexibility is not because the developers of the plasmoids are smart, but because the very API they use allows this kind of flexibility without any extra effort!

And, to conclude - the flexibility of GNOME. GNOME is notoriously inflexible when it comes to configuration. This might be a good thing at first, but after a while GNOME gets boring. It takes away the excitement of learning something new because there is really nothing to learn.

I will be watching the GNOME shell project closely, but unless they change the underlying API (i.e. GTK) it is doomed to suck too.