Is SaaS Killing Native Linux App Development?

jfruhlinger writes "In a world where 'app' is the new buzzword, the development of native Linux apps is lagging. Some of this can be attributed to the usual community infighting (the latest version of which is argument about Ubuntu's Unity interface), but there may be something deeper at play: Linux advocates have for so long advocated browser-accessed software as a service as a way to break out of Microsoft's proprietary desktop. Now that this world has arrived, there's less incentive to work on native Linux apps. But of course, entrusting your functionality and data to a cloud provider like Google has its own set of concerns for free software fans."

Comments

SaaS = Vendor Lock-In As A Service

[Anonymous Coward]

The hardest most expensive portion of SaaS is hosting it. That's why native apps are lagging.

Re:SaaS = Vendor Lock-In As A Service

[SharkLaser]

Which is even more true for Google's services. At least when I buy Office from Microsoft I know the software will continue working and they don't care about datamining me, as they got my money already. Google can, and does, shut down their services at will. It's DRM^2^2.

Re:SaaS = Vendor Lock-In As A Service

[ArhcAngel]

Depends on which version of Office you are referring to.

Re:SaaS = Vendor Lock-In As A Service

[pentadecagon]

Could you be more specific what's wrong with Google's offer? So far Google didn't do anything wrong, so no reason to assume they will. They provide a free service, they provide you options for downloading all your stuff, there is no lock in. And about data-mining: They probably do, but this doesn't hurt you at all. And as long as they don't do anything bad or wrong with all that information you don't need to care. And so far there is no sign of them doing the wrong thing.

check out this awesome song

[Anonymous Coward]

put on your best cookie monster voice and sing along with me!

D is for download, that's good enough for me.
D is for download, that's good enough for me.
D is for download, that's good enough for me,
oh! Download download download starts with D.

hang on why are we singing about downloads again?

oh yeah linux software :)

my damn alzhbergers is playing up :(

Re:check out this awesome song

[ledow]

Freshmeat is now called freecode.

Re:check out this awesome song

[Anonymous Coward]

And what the hell was the guy who came up with Freshmeat thinking? LOL

Re:check out this awesome song

[Chrisq]

Great stuff, As a vegetarian I always had a "yuck" factor when going to that site.

Re:check out this awesome song

[SharkLaser]

Meat is actually extremely healthy food. It's low on carbs, it has the good fat, and is high on protein. Bacon, eggs, whole-meat sausages and steaks are the same. You get a much healthier lifestyle if you avoid eating carbs. Yes, it means giving up on french fries, hamburgers and pizza, but that's easily solvable by making more bacon. I lost almost 30kg that way and still felt good while in low carb diet, after the first week anyway.

Re:check out this awesome song

[f0rk]

You get a much healthier lifestyle if you avoid over eating or over doing any of the three main nutritional groups.

There, fixed that for you.
Carbs ain't bad for you. Shit loads of carbs is. To little carbs is also bad for you.

Btw, there's more stuff in meat then just protein, carbohydrates, and fat. A lot of it is carcinogenic, even more in cooked meat. To much meat ain't good for you either. A balanced diet of everything is the best.

I'm a vegetarian my self, and restrain my self from eating meat out of political reasons (i'm an anti-consumerist). I can get all the nutrients I get from meat, from veggies. It just requires you to do some reason your self.

The problem I see (and experienced*) with a special lower-then-normal-carb diet, is that as soon as you hit your desired weight. You will go back to your old habits.
This is cuz' a diet ain't permanent, only temporary, to solve a temporary problem you have with your weight.

What I'm trying to say is that it's a typical scenario of "Don't medicate the symptom, medicate the cause". Special weight loss diets are medication for your symptoms of overweight, not your problematic lifestyle. Think about it, your not going to WANT to eat Weight Watchers crackers for the rest of your life.

* I went from 130kg to 80kg with a low-carb diet in just under a year. Then jumped back to 120kg within 2 years. Now I'm at 75kg, cuz' of a proper lifestyle, free from over sweetened food, additives, and cheap meat (If you want to eat healthy meat, it will cost you. It will cost you a lot, cuz' proper meat ain't cheep).

Re:check out this awesome song

[angel'o'sphere]

You get a much healthier lifestyle if you avoid eating carbs.

That is bullshit.
Eating to much meat may make you loss weight, but is in no way healthy. Meat is actually extremely healthy food No, it is not. Especially not the meat you get in our days in the super market.
To be healthy you need a balanced nutrition. To get an idea I would suggest to read an actual nutrition science book instead of a magazin diet. However as long as you eat enough / the right vegetables to your steaks you get enough carbs anyway (without you noticing ;D )
Stop drinking super sweet cokes and othere bullshit, and especialy don't eat diet shit with no fat an no sugar but full with artificial sweeteners.

Re:check out this awesome song

[somersault]

If you went back to your old habits after knowing the weight benefits as well as simply how good it feels to be eating a healthy diet, that's your own fault.. I started eating low carb (and then just "low GI" when I was thinking I was getting too light) after I had already lost 9kg (from my lifetime maximum of ~85kg), and I'm still eating only whole carbs. I don't see any benefits to eating white flour based or sugary foods. To those who know how bad sugar is for you to then go and keep eating sugary junk is as bad as being a smoker, heavy drinker (light drinking is actually healthy), or drug user IMO.

Re:check out this awesome song

[TheLink]

A lot of it is carcinogenic, even more in cooked meat.

A lot of plants we eat have poisons to stop stuff from eating them. Chocolate, coffee, onions, garlic, various beans and many others. Even potatoes can be toxic.

Just because stuff is listed under GRAS (generally recognized as safe) doesn't mean they're actually as safe as other foods and drugs that have to pass the tests. They can be listed just because of:
http://edocket.access.gpo.gov/cfr_2009/aprqtr/21cfr170.30.htm

(c)(1) General recognition of safety through experience based on
common use in food prior to January 1, 1958, may be determined without
the quantity or quality of scientific procedures required for approval
of a food additive regulation. General recognition of safety through
experience based on common use in food prior to January 1, 1958, shall
be based solely on food use of the substance prior to January 1, 1958,
and shall ordinarily be based upon generally available data and
information. An ingredient not in common use in food prior to January 1,
1958, may achieve general recognition of safety only through scientific
procedures.

There's usually little profit in finding out that common popular foods are actually not that safe (I mean in actual proper live studies - not in-vitro studies which seem to be able to show almost anything you want ;) ).

As for vegetarian diets, there's plenty of scientific evidence that show that humans do better on diets that contain fish. Just too bad many fish are getting endangered and more toxic (mercury, pcbs etc).

Re:check out this awesome song

[hairyfeet]

Or here is a thought, just throwing it out there, how about everyone leave everyone else alone and we all take personal responsibility for our actions, how about that? you may think this is just a healthy argument but you're wrong because i've seen where this goes. The smokers turned out to be the canaries in the coal mine, as we now have nanny types advocating cranking the taxes up on sugar and red meat, what's next you getting weighed and being fined by the state if you don't match their target weight?

I so miss the days when people actually stood up for each other instead of acting like a bunch of whiny bitches trying to control each others lives. the latest excuse is "Oh you'll burden us all with increased medical costs" while fucking ignoring Jose and his 14 kids that snuck across the border and are now getting food stamps,welfare, medicare and everything else. Oh and don't say its bullshit because I know a gal that actually has to go out and check on them as part of her job with the state and she is sick of seeing whole houses filled to the brim with those who just snuck over the border as late as last week and who ONLY know how to say in English "How do I get check?". And of course she ain't allowed to say shit because according to her boss "its politics'.

But if the nanny state is so damned worried about me possibly costing them a dollar, even though they can shit money out to the whole damned third world, every "el presidente" that will toe the line AND half of Mexico, I propose this: I will sign an ironclad contract that says if ANYTHING bad happens to me the ONLY treatment I'll get is morphine which is dirt cheap, and in return you leave me the fuck alone and exempt me from your sin taxes, do we have a deal? I have actually walked up to politicians and proposed this and they hemmed and hawed and in the end it all came down to "We know what is best for you". Well kindly go fuck yourself state, you can't even balance your checkbook or keep from taking bribes long enough to do something about a sinking economy, I frankly don't trust any of you to be able to even tie your damned shoes!

As for TFA, the economy is down and Linux is going down, surprise surprise. programmers actually have bills to pay too ya know and while RMS seems to believe a communist utopia is possible (of course most of us aren't allowed to be permanent squatters at MIT, don't take my word for it, look it up, that is how he describes his living arrangements) everyone else out there is worried about keeping a roof over their head or having their job shipped to Bangalore. So no shit a software whose philosophy is "free as in beer!" is having trouble getting new programmers, they are all trying to keep their heads above water like everybody else. But hey in 25 to 30 years when the economy isn't a corpse anymore (look up the last depression, it went from 1929-1953, that's 24 years folks) I'm sure the programmers that have jobs and money won't mind donating to the cause. its just right now they are trying to keep from ending up living under an overpass and if SaaS or MSFT or frankly anybody else is offering more than Linux, which is "Thanks for your donation citizen!" then duh! That's where they are gonna go.

Re:check out this awesome song

[CronoCloud]

Probably played too much Diablo:

"Ahhh, Fresh Meat" --- The Butcher

Re:check out this awesome song

[cyber-vandal]

Is there a command line in Windows?

Re:check out this awesome song

[Kjella]

Pure meat, that is not mixture products with a ton of fats and whatnot is not particularly unhealthy. Unless you choose to fry in a ton of butter and dip it in a ton of BBQ sauce or other fat stuff, of course.

As for the diet shit, yes thank you. No fat, no sugars, maybe it'll fuck up something else but as far as weight goes it's working perfectly. I lost a lot of weight while drinking a lot of diet coke.

Re:check out this awesome song

[Anonymous Coward]

Haven't more jobs gone to Shanghai/Beijing/Suzhou/Shenzhen/... than to Bangalore? Aside from some clothes (which happen to come from every country - Honduras, Sri Lanka, Cambodia, et al), I don't see too many things 'Made in India'. 'Made in China' is of course ubiquitous.

Re:check out this awesome song

[unixisc]

There is, but one has to look for it, or know it in advance. Either Start -> All Programs -> Accessories -> Command Prompt. Or Start -> Run... and then type cmd.

Re:check out this awesome song

[hairyfeet]

The better and more telling question would be thus: "If you remove command line from the OS, will it continue to function and will the user even know it is missing?" and the correct answer for Windows is YES as its not a vital subsystem and NO a good 98% of the users will NEVER know it is even missing. Now try that in Linux, go on, I'll wait...the OS died you say? and if anything goes wrong without CLI you have no prayer of fixing it? welcome to reality.

Here is a NASTY bit of truth that is guaranteed to make Linux users foam at the mouth but it is a fact, ready? In its current form Linux IS Windows 98, no more and NO less and I shall elaborate. What was Windows 98? it was a CLI OS (DOS) with a GUI shell, that's all. While the GUI shell was fine for basic tasks one often had to drop down to CLI to get anything complex accomplished, and there was even a way to bypass the shell completely and only use the CLI underneath.

Now what is Linux? Its a CLI OS (Bash, Korne, Bourne, you do get choices on the CLI, kinda like how you could run different DOS under Windows 2.x) and like Win9x it has a GUI shell. Like Win98 the CLI is a VITAL SUBSYSTEM and can NOT be removed because like Win98 the GUI is just a shell, it isn't vital, in fact one can run the whole thing without the shell even being loaded.

While you might consider this choice I am about to point out why this is actually a handicap that ensures that Linux in its current form will never gain squat in the vital consumer market. You see by having the GUI be a second class citizen this "cripples" for want of a better term, anything to do with the GUI. a developer can build for the GUI, he can build a barely functional GUI shell, he can ignore it completely. How many programs have you seen in Linux where there are 15 pages of CLI options and only half a page for the GUI? I bet you can find quite a few, in fact i'd bet there are more that meet this description than that are the other way around. hell you'd probably be hard pressed to find even 5 programs that have more options in the GUI than the CLI.

But the nasty truth nobody wants to admit is you'll never get home users to run CLI because as far as they are concerned that is a huge step backwards, and they would be right. If you want to have a CLI that is fine, hell with PowerShell (which isn't shipped by default with a single home machine that I'm aware of) you can script the whole damned thing, local or network. but this is if you WANT TO but it is never a HAVE TO. With Linux it is the opposite because Linux is still stuck in the "GUI optional" phase.

So like it or not what I said was the truth, as far as users are concerned there is no CLI in Windows. It is buried in the bowels of the OS, they'll never see it, never look for it, and never call on it. Can you say the same for Linux friend? can you HONESTLY say that if you made a Linux distro with NO CLI that it would be able to remain functional while updating from one of the main distro branches? hell would it even boot? probably not.

Re:check out this awesome song

[angel'o'sphere]

Pure meat is particular unhealthy! I simply don't get why you think different. Go read a book about nutrition. At least as I understood you, you mainly eat meat and nothing else. Humans are not made to live from meat. Except for some races like Inuit and to a small extent Maori. Meat is incredible hard to digest and creates "poisons" or in other words lots of nasty side products which make you age faster and more likely to get cancer.
Yes, you might lose weight and feel better if you have been "fat" before. But that has nothing to do with "healthy". More than like 300g / 400g (yes roughly a pound) meat per week is NOT HEALTHY. That has nothing to do with the meat itself, it has some vitamines, you need proteins ofc, it is mainly water anyway, but with our digesting apparatus. Neither Liver nor Kidneys nor your Pancreas are really happy about to much meat.
That yo lost weight while drinking diet coke is cool for you. Most people I know that drink diet coke are fat, really fat. Slim people neither drink normal coke nor die coke. But as you said you lost weight, it seems your body can distinguish between digestible sugar and non digestible suger pretty well and is not running mad in spilling out insulin.

Argument about Unity?

[Chonnawonga]

There is no argument about Unity. We all agree that it sucks. There is minor disagreement about the degree to which it sucks.

Does that really stop anyone from writing Linux applications?

Re:Argument about Unity?

[0123456]

Does that really stop anyone from writing Linux applications?

Developers don't really like using a tablet GUI on their desktop machine.

Oh, but at least Windows 8 will level the playing field there...

Re:Argument about Unity?

[Chrisq]

There is no argument about Unity. We all agree that it sucks.

The guy that wrote it probably doesn't

Re:Argument about Unity?

[Anonymous Coward]

I've used Leenooks since downloading Slackware 1 on 20+ floppies some 17 years ago or something. I like Unity. Fook you.

Re:Argument about Unity?

[Anonymous Coward]

Unity is used by one (1) distribution. If you don't like it, switch to a different one. I did; it isn't hard.

Re:Argument about Unity?

[slydder]

I'm betting suicide.

How did Sam Kinison put it... "I hope you slide under a gas truck and taste your own blood!!!!!"

yeah. if not then hopefully somebody helped him. ;)

Re:Argument about Unity?

[0123456]

Unity is used by one (1) distribution. If you don't like it, switch to a different one. I did; it isn't hard.

The problem is that the other distros are switching to Gnome 3, which sucks just as much but in slightly different ways.

Re:Argument about Unity?

[diegocg]

Many people like it.

Re:Argument about Unity?

[F.Ultra]

Actually no, there is a huge group of people who like Unity and a probably even bigger group who don't care.

Re:Argument about Unity?

[Anonymous Coward]

Oh, and also the 90%. Because you're a fucktard that dislikes any change, doesn't mean everybody is. You're free to go to Gnome3, KDE4, Xfce, LXDE, Enlightment and who knows what else, but yet you choose to complain about Unity. It's more fucking straightforward than any of those!

Re:Argument about Unity?

[Anonymous Coward]

I figure Debian stable will keep using Gnome 2 for at least two more years and by then somebody will have forked it.

I suppose I can't complain about Unity too much; I switched from Debian to Ubuntu way back when it was new because I was annoyed at Debian's "slow update cycle". Guess I got what was coming to me.

Re:Argument about Unity?

[Anonymous Coward]

That's because he's using it on a tablet.

You'd think they would have left the desktop alone and forked Ubuntu to Tabuntu with Unity on it for tablets. On a desktop Unity is a joke.

Re:Argument about Unity?

[heathen_01]

You can be fairly sure that Kubuntu will not switch to Gnome3.

Re:Argument about Unity?

[rubycodez]

Plenty of distros are getting into other desktops due to concerns about GNOME, KDE, Unity. Debian has LXDE and Xfce4 version, Linux Mint has xfce, just for a couple examples. There are other awesome desktops too.

Re:Argument about Unity?

[Anrego]

Sure, but KDE4 is just as bad (or at least was when I tried it, but based on what I hear, has improved none).

The two go to desktops for new users have completely failed .. and unity..

I'm happy with openbox, but I recognize that not having a serious desktop environment to point at that doesn't suck is a big deal for drawing in new users.

Re:Argument about Unity?

[mikechant]

I figure Debian stable will keep using Gnome 2 for at least two more years

And RHEL/SL/Centos should keep using it for *six* years (if you can put up with an RPM distro).

Re:Argument about Unity?

[Stewie241]

I use Unity on my laptop (which serves as my desktop) and I don't have any major issues with it.

I really don't understand what the fuss.

I mean, I got people like my parents who don't like change at all (they get uneasy if you move the icons on their desktop), but I find it surprising seeing the complaints come from IT professionals.

I also get it not being everybody's first choice. But the animosity directed towards it seems unreal.

Re:Argument about Unity?

[junk]

I can't find any evidence to support that statement. Shuttleworth didn't write it, he just blows a lot of smoke about it. I would love to hear a primary dev on the Unity project talk about it and then do a Q&A. Unfortunately, Shuttleworth being the mouthpiece is received as well as Lars Ulrich being the voice of the RIAA.

Re:Argument about Unity?

[lvxferre]

Say hello to Xfce. And dream as many of us that old GNOME 2 code will make its way to Xfce as xfce-goodies-2.

Re:Argument about Unity?

[Anonymous Coward]

I was disgusted with Ubuntu 11.10 because:

* It removed the useful icons from the toolbar, where they are never hidden behind windows (there is a Gnome extension that is supposed to fix that, but it didn't work on my system)

* It replaced OpenOffice, which works fine, with LibreOffice, which doesn't work with several of my spreadsheets

* The Unity interface is a mess, with everything that used to be in orderly menus now piled haphazardly into a "Dashboard" where it's hard to find

* Even the Gnome shell hides the former System menu under Applications, making it more of a nuisance to get to

So I ended up going back to Ubuntu 10.10. I will not be upgrading the OS again.

I am absolutely sick of brain-dead developers who insist on removing useful features, and ignoring the needs of the Linux community in general. When I next buy a computer I will not be installing Ubuntu on it, unless I find no distributions that suit my needs better than 10.10.

Re:Argument about Unity?

[discord5]

We all agree that it sucks. There is minor disagreement about the degree to which it sucks.

Remember when we had arguments about X being better than Y? I miss those days...

Does that really stop anyone from writing Linux applications?

I'm actually writing an application that makes Unity hate posts on slashdot. It works in Linux.

Re:Argument about Unity?

[TheLink]

Yes, if you actually want many users.

Re:Argument about Unity?

[DuckDodgers]

Unity was unusable due to crashes for me in Ubuntu 11.04. In 11.10, it's stable. Now that it doesn't crash, I like it a lot.

I suspect there are a lot of satisfied Unity users out there that just don't feel inclined to get involved in flame wars over what they perceive as a non-issue.

Re:Argument about Unity?

[Anonymous Coward]

I like it. Your argument is invalid.

Re:Argument about Unity?

[AngryDill]

KDE4 is just as bad (or at least was when I tried it, but based on what I hear, has improved none).

You might want to give it another look. My experience (which seems common compared with others ' remarks) is that it improved with each point release to being quiet stable and usable around 4.5 or so.

Re:Argument about Unity?

[WaywardGeek]

I hated Unity in 11.04 and am glad to be rid of that version. I currently use Unity 11.10 2D, with the Compiz window manager and a couple of config hacks to make it work better with Unity 2D. Honestly, it's pretty cool. It's annoying having yet again to learn all the new keystrokes to launch stuff, but oh well... at least the new system looks pretty good.

Re:Argument about Unity?

[WaywardGeek]

Well, I run on a small screen and have poor eyesight, so Unity in 11.10 is a welcome release. The auto-hide feature in Gnome 2.0 always sucked so badly I always had to turn it off. A large icon bar on top and a large bar for app icons on the bottom really squeeze the vertical usable area, especially when we we haven't even added the window frame, menu tool bars and status bar. Having it on the left and with a decent auto hide is awesome. They even integrated the menu bar into the window title bar, and integrated that into the top bar with the system menu and system tray icons. Very well done, IMO. I'm able to launch apps faster than ever, by typing the Windows key and then one or two letters and hitting enter, just like I can in windows. If Unity didn't include this feature, i would have sorely missed this from Windows 7. Instead, Unity added the dash, which is as far as I can tell, a nice improvement over what Windows does. With this desktop, I can use the same window manager on a netbook as on a desktop. Kudos to Ubuntu for a decent unity in 11.10, IMO. They even have a decent effort for accessibility in progress, and so far as I can tell are taking community feedback, bug reports, and patches very seriously. The port to Qt paves the way for a decent tablet play, though GNU/Linux has many bigger problems to fix to before they'll have much presence there.

Re:Argument about Unity?

[unixisc]

If you don't even like 4.7, go to Trinity. Or better still, pick something that offers you the choice of Window Maker or Etoille.

Re:Argument about Unity?

[WaywardGeek]

Actually, I rather like the latest incarnation of Unity. A nice improvement over Gome 2 Shell, IMO. That's not why people don't write Linux apps. Imagine trying to write the super-popular "Fart" application, which prints money on Android, for Linux. First you write it. Take maybe a week to do it well. No biggie. Now what? You can put it on Sourceforge and watch 10 users per month download it, or publish it on Freshmeat.net, and watch downloads go to 10 per day for a couple of days. What if you want real exposure? Well, the distros are binary incompatible and massively fragmented, but there's a decent number of Ubuntu users. If you want to reach them, the easiest way is to get into the Debian disto and wait a very long time for it to show up in Ubuntu. Now this will naturally turn off your Fart app developer, but even if he wanted to be in Debian, he now has to go through hoops that take more effort than developing the app did. And even after that, there's no way the Fart app will attract a Debian sponsor.

So, the real reason there aren't many Linux apps is because in most cases it's literally impossible to get them to users. Pretty simple, really. Maybe someday soon someone in GNU/Linux land will wake up and realize they have a problem. What we need is a cross-platform app-store that includes a packaging solution that automates building packages for Windows, Mac, Linux, and Android (sorry iOS...). Users should be able to use the same app store on all platforms to find and install the apps they want, whether paid or FOSS. If that happens, Linux apps would skyrocket.

Re:Argument about Unity?

[Yvanhoe]

You know these things are open source, right ?

Re:Argument about Unity?

[JanneM]

As a developer, I rather like it. The integrated top bar gives me another vertical line for code and text, and the pp-out scrollbar gives me another horizontal position (gvim isn't adapted to it yet, unfortunately).

It does need some tweaking to be good for a power user, though, and I think Ubuntu is making a big mistake in not exposing those settings properly. You have to google for a lot of those settings and they're not obvious. That's really the main issue, not Unity itself.

Re:Argument about Unity?

[unixisc]

Yeah, but not everybody knows programming. They have to search for something already out there.

Re:Argument about Unity?

[rdnetto]

There is no argument about Unity. We all agree that it sucks. There is minor disagreement about the degree to which it sucks.

Does that really stop anyone from writing Linux applications?

I'd say it does. Until now, Ubuntu has been regarded as the de facto OS for new users, and accordingly, developers have targeted it for development (along with maybe Fedora cause they like it). More capable users generally don't have any problems with downloading, compiling and installing the source. Now we have to seriously consider which distro to target, and the decision isn't as easy.

Re:Argument about Unity?

[DuckDodgers]

If you sudo apt-get install compiz compizconfig-settings-manager you can change the shortcut keys.

Re:Argument about Unity?

[vinitagrawal]

Unity is used by one (1) distribution. If you don't like it, switch to a different one. I did; it isn't hard.

The problem is that the other distros are switching to Gnome 3, which sucks just as much but in slightly different ways.

whats so bad abt Gnome3?

Native GUI app development is a pain

[ickleberry]

Due to lack of good tools. With MS Visual studio / VB any old monkey can make GUI apps easily, with Linux its not that easy There are plenty of GUI creation kits out there for Linux apps that are

*Easy to use
*Widely supported
*Actively maintained
*Designed for use with a somewhat mainstream language

But it seems to be a case of "pick any 3", or sometimes only 2.

Re:Native GUI app development is a pain

[Anonymous Coward]

Nonsense! Qt Creator is a perfect counterexample!

*Easy to use: Check
*Widely supported: Check
*Actively maintained: Check
*Designed for use with a somewhat mainstream language: Ch... well... ahn... wait, but what is a "somewhat mainstream language" anyway?

Re:Native GUI app development is a pain

[Anonymous Coward]

Qt Creator works well with C++ and Python, I'd call those both pretty mainstream.

Re:Native GUI app development is a pain

[Anrego]

This is exactly what I came here to say!

Much as I dislike Windows, I have to admit that visual studio is a very well put together package and lets you hammer out those GUI apps very quickly. We do have some ok-ish GUI designers (like Qt-designer) .. but nothing with the kind of "drag a button to the screen.. double click.. insert logic" workflow that VS has.

Re:Native GUI app development is a pain

[TheGreatOrangePeel]

Another vote for Qt. Also, if you can show me an IDE that is easy to use, widely supported, actively maintained, but isn't using a mainstream language, I'll show you open-sourced VB6 that never actually went open-source.

Re:Native GUI app development is a pain

[Anonymous Coward]

http://qt.nokia.com/products/developer-tools/

Re:Native GUI app development is a pain

[Ogi_UnixNut]

wxglade works well with me, and being XML based can work with any language you want (and uses native widgets on all the OS'es, so it always looks well integrated). I use it with Python primarily, but it should work for anything (look on their site for officially supported languages).

Have a look here if you're interested: http://wxglade.sourceforge.net/

Re:Native GUI app development is a pain

[characterZer0]

Eclipse, Netbeans and Java.

As a bonus, your programs will work on Windows too.

Re:Native GUI app development is a pain

[Anonymous Coward]

Personally I like Qt which I'd say ticks all 4.

Re:Native GUI app development is a pain

[Windwraith]

Really? I always thought developing in Linux was way easier, and I am no software engineer. The few times I tried to compile Win32 binaries I found it painful and slow...

Seriously, just with Geany and a few packages and I am ready to go. That takes like 3 minutes during a Linux install.

Then again I am a C/Lua/SDL/GL guy, maybe that's why.

Re:Native GUI app development is a pain

[zlogic]

Qt is an incredible toolkit. The only problem is that it looks "different" in Gnome, but about a year ago this problem seems to be solved.
Qt Creator is one of the highest-quality IDEs - very easy to use, powerful and not getting in the way.
Plus as a bonus Qt apps can be easily ported to Windows and MacOS, especially if they aren't using anything outside the Qt toolkit - many apps will simply compile and run with zero changes. Qt includes stuff like XML parsing, sockets, OpenGL etc. so you can probably will never need anything except Qt.
The only thing some people don't like about Qt is the need for a preprocessor and duplication of C++ stdlib stuff (like containers, I/O operatuins), which supposedly fragments C++ development. But I actually like this - no need to use use multiple library dependencies, everything is included in Qt SDK, along with great documentation and works out of the box.
There are concerns about Qt's future, since it no longer fits in Nokia's strategy. But it's quality definetly exceeds most toolkits like wxwidgets, gtk, mono and many others. Probably because the "boring" stuff like documenting and testing was done by full-time employees.

Oh, and Windows native C++ development is horrible. Just look at how you're supposed to display the standard "open files" dialog and get the selected file names.

Re:Native GUI app development is a pain

[Anonymous Coward]

Yes, with Linux it is that easy.

Gambas is basically (pardon the pun) a clone of VB for Linux. It's amazingly easy to install and use and has great documentation. Anyone who can use VB to make Windows apps would have no trouble making Linux apps with Gambas. It's actively maintained.

Qt Creator is at least as easy to use as MS Visual Studio. Making C++/Qt apps is point-n-click simple. It's widely supported, has great documentation and is actively maintained.

Eclipse for Java development, also very popular, widely used (on both Linux and Windows), lots of documentation, etc.

So unless you consider C++, Java and BASIC not "mainstream languages" I'm afraid your argument doesn't hold water.

Re:Native GUI app development is a pain

[Anonymous Coward]

And amazingly none of those are *native*

Re:Native GUI app development is a pain

[Anonymous Coward]

Oh, and Windows native C++ development is horrible. Just look at how you're supposed to display the standard "open files" dialog and get the selected file names.

And have you tried doing that in native X? At least doing it in native Windows is feasible, for those who have never heard of frameworks or libraries.

Re:Native GUI app development is a pain

[Robert Zenz]

With MS Visual studio / VB any old monkey can make GUI apps easily...

And I'm happy that the "I know how to code in MS Access", "Who needs Object-Orientation in VB.NET? I can code like in VB6!" and "What is a parameterized query?" stays away from my platform of choice. Microsoft might have the bigger ecosystem of developers, but they're also crappier for the most part.

*Easy to use *Widely supported *Actively maintained *Designed for use with a somewhat mainstream language

Eclipse, NetBeans, QtCreator, MonoDevelop and CodeBlocks come to mind. I'd consider those all 4 (well, except MonoDevelop maybe) of your points. I mean, if you consider Java and C++ mainstream...it was at least for the last 15 to 25 years.

Don't get me wrong, I know what you mean...but it isn't like that anymore. I'm using Visual Studio everyday...and there's not one day I don't want to stuff it down someones throat or light it on fire...

Re:Native GUI app development is a pain

[Anonymous Coward]

Yes because compiling binaries is the bulk of the time/work spent developing software. You are doing it wrong, or certainly not for any meaningful living.

Re:Native GUI app development is a pain

[0123456]

And amazingly none of those are *native*

If Eclipse isn't native, how come it's sucking up two gigabytes of my RAM?

Re:Native GUI app development is a pain

[liquidweaver]

I have to say Qt and Python with various GUI technologies not only prove idea wrong, but I would say you can develop faster with them.

That said, I wish there was a Linux equivalent to visual studio - although, Qt Designer is pretty damn close.

Re:Native GUI app development is a pain

[Anonymous Coward]

And I'm happy that the "I know how to code in MS Access", "Who needs Object-Orientation in VB.NET? I can code like in VB6!" and "What is a parameterized query?" stays away from my platform of choice. Microsoft might have the bigger ecosystem of developers, but they're also crappier for the most part.

No, that crowd is on that platform, too. They just use different tools, like PHP and MySQL*.

* Note, I am not claiming that the tools themselves are responsible. Good code and bad code can be written on any platform using any tool set, but some tools have lower barriers to entry and as a result they attract those with less experience and knowledge of the fundamentals of good design.

Re:Native GUI app development is a pain

[Anonymous Coward]

Part of that is because it isn't native. It's Java with an SWT GUI, neither of which is likely native to your system. Java virtual machine overhead isn't the issue it was in 2001, but it's still very real.

It's a weird day on slashdot when supposed geeks consider Java to be "native"

Re:Native GUI app development is a pain

[F.Ultra]

Visual Studio? The only version that Microsoft made that was any good was 6.0, all the others are crap. Yeah some noobs might make GUI applications a lot faster using VS but who wants to run the applications that these produce?

Re:Native GUI app development is a pain

[characterZer0]

http://www.google.com/search?q=compile%20java%20to%20native

Re:Native GUI app development is a pain

[Anonymous Coward]

Incredible, javacc exists. Good luck doing that for Eclipse. Why do you think Android apps still run with a JIT? Do you not think Google would have made them precompile (or periodically compile when plugged in and idle?) when you installed them if it were really that trivial.

Re:Native GUI app development is a pain

[Desler]

If build time is that big of your total time spent developing then you are most likely creating only trivial toy apps. Secondly I can configure and build the entire Qt codebase on a 3.5 year old laptop on Windows in less than 30 mins and I imagine it it's far larger and more complex than the shit you're churning out. In conclusion, you're doing it wrong.

Re:Native GUI app development is a pain

[Zo0ok]

Oh, and Windows native C++ development is horrible.

I have a stupid question... What is actually C++ about Windows Native Libraries. All I ever saw was just completely twisted C - twisted as in just weird datatypes, weird includes, and a Macro hell. I admit, you can declare variables anywhere (just not on top, as in C), but otherwise I cant understand how Microsoft can call it C++.

Real C++ is so different from Microsoft C++.

And QT is very nice. It combines the best of C, C++ and Java. QT Creator is the only IDE I ever appreciated. I really like the .pro file that contains everything about the project, and is still a very small simple text file; nothing like a configure-script or even worse a Microsoft Solution file. qmake uses the pro-file to make a Makefile.

Re:Native GUI app development is a pain

[Anonymous Coward]

Since most Slashdot posters seemed to have washed out of IT like 10 years ago, you probably haven't heard that almost nobody builds VB thick-client Windows apps* anymore. 90% of the Microsoft IT developer community has been doing webapps for the last decade.

* and if they did it would probably be C# and 'Winforms', not VB.

Re:Native GUI app development is a pain

[JBMcB]

It's not just the IDE, but the third party controls available. There are some *really* outstanding controls available for .NET/VCC/WPF. They are usually the first thing I miss when playing around with GTK/Qt development. Grids especially. The DevExpress grid is amazing - built-in filtering, sorting, grouping, basic report generation, paging, editor embedding, multiple binding modes, hybrid data types, runtime layout retention - that's weeks worth of coding for a couple thou (which gives you the rest of the suite)

Re:Native GUI app development is a pain

[zlogic]

Technically I was probably wrong - although some stuff in WinAPI looks like C++, such as
- pre-defined pointers (LPCWSTR etc.) are close to C++ references
- structs are used far more often than in C
- less memory tricks, such as pointer arithmetic
And MFC is C++, but most developers give up on using this monster, and instead implement their own C++ wrappers for making WinAPI usable.

Re:Native GUI app development is a pain

[Anonymous Coward]

> and being XML based can work with any language you want

The only languages I like are programming languages, and data languages that are not XML.

Re:Native GUI app development is a pain

[Anonymous Coward]

QtCreator has this. However it is a right-click, then click on preferred signal.

Re:Native GUI app development is a pain

[Anonymous Coward]

You obviously have never had to deal with visual studio project files when you have more complicated dependencies and graphics libraries. QT is the saner of the options for visual studio and even that is somewhat painful, as it relies on "Custom Compiler Commands" on a per file basis for your .ui's and your headers for moc generation. Nothing compared to the real headache of the win32 API, though.

Makefiles are so much easier to deal with. I like being able to follow a dependency tree without having to resort to a giant mess of XML.

Re:Native GUI app development is a pain

[Anonymous Coward]

well, yeah, I know very well it works with C++, but God knows what kids are doing these days...

Re:Native GUI app development is a pain

[DuckDodgers]

You think GUI development with Java is not a pain? What world do you live in? ;)

I think Swing is a sadistic joke foisted on the industry by some psychotic Sun employees.

Re:Native GUI app development is a pain

[simcop2387]

I remember that kdevelop was integrating QT-Creator into it for some of that but I'm unsure of how far that's gotten. Anyone have any idea?

Re:Native GUI app development is a pain

[AdamJS]

KDevelop is a really, really, REALLY nice IDE, IMO.
Or did you mean the .NET frameworks with the full integration into the IDE/compiler?

Re:Native GUI app development is a pain

[shutdown -p now]

Linux has a development tool that is explicitly designed in the same vein as Visual Basic, and which even uses BASIC as a language Gambas. It's been there for a long time, too.

For C++, as others have noted, there's Qt Creator. For visual design of C++ components, it does much better than VS/C++/MFC.

For Java, NetBeans offers a very nice visual designer for Swing, and otherwise is pretty close to VS feature-wise, at least for things that you need to write a desktop apps. Then, of course, there's Mono & MonoDevelop.

I don't think that lack of good tools is the problem. Reluctance to use tools that are there may be, though.

Re:Native GUI app development is a pain

[Anonymous Coward]

No, everything after Visual Studio 4.2 was trash. "Internal compiler error" my ass! I can't tell you how much valid C++ code I had to rewrite because the fool compiler couldn't handle it.

Re:Native GUI app development is a pain

[liquidweaver]

I'll have to give kdevelop another chance. Last time I used it, kdevelop4 was brand new, and it was buggy as hell. I'll have to check it out.

Re:Native GUI app development is a pain

[Anonymous Coward]

I love Qt. Native-looking on every platform, straightforward API with a completely reasonable architecture, and it handles pretty much all the hairy cross-platform nonsense for you. wxWidgets works, but it's not pretty; the API was rife with inconsistencies last time I tried to use it, and the documentation is opaque. Qt Creator is also a huge boon for UI development, and a solid IDE even for plain C++ dev.

Re:Native GUI app development is a pain

[Anonymous Coward]

Qt Creator works well with C++ and Python, I'd call those both pretty mainstream.

Qt requires a special build system to convert their non-standard C++ syntax into standard C++ for compilation, so Qt-C++ is not mainstream-C++.

Re:Native GUI app development is a pain

[Anonymous Coward]

wxglade works well with me, and being XML based can work with any language you want (and uses native widgets on all the OS'es, so it always looks well integrated). I use it with Python primarily, but it should work for anything (look on their site for officially supported languages).

Have a look here if you're interested: http://wxglade.sourceforge.net/

wxWidgets c++ api is one of the most braindead API's ever invented for c++ gui's. I can't imagine the python API being that much better.

Re:Native GUI app development is a pain

[Anonymous Coward]

Yes, and no. Most people could confuse the meta-language tools as simple macro trickery (which it isn't always, but is always very similar to). "Qt-C++" (as you name it) is a strict superset of C++, and with the aid of the SDK you never have to write a line of code that isn't syntactically valid C++.

Qt is rather mainstream in the C++ world anyway, so "Qt C++" could be considered mainstream as-is.

Re:Native GUI app development is a pain

[AdamJS]

It's still buggy, but most of what I've seen is just the occasional random crash. The program seems smart enough to remember all documents/folders/project states though.

I'm a sucker for its "successor to Kate" spacing/formatting and autocomplete suggestion boxes.

Re:Native GUI app development is a pain

[msclrhd]

You're not keeping up with the Microsoft tredmill. WinForms was deprecated for WPF which has now been replaced with WinRT.

Re:Native GUI app development is a pain

[F.Ultra]

Could be, 6 was the first one I tried so :-)

Re:Native GUI app development is a pain

[BenoitRen]

Oh, and Windows native C++ development is horrible. Just look at how you're supposed to display the standard "open files" dialog and get the selected file names.

I have in the past. What's horrible about it? Seemed pretty easy to me.

Re:Native GUI app development is a pain

[TheDarkMaster]

Too much buzzwords on your comment. A good code can be made with or without object orientation, parametrized querys and etc, etc, etc. If you think code MUST be "object-oriented" (or needs to use any another shiny new buzzword) to be good, you are very, very wrong... Hell, you can create very good code even using the "evil" GOTO, if you know what you are doing.

Re:Native GUI app development is a pain

[zlogic]

Just a quick example: to create a thread in Boost, I need just one line of code, with relevant and understandable. A mutex takes one line (no parameters required) and a mutex locker is also a one-liner. Now in Windows, this stuff requires a lot more boilerplate code, with pointers and illogical function names such as WaitForSingleObject to lock a mutex.
I've created a simple image-viewing UI in win32 api and it took about 600 lines of code, 400 of which are related to scrollbars. In Qt the same application took about 50 lines of code and I didn't even have to write a single line of code for the scrollbars to work.
Another thing which kills me is poorly documented function parameters, some of which are even conflicting! It took me about 2 hours just to make a progress bar work.
If you learned Win32 API properly and have used it a lot, you probably are OK with it. But compared to modern toolkits/frameworks (Qt, Boost, .NET Framework, Java etc) it's much more difficult to use and writing an app in WinAPI feels like a waste of time.

Re:Native GUI app development is a pain

[cyber-vandal]

All the people that use ASP.NET websites like, for example, this.

Re:Native GUI app development is a pain

[Anonymous Coward]

So says the guy who cut his teeth with Visual Basic. Isn't it a truism that we always prefer the tools we grew up with? I grew up with IDEs but decided that IDEs are limiting my skills growth so I decided to learn how to do it the old fashioned way: with text editors, debuggers, make, command line compilers, etc. Now I can develop for Windows, Linux and Mac. I know what an SDK is, unlike this young person I talked to who does not even know what a compiler is, much less what APIs and SDKs are. And that person develops with Visual Studio! He wants to try out Android programming and he sat next to me at the Android Conference.

What is my point? I guess I have none. But to diss Linux development because all you know and understand is Visual Studio is too naive and narrow-minded. You are a developer! You should be better than the users! Maybe you should try to break from the island you are living in and explore the wider world around you.

Re:Native GUI app development is a pain

[Anrego]

You don't have to be a brain dead VB kiddy to recognize that there are times when you just want to throw together a front end in an hour or so, and Linux really lacks the tools to do that (although someone pointed me towards Qt's newer stuff.. which actually looks promising).

For the record, I "cut my teeth" on a TRS-80.. eventually progressed to c (using turbo c) .. stayed there for a long time, did some windows programming in c (using win32 api and the lcc-win32 IDE) .. then moved into Linux where I became a fairly hard core vi elitist type and mainly used c++/Qt .. then went through a weird eclipse/java phase.. and am now in a kind of getting back to c++ phase.

Re:Native GUI app development is a pain

[Anonymous Coward]

Just choose your favourite IDE and don't worry about what goes on under the hood. I prefer Lazarus, but that's just because I come from a Delphi background. There are no doubt others.

Re:Native GUI app development is a pain

[Anonymous Coward]

The only languages I like are programming languages, and data languages that are not XML.

We're not interested in your personal problems.

Re:Native GUI app development is a pain

[Anonymous Coward]

I'd say wxWidgets fills all of those, if you like C++. The forums are a good place for help, and the samples and documentation are pretty amazing too. It isn't rocket science.

Re:Native GUI app development is a pain

[caseih]

Technically QT's moc is not a preprocessor. The .cpp files still get compiled directly. moc generates some signal glue code from the cpp files. Moc is required because the way that qt implements signals is a bit hackish. Qt was developed when the C++ STL and templates were not as standardized across compilers, so signals and slots were implemented without support from those modern C++ features. Unlike in GTK+ and GTKmm (or pyGTK for that matter), signals are referenced by string lookup, which is part of what moc does. Your cpp file, though, will refer to signals as if they were real object references, even though they are not. This caused me a really strange bug once that took some effort to track down.

In GTKmm, signals are actually object references, and connecting them involves calling templated methods of the signal object itself.

Qt's biggest weakness is that its object model is based on C++. Hence when you wrap Qt in a language like Python, the result is rather not pythonic. PyQt is pretty much like writing C++ code with a Python syntax. PyGtk, is a bit more pythonic if that is important. Qt's biggest strength is, as was said, providing a full cross-platform toolkit that provides virtually everything you need to develop a modern GUI app. Its use of CSS to help style widgets is pretty slick too.

Re:Native GUI app development is a pain

[zlogic]

Thanks for sharing the info! Qt signals also have problems with passing non-Qt classes. If I recall correctly, in order for passing non-Qt classes, they have to be inherited from QObject, and have the QOBJECT macro defined in the class definition; also some code is required to declare the class at runtime (qRegisterMetaType). It's strange that Trolltech didn't use C++ templates instead of moc when they developed Qt4 and broke compatibility with earlier Qt versions.
But moc is still required, so Qt compilation may be tricky if you're using a non-standard toolchain (Intel's C++ compiler for example) and not using qmake to build the project.

Re:Native GUI app development is a pain

[F.Ultra]

So you say that everyone who uses websites wants to run GUI applications done be noobs using VS, does that even make sense :) ?

Re:Native GUI app development is a pain

[BenoitRen]

Okay, but you dodged my actual question. That doesn't explain your first example: opening a file open dialog and getting the selected files.

By the way, Borland's WINAPI documentation is better than Microsoft's.

Re:Native GUI app development is a pain

[cyber-vandal]

More sense than your post did. There are a lot of really good applications created using Visual Studio. Stack Overflow is one of them.

Re:Native GUI app development is a pain

[zlogic]

Okay, found the example in my code archives:

CFileDialog fDlg( TRUE, NULL, NULL,
        OFN_PATHMUSTEXIST | OFN_FILEMUSTEXIST | OFN_ALLOWMULTISELECT | OFN_EXPLORER,
        NULL, this );
DWORD MAXFILE = 2562;
fDlg.m_ofn.nMaxFile = MAXFILE;
fDlg.m_ofn.lpstrFile = new wchar_t[MAXFILE];
fDlg.m_ofn.lpstrFile[0] = NULL;
fDlg.m_ofn.lpstrFilter = L"Images\0*.bmp;*.jpg;*.jpeg;*.png;*.tif;*.tiff\0"
        L"All files\0*.*\0";

#ifdef NDEBUG
fDlg.m_ofn.Flags&= ~OFN_ENABLEHOOK; //Vista style support
#endif // Display the Open dialog box.
if (fDlg.DoModal()==IDOK){
        for (POSITION pos = fDlg.GetStartPosition();pos != NULL;)
        {
                CString fname= fDlg.GetNextPathName(pos);//here's the filename!
        }
}

More or less the same stuff in Qt:

QFileDialog dialog(this);
dialog.setFileMode(QFileDialog::ExistingFiles);
dialog.setNameFilter(trUtf8("Splits (*.000 *.001)"));
QStringList fileNames;
if (dialog.exec())
        fileNames = dialog.selectedFiles();//List of files is returned

The WinAPI version required a lot of trial and error to get the input parameters correct. Plus, there's no way of stating "allocate as much memory as required for output", so you have to pre-allocate memory and hope the user doesn't select too many files. This issue turns a "lots of boilerplate code pasted from MSDN" inconvenience into a real problem with no solution except for displaying an error message.

Re:Native GUI app development is a pain

[F.Ultra]

So you just selectebly ignored that I was only talking about noobs creating GUI apps quickly and the applications that those produced. In no way did I ever talk about the quality of applications that can be produced with VS. I'm only commenting on the argument that VS is great for creating GUI quickly. Skilled developers can create GUI apps quickly regardless of platform.

Re:Native GUI app development is a pain

[Anonymous Coward]

I've created a simple image-viewing UI in win32 api and it took about 600 lines of code, 400 of which are related to scrollbars. In Qt the same application took about 50 lines of code and I didn't even have to write a single line of code for the scrollbars to work.

Apples and oranges. You're comparing development with the raw Win32 API to development with the Qt framework. A better comparison would be Qt to MFC (Qt would probably still come out ahead; I detest MFC) with MSVC++, or Qt to VCL with Embarcadero C++Builder (used to be from Borland/CodeGear). These frameworks (and others) exist because it doesn't take much experience programming against the raw Win32 API before you realize that it involves too much boilerplate.

FWIW, the effort you describe for a simple image viewer in Qt would be about the same using C++Builder.

- T

Re:Native GUI app development is a pain

[BenoitRen]

I got the input parameters correct on the first try because I used a WINAPI tutorial and Borland's WINAPI documentation. So I guess the lesson is that MSDN's documentation sucks?

I hate this aspect of the 21st century

[masternerdguy]

"App" is short for application, of which Linux (and any OS worth its salt) has plenty of.

Re:I hate this aspect of the 21st century

[tepples]

Linux (and any OS worth its salt) has plenty of [applications].

GNU/Linux has plenty of free applications but not a lot of well-known non-free applications. There are some kinds of applications for which nobody has figured out how to make a free software model work. Let me know when these applications get ported.

Re:I hate this aspect of the 21st century

[masternerdguy]

Linux (and any OS worth its salt) has plenty of [applications].

GNU/Linux has plenty of free applications but not a lot of well-known non-free applications. There are some kinds of applications for which nobody has figured out how to make a free software model work. Let me know when these applications get ported.

Oh not one of these again. Let me fix that for you.

Netflix Watch Instantly - Admittedly limited (VM or Duelboot, Use Hulu)

Adobe Photoshop, including those high-end features that distinguish it from GIMP mods such as GIMPshop - wine

Adobe Flash CS3 - wine

TurboTax - wine

Stone Edge Order Manager - wine

Sonic 3 & Knuckles - wine

Diablo II - wine

Starcraft - wine

Street Fighter IV - wine

Call of Duty: Modern Warfare 2 - wine

According to WineHQ these applications don't present a problem. Here's your precious photoshop

http://appdb.winehq.org/appview.php?appId=17

Gold/Silver consistently!

Was that hard?

Re:I hate this aspect of the 21st century

[tepples]

So in other words, it appears you're telling me it doesn't matter one bit whether or not developers make and publish Linux-native applications when they can just make a Windows-native application and foist all the compatibility work on the Wine team.

Re:I hate this aspect of the 21st century

[0123456]

GNU/Linux has plenty of free applications but not a lot of well-known non-free applications.

Probably because the commodity applications (word processing, basic image editing, etc) are already 'good enough' and there's not enough of a market for more specialised apps. I almost exclusively use the native Linux apps, but I have Wine installed to run Windows apps for screenplay and novel writing, for example; though the novel writing app is supposed to be coming out native for Linux sometime soon.

Actually, there's even a Linux version of my video compositing app but I've never sat down and figured out what I'd need to do to get the DRM working. If Avid supported Linux too I could probably get rid of Windows.

Re:I hate this aspect of the 21st century

[masternerdguy]

No, I'm telling you that "like omg my favorite app don't work on de linuxes" isn't an excuse to say that Linux isn't useful.

Re:I hate this aspect of the 21st century

[hazah]

So the responsibility of an app being cross platform is with the OS?

Let me ask you this: as each of these apps were developed with the specific design for a single platform, behind closed doors, with the intended purpose of keeping a tight control over that code, please explain how these applications can then be ported in the manner in which you specify (perfect function).

I'm very curious to know your technical solution to this political problem.

Re:I hate this aspect of the 21st century

[tepples]

as each of these apps were developed with the specific design for a single platform

On that list are Photoshop, Diablo II, and StarCraft, which are also ported to Mac. On that list is Sonic 3 and Knuckles, which is a port from the Sega Genesis.

I'm very curious to know your technical solution to this political problem.

As am I to know your political solution.

Re:I hate this aspect of the 21st century

[hazah]

Keep the politics out of it, obviously... but then that's a social problem, and no, I wont be offering solutions to that.

SaaS killing any native app development

[MrBoring]

This triggers my rant reflex...
I started my career in native development, and only in the last say 5 years have I done almost exclusively Java based web development, mostly due to market demands and needing a paycheck. I miss the quick response times, quicker builds and simplicity where it was appropriate. I suspect the best hope for any native development now is maintaining legacy systems and mobile apps. People used to be in client/server development, but that's largely been replaced by the SaaS model due to comparative simplicity, but now we have a myriad of new technologies and frameworks globbed together. The industry's answer to any amount of complexity is yet another platform or framework and more indirection. It's hard to secure and know that it's done properly, and harder to know that someone else did it properly.

Go ahead and shoot me, but I miss the real native development days, regardless of the platform.

Re:SaaS killing any native app development

[Chrisq]

Go ahead and shoot me, but I miss the real native development days, regardless of the platform.

I guess you never used DCOM then.

Re:SaaS killing any native app development

[omglolbah]

Ugh..... DCOM...... kill me

Re:SaaS killing any native app development

[idontgno]

I see that didn't make you puke. Did I mention... CORBA?

Re:SaaS killing any native app development

[Anonymous Coward]

Ugh..... DCOM...... kill me

Don't worry, in your next life you'll be doing MFC.

Re:SaaS killing any native app development

[coder111]

Actually I think you can expose services via RMI-IIOP using Jboss easily. I haven't tried invoking any services from Java over IIOP though.

http://www.jboss.org/jbossiiop
http://en.wikipedia.org/wiki/RMI-IIOP

--Coder

Re:SaaS killing any native app development

[Anonymous Coward]

I am pretty new to the web dev side, but the things I always hear people complain about for complex native apps trying to move to the cloud is:

Response times
Inability to work with very large datasets locally
No multithreading in the browser

Re:SaaS killing any native app development

[Alex Belits]

I started my career in native development, and only in the last say 5 years have I done almost exclusively Java based web development, mostly due to market demands and needing a paycheck.

Good! You are clearly incompetent, and now you can't shit up anything that matters for other people because no one will reuse your code.

Lock in? Take out.

[tepples]

Google can, and does, shut down their services at will. It's DRM^2^2.

How, when Google offers Takeout to copy your data out?

Re:Lock in? Take out.

[SharkLaser]

Yes, like you can "take out" your data from Office too. Still people are crying about Microsoft using their own file formats.

Re:Lock in? Take out.

[Anonymous Coward]

No... you cannot. The local copy of an Outlook-Exchange email database is Microsoft proprietary-encrypted and cannot be exported out of Outlook.

Re:Lock in? Take out.

[realityimpaired]

And yet, oddly, it can be imported into Thunderbird when you install it on a system that has Outlook installed....

Re:Lock in? Take out.

[somersault]

You can export PST files from Outlook. There are import tools to import these into other email clients. If you wanted to be a complete charlatan, you could just drag and drop the emails into an explorer window and copy your mail that way.

Re:Lock in? Take out.

[somersault]

PSTs are user created files. You might be thinking of the OST file, which still holds mail locally if you're using a cached Exchange profile. If you're not using cached Exchange then not much will be stored permanently on the client side though.

How the hell do you think workplaces will stop employees backing up confidential email if they want to? All it would take is a BCC to another mail account and boom, backup. If they can send mail, they can make backups. If you're doing proper backups server side, why would they care about client side backups though?

Re:Lock in? Take out.

[LS]

There's a difference. Code vs data. The app = code. The document = data. With google, you can export your document and keep it, yes, but the app - they can shut it down at any time. With MS office, you have a copy of the app, so Code = Data. They can't shut down the app. They can't take away the code. With google, you never have the code.

LS

Re:Lock in? Take out.

[tepples]

With Google, I admit that you don't control the app, but you can switch to anyone else's app. Or do you claim that Google is the only company to make a certain kind of app?

Re:Lock in? Take out.

[bmo]

>Workplaces do not want employees to backup confidential email.

The only way to stop this is to treat employees better.

Because technology is defeated the instant that someone does a screen-grab by the fabulously new widget in Vista and 7, or brings a camera to work, or runs a printer and scans the printout or whatever.

It is an arms race that has already been lost.

You want your employees to be loyal? Stop being a douche.

--
BMO

Re:Lock in? Take out.

[somersault]

I'm not sure where you're going with the second paragraph. Of course you can't access your mail if your account has been disabled. I was talking about current employees anyway. But in the case of fired employees, IT staff are sometimes lax, or simply not notified of people being fired, so people may still be able to access their mail for a while.

Re:Lock in? Take out.

[MikeyC01]

In Outlook 2007 and earlier (I don't have 2010 here at the office and aren't sure where it is in 2010), click on File | Import and Export ~ Export to a File and select your poison (PST being the "best" of the choices) and export away. Can this be GPOd out? *shrug* In fact, unless you clicked on (to paraphrase) "Don't do this and don't ask me again", Outlook may be building an "archive" PST folder buried deep in the bowels of your /%USER_PROFILE%/AppData directory for you of certain folders I can't remember at the moment.

Multi-select, click on "Forward" will get you copies of multiple messages to the mailbox of your choice

I'm not sure what happens if your corporation has implemented a server-side archiving application. It may be a little more difficult to get the mails out then (since only stubs to the actual messages are stored in the Exchange Server mailbox), but even then, if you can see it on the screen, you can copy it ;)

There are many ways to skin the proverbial cat.

Re:Lock in? Take out.

[SomePgmr]

Regardless, Microsoft distributes a tool for Exchange Server that is specifically made to export exchange mailboxs out of an information store and into a PST for portability. No need to go hunting for local mailbox caches or run exports from the client unless that happens to be more appropriate in your use case.

Though as you said, not sure how that works if you have an archiving solution that uses stubs for old email. Most likely you'd have to export that data out separately.

Re:Lock in? Take out.

[SomePgmr]

Some people are assholes anyway. But your point stands. If you can't trust an employee not to intentionally hurt you, they shouldn't be working there at all.

The primary goal of systems I build is to keep people from making dangerous mistakes, making a mess because they don't know better or looking through things they shouldn't just because the opportunity is there. No, I'm not building banking systems. ;)

Ultimately, you'll never get to a point where you can employ someone that's totally untrustworthy and be able to sleep at night. Some things just aren't IT problems... they're administrative ones.

Re:Lock in? Take out.

[MikeyC01]

ExMerge assumes you have physical access to a command line on the Exchange Server logged in with appropriate privileges. BTW ExMerge with Exchange 2003 was awesome to use when the mailbox is greater than 2GB ;)

At least you can filter by date

Re:Lock in? Take out.

[SomePgmr]

I hear that... I still have nightmares of dealing with > 2GB mailboxes, even just migrating boxes from IS to IS. :(

Re:Lock in? Take out.

[LS]

No, I was just explaining that the parent poster misunderstood the grandparent's point, which is that the apps can shut down at any time. I actually am a big google fan, and utilize their services to the fullest. I'm just pointing out a fact.

The challenge...

[Junta]

I always go with local applications rather than webapps when possible. The issue is that with the google products I used, the critical aspect was how to get data to and from other people. Google reader takes away the only feature that made me use that instead of something else, but something else simply cannot realistically replace the 'sharing' capability without relying on some service that can be shut down at the whim of the provider.

Incidentally, why the hell did everyone start going with 'SaaS' instead of 'webapps'? The concept is nearly as old as the web itself.

Re:The challenge...

[tepples]

Get your own domain, and host your e-mail on it. Then you can share whatever you want. You can even use Google Apps for this; if Google decides to shut down Google Apps, you can switch your e-mail hosting to someone else.

Re:The challenge...

[mwvdlee]

Incidentally, why the hell did everyone start going with 'SaaS' instead of 'webapps'? The concept is nearly as old as the web itself.

Because SaaS is on the cloud, whereas webapps are on the internet.

Re:The challenge...

[Anonymous Coward]

No it's not, it's simply what it says, Software-As-a-Service. Leased. Nothing to do with the cloud specifically, it's just the latest buzzword for a the that centralized controlled thin-cliented model.

Re:The challenge...

[icebraining]

That's nice for email, and I do just that (couple with offlineimap to backup it all).

It's useless for Google Reader, which is what GP mentioned.

Re:The challenge...

[Noughmad]

I think (or at least hope) that GP was joking, he's trying to say that SaaS is the same as webapps, only buzzword compliant.

Re:The challenge...

[aaalex52]

Check out Tiny Tiny RSS. It's an RSS server program you run on your own machine. Like Google Reader but under your control.

Re:The challenge...

[Junta]

I was thinking less about email and more about things like 'google reader', where I can share all day long on my own domain but my content will not be interleaved with others. email/jabber are nicely distributed, but some things do have an inherent network effect that is difficult to overcome.

Bad design sense doesn't bring in new business.

[Anonymous Coward]

The lack of a really good desktop environment discourages myself and most people at my firm from even using linux for our personal computers.

Bad design sense doesn't bring in new business.

Re:Bad design sense doesn't bring in new business.

[Darfeld]

Well for that matter, I don't know really good desktop environnement outside Linux either. I mean, I'll choose Gnome or KDE over any windows env any day. Not sur about OSx though, cause I don't know it enough, but It doesn't seems ultimate to me. Sweet all right, but still...

Now the desktop env shouldn't really matter when writing app. In the end it's more or less a matter of windows decorator...

Re:Bad design sense doesn't bring in new business.

[Anonymous Coward]

You obviously haven't tried any.....

Re:Bad design sense doesn't bring in new business.

[Anonymous Coward]

</sarcasm>

I use KDE4 every day, sadly.

Re:Bad design sense doesn't bring in new business.

[Lisias]

Well, I was a OSX 10.3 heavy user for almost a year, and I can say that it was, at that time, the most beautiful and coherent UI I ever used.

But it was not perfect. There were a few common tasks where I missed Windows XP (some file manager operations were easier on XP, I swear!).

For the sake of comparison, I still prefer Windows XP UI over the Win7 one. Win7 is prettier, no doubt, but I can reach what I want easier with WinXP (not to mention that some over-simplification on Win7 UI renders some customizations almost impossible - as each folder having its own visualization settings).

For the past 18 months, I'm a heavy Gnu/Linux/Gnome user. Gnome is not aesthetically well done as OSX, but it works fine to me and after a very few customizations (as adding a shell box on each filemanager window) is now the most effective UI to me (but I still miss some WinXP features).

However, I'm not a ordinary user. I'm a software developer - emphasis on mobile (Android and J2ME) and web applications.

Re:Bad design sense doesn't bring in new business.

[Anonymous Coward]

You poor thing.

Re:Bad design sense doesn't bring in new business.

[lvxferre]

I'll just say that you can tweak a GNOME 2 enough to look like Aqua [OS X interface], but the inverse is not true.

Re:Bad design sense doesn't bring in new business.

[hazah]

In my experience the OSX GUI model is still too clickity click oriented. Main pet peeve is the insistence on having cmd-c cmd-v style of copy paste, where as in linux is simply highlight and middle click.

Re:Bad design sense doesn't bring in new business.

[Anonymous Coward]

The lack of a really good desktop environment discourages myself and most people at my firm from even using linux for our personal computers.

I spent decades in front of beehives, VT52s, VT100s and 3270 terminals so I have no problem finding a really good desktop environment. It's any multitasking, multiscreened environment where I can cut and paste across processes. I don't care what OS, what shell, whatever... just let me type! The mouse is so slow and tedious; I can type a full line of bug-free code in the same amount of time it takes a twenty-something to "perform a mouse gesture".

But I swear, you tell this to the kids today, and they won't believe you.

And since we're on that subject, when I was a lad, I had to get up in the morning at 10 o'clock at night, half an hour before I went to bed, eat a lump of cold poison, work 29 hours a day down at the mill, and pay the mill owner for permission to come to work, and when we got home, our Dad would kill us every one, and dance about on our graves singing "Hallelujah, hallelujah."

Re:Bad design sense doesn't bring in new business.

[cyber-vandal]

Which is shit if you want to replace some text, e.g. a url in a browser.

Re:Bad design sense doesn't bring in new business.

[DarwinSurvivor]

So instead of switching to a fairly good DE (linux), you decide to stick with the most useless, feature-free DE in the computing business (windows)? The *only* innovative thing Microsoft has done on the desktop that I can even remember was the "maximize to half-screen by dragging" thing and now almost every major DE out there (unity, gnome, kde) has it. In fact, gnome-2 had it working before windows 7 was even RELEASED!.

Windows hasn't had the right to claim "superior desktop environment" for over a decade.

Re:Bad design sense doesn't bring in new business.

[Darfeld]

ctrl-c ctrl-v still works in this case. And I really appreciate the highlight middle-clic thing. To copy paste things in the command line, for example, its just great.

So why would i need native apps again?

[Anonymous Coward]

Everything i need to do today should be technically possible via a browser. I'm sure there's still some use cases for native apps (multimedia, 3D), but hey are getting thinner every day.

Re:So why would i need native apps again?

[Jeng]

Everything I need to do today should be technically possible via a browser, but also allow me to use the application when my internet connection is down.

Not SaaS it's FAIB

[sproketboy]

It's not SaaS Killing Native Linux App Development it's the FAIB (free as in beer) crowd doing it. Basically there is no money to be made in desktop Linux development out there since people simply are not paying for the software.

Most developers would use Java or QT and do all their work in Windows and then do a quick port and QA to Linux - if you're lucky.

Re:Not SaaS it's FAIB

[jerryjnormandin]

Wrong dude. There is plenty of money to be made Native Linux App development. It's called support! I think this post is just FUD. Have you ever checked out Freshmeat.net or Sourceforge ? There are lots of new applications being developed for Linux.

Re:Not SaaS it's FAIB

[Anonymous Coward]

You're so delusional that you don't even deserve a response. Just LOL at the Freshmeat.net and Sourceforge suggestion.

Re:Not SaaS it's FAIB

[JustSomeProgrammer]

Support for an application? Seriously? If an application is so complex I need to buy support to use it. I'M NOT USING IT*.

Do you buy support for LibreOffice before you start using it or even anytime after?

* -- unless I'm forced to use it for business reasons and don't want to care about it. But I don't think those are the kinds of applications they are talking about here. And even then I'm having trouble thinking of any that I personally use without it being just a webapp or a piece of an application that I'm build (such as a web server).

Re:Not SaaS it's FAIB

[Anonymous Coward]

just search for Humble Bundle multiplatform pay-what-you-like games - who's paying most? Linux users...

Re:Not SaaS it's FAIB

[Anonymous Coward]

You call the stuff Freshmeat or Sourceforge professional software? You can call my post FUD - I call it reality.

Re:Not SaaS it's FAIB

[multipartmixed]

That's a really good point.

I was about to post, "Hey, I have been doing SaaS for 15 years and only recently started releasing end-user stuff that runs on Linux".

But I'm not making any money from it -- I'm releasing it under a FOSS license.

Re:Not SaaS it's FAIB

[ardeez]

I'm curious to know, how many of the people releasing projects on Sourceforge
or Freshmeat area actually making a profit on their apps?

Also, how do you make lots of money just from supporting an app unless you intentionally
make it hard to use, unstable or force people to do expensive migrations between
each release?

You say there's plenty of money to be made, but I just don't see how.

Re:Not SaaS it's FAIB

[sproketboy]

jerryjnormandin won't reply because there is no good answer. Linux users simply will not pay for anything. They feel entitled to get everything they want for free. Software, Music, Videos, TV, etc...

Re:Not SaaS it's FAIB

[sproketboy]

Exactly. Thanks. I was really expecting to me modded troll posting this on slashdot.

Desktop standards

[bonch]

It's not web apps--native apps are actually experiencing a major comeback, especially on mobile devices. Native Linux app development sucks because of a lack of standardization. Conflicting projects, changing APIs, and aggressive attitudes from the community have all contributed to the failure of native Linux development. If a company wants to develop a commercial product for Linux, it will get attacked right out of the gate for being closed source and not be available for free, and the company has to support a staggering number of possible desktop environment configurations and APIs--many of which might get supplanted by something else in a couple of years.

Native Linux development will always suck until it's worth it to actually develop for.

Re:Desktop standards

[Lumpy]

"Native Linux app development sucks because of a lack of standardization. Conflicting projects, changing APIs, and aggressive attitudes from the community have all contributed to the failure of native Linux development."

No not really. Statically compile your binaries and call it done with the stupid dependency hell and API change hell. That way you can release binaries that "just work"(tm) instead of the 6 hours to get this damned bullshit app to run that uses hyper alpha library Development version 6.32.01.2 and anything else will not work because the programmer is a fricking moron.

Honestly, why did linux embrace the "spread crap everywhere" windows software installation model, instead of the Apple software install model? IT's retarded to have to install files shotgun style all over the fricking FS.

Re:Desktop standards

[Anrego]

The big problem I see is that the stuff you describe is part of what makes Linux great for geeks like me.

Standardization tends to run up against flexibility. Linux is a hackers platform.. screw with anything you want.. you have the code! Don't like how something works.. make your own version of it!

This of course, as you said, makes Linux a nightmare for commercial development. For open source software, people can make the little changes necessary to make it work on their particular distro.. with closed source software, you can't do this.

Re:Desktop standards

[0123456]

Honestly, why did linux embrace the "spread crap everywhere" windows software installation model, instead of the Apple software install model?

Because we don't want to waste vast amounts of memory loading the same libraries in different applications and don't want the security nightmare of every application having its own copy of DLLs with ancient security holes that will never be updated?

Re:Desktop standards

[bonch]

Apple software isn't statically compiled. It either installs frameworks to a system folder or ships them in the application bundle.

Static compilation would not solve the problems of their being no standard desktop environment, no standard configurations for commercial developers to support, and anti-closed source attitudes from the userbase.

Re:Desktop standards

[Anonymous Coward]

I have apps that run under Kde, Gnome 2 and 3, XFCE, and WM. if you make calls directly to a Windows Manager's API you already are at Epic Fail for Linux development.

QT libraries work under ALL Window managers in linux.Hell call the core X widgets if you want a retro look instead of the Qt look.

the ONLY you app should talk tot the WM is if there are events it needs to pass to it, otherwise for 98% of the time you dont even need a WM.

Re:Desktop standards

[Anonymous Coward]

Or the most traditional method: source distribution.

Re:Desktop standards

[jbolden]

Honestly, why did linux embrace the "spread crap everywhere" windows software installation model, instead of the Apple software install model? IT's retarded to have to install files shotgun style all over the fricking FS.

You have a low user number, I'd assume you had this argument a dozen years ago or earlier. Linux doesn't use either model, it uses the huge OS library model which is frankly closer to what mainframers use than either Windows or Apple.

Re:Desktop standards

[cyber-vandal]

They could always do what the server application vendors do and only support certain versions. I doubt very much whether Larry Ellison gives a shit about what RMS thinks, so why would a desktop app vendor give a damn.

The reasons behind it

[bananaquackmoo]

Really this is not an argument about "in the cloud" or "webapp vs native app." One has to look at the reasons this is happening. The major reason is because most webapps are universally cross platform and usable anywhere. Other bonuses include that they are generally lightweight, don't require an install, and they sync your data between computers. Looking at that list of reasons, it is entirely possible to make native apps which do this, its just that most developers can't be bothered to do so (or aren't allowed to do so).

Saa

[bWareiWare.co.uk]

The architecture (thin v fat client) is tangential to whether you in-source or outsource system administration.

Most of the software we rely-on most is accessed via the browser:
Roundcube, Gallery, phpMyAdmin, LDAP Account Manager, Trac, and most importantly our own internal systems.

Once LibreOffice makes the switch my work will probably go days between firing up a GUI besides Firefox.

Is SaaS Killing Native Linux App Development?

[Anonymous Coward]

Maybe. Why do you ask?

Does anyone use Linux native apps?

[spike_gran]

Despite being in free software for a bajillion years and using it as my desktop, I can't say that I've used any native Linux apps for anything really. For the past few years, it has just been a way to get a webbrowser running and to get online, and as a place to cache content. I also use Emacs and the dev environment to make my own (web) apps, and Apache to serve them.

The only native apps I use are games that need native audio/video control.

Re:Does anyone use Linux native apps?

[JohnFen]

Native Linux apps are all I use. I don't do use SaaS stuff at all, and don't intend to start. I simply don't trust any third party enough to make myself so vulnerable to and dependent on them.

Or, maybe Linux is dying...

[Anonymous Coward]

Linux is dying (or perhaps dead). At least it is on the desktop. It was stillborn and never really had a chance. Everyone (myself included) spent so much time worrying about Microsoft that we ignored what Apple was doing - and then *wham*, OS X comes out. A Unix-backed desktop OS with a gorgeous UI that non-techies can actually use.

Without a strong desktop presence, there is very little need for native apps. We sysadmins prefer our command line tools - and nobody can argue that there aren't a lot of those ... but general app development on Linux dropped off years ago and I don't see it improving anytime soon.

And you know, I really don't have a problem with that. I started with Linux in the early 90s, with one of the first releases of Slackware. Back then monitors were fixed-frequency and you had to calculate your video card's dot-clock & other timings in order to not blow up your monitor... I became a full-fledged sysadmin in 1995 and worked for a number of big Linux companies. I drank the kool-aid... all my machines at home ran Linux and it was good. In 2000 I switched gears and became a Linux developer - working on both embedded and desktop projects. Had some great experiences back then.

But somewhere around 2002 I started to feel betrayed. Here I was, nearly 10 years later and Linux still wasn't on the desktop - at least not in any kind of meaningful way. Sure I kept hearing how 1997...1998...1999...2000...2001 were all going to be the "Year of Linux on the Desktop" - but it never happened. The various Linux forums were the same old thing - people complaining about Microsoft (and now Apple) - all the while lifting as many UI ideas as they could from each OS, expecting that someone a floaty OS X dock-like thing would attract hundreds of thousands of new desktop users to the platform.

Here we are in 2011 and I'm seeing the same old shit. I'm just about ready to give up Slashdot because 90% of the Linux-related news stories just remind me that we haven't made any progress (and yes, as a developer I tried to help out in that area). Ubuntu's now saying that they're going to make a tablet UI. Yay ... only they haven't cracked the desktop (they made a lot of progress and undid it all with Unity).

*sigh*

Re:Or, maybe Linux is dying...

[masternerdguy]

Oh man, only an anonymous coward would post something like that. Mod this -1 Flamebait.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

no i won't. cower in my shadow you ignorant hypocrite.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

Linux is dying (or perhaps dead)

It's not over until we win, or we all give up ...and I and my fanatic friends will never give up. You underestimate our fanaticism.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

We're all well aware of your fanaticism, we doubt your ability to win.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

I am not really sure if I want more `GUI-user-oriented' Linux though. I have noticed with Ubuntu, the main push is to make the system `easier to user' meaning easier to do the simple things, which, strangely, leads to much harder times for users who prefer thing like CLI, etc. `Make something that even a fool could use and only fools will use it' someone said and to my astonishment, this is certainly true with almost anything on the desktop. I do not want a fancy crippled DVD player, I want a real computer instead.

Re:Or, maybe Linux is dying...

[serviscope_minor]

Linux is dying (or perhaps dead). At least it is on the desktop.

Linux desktop seems steady at around 1%. Since computer usage is growing, that also means Linux desktop usage is growing. If I ever capture 1% of the global computer market, I would consider it a success beyond anything I could comprehend. So, you have a very odd definition of "dead".

general app development on Linux dropped off years ago

Huh? Like what? There are now decent 3D modelleers (blender), audio editors, photographic maniuplation, drawing, image stitching programs, indie games, etc. There is so much more than there was 10 years ago.

But somewhere around 2002 I started to feel betrayed. Here I was, nearly 10 years later and Linux still wasn't on the desktop - at least not in any kind of meaningful way. Sure I kept hearing how 1997...1998...1999...2000...2001 were all going to be the "Year of Linux on the Desktop" - but it never happened.

So you feel betrayed that the rantings of a bunch of usenetters didn't come to much? Wow. You have a low threshold for betrayal.

all the while lifting as many UI ideas as they could from each OS

Yeah well, that does kinda suck. The blind copying is making a mess of the Linus desktop experience.

Here we are in 2011 and I'm seeing the same old shit.

Like...?

I'm just about ready to give up Slashdot because 90% of the Linux-related news stories just remind me that we haven't made any progress

C' ya

Re:Or, maybe Linux is dying...

[JasterBobaMereel]

Desktop is a strange notion that seems to mean Office PC - which means OS is irrelevant, but the Apps it runs Office, DB, Company app, Webbrowser Outlook etc is and only Windows can run these... so we are stuck with Windows

OSX is pretty and works well, but is still the niche market is always was, and it is the apps it runs, image editing , video editing, design tools , that really sell it

Linux has it's place as well, but it is not the Desktop because all the Desktop apps people currently use run on Windows or OSX ... There is not native Linux killer Desktop App

Cloud and Web Services may actually help this, if you can buy an Office PC without Windows (i.e. ready for a minimal Linux install and Web browser) and only run cloud apps, this will be much cheaper ... and so be the Desktop Linux ....

 

Re:Or, maybe Linux is dying...

[Anonymous Coward]

Ferrari are dying, Mclaren more so. Just because the masses aren't buying or using something doesn't mean they are at death's door. Stop being so melodramatic, and stick to your dumbed down junk on OS X. Gorgeous UI? Bwhahahaha.

Linux controls the embedded market, it's a major player in mobile technology, it's in most medical devices, HDTVs and media players. It's controlling most blu-ray players, and the reason HDTVs are now web enabled devices supported more media codecs than both Win7 and OSX combined out of the box.

So what if your sad dream of Linux desktops never happened. The world is much bigger that a single view from a dweeb like you.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

Linux may not be on _your_ desktop but it is certainly on mine, and it will remain so far into the future.

Linux is a niche OS and it caters to an elite (in the technical sense). Linux will never become a part of the mainstream. In order for it to do so there would need to be a great surge in technical literacy among the masses, and _that_ is never going to happen.

The fault is not with Linux. The fault is the failure of the common person to properly educate himself in the face of burgeoning computer application.

Re:Or, maybe Linux is dying...

[flydpnkrtn]

You should come on over to Hacker News if you're looking for what Slashdot used to be circa 1998... http://news.ycombinator.com/

I agree with you... the comments were the big draw, and mostly it's "herd mentality" on /. any more.

Re:Or, maybe Linux is dying...

[roman_mir]

Well, I switched to only OpenBSD and GNU/Linux about a year ago, before that I mostly worked on Windows and deployed on Unix.

However I do NOT like either Gnome, I do NOT like KDE, I do NOT like XFCE, I cannot stand Unity.

I don't like ANY of the windowing environments that Unix/GNU Linux offer.

I HATE Mac.

So in reality I only like Windows Windowing environment, specifically Windows 2000 or Windows XP. I think those are very solidly done and they FEEL solid. They are very integrated, all of the cut/paste, all of the menu looks, all of the fonts, borders.... It's no contest.

I HATE the way the non-Windows windowing systems are done, but I use them anyway, because I have to, because the underlying technology is what I need in GNU/Linux or Unix, not in Windows, and again, I don't like anything from Apple.

So unfortunately for me, I have to suffer through this.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

I manage some very large general-interest sites, and have never seen Linux usage approaching 1%. More like 0.1% down in the noise.

Desktop Linux isn't dying, it was stillborn to begin with. As in, never got any traction outside of Computer Science sophomores, never will.

Re:Or, maybe Linux is dying...

[sauge]

Niche products do not have capitalization larger than oil companies. http://www.marketwatch.com/story/apple-crowned-no-1-with-biggest-market-cap-2011-08-10 You are correct about the old apple of the pre-2000's. These days Apple OS X is long past niche.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

1) why in the hell would you want slashbots posting on hn?
2) their heads will explode when they find out perl cgi scripts are no longer cutting edge technology

Re:Or, maybe Linux is dying...

[Anonymous Coward]

The great thing about software ecosystems is that they are much larger than any natural organism, and therefore they take years and years to die, because they really just fall into hibernation.

They may lose momentum, but in order for a project to die, you have to lose a lot of interest from everyone involved.

I don't expect you to use Linux right now. The kernel itself is okay (I guess, because I've never had to touch it), but people are still learning what they want from human-computer interaction and from APIs and many other things, so it will take time before the userspace is up to the expectations of every-day users.

But for me, I already use it every day, so there's at least one person thinking about these issues almost constantly. Even if I'm not making much progress, the idea of "Linux for end users" won't die.

Speaking of projects that are actually dying, I think more people should look at GoboLinux... I'm fucking sick of having to run package managers when the tools can be written to just put every package in its own folder and then use symlinks for backwards compatibility. I've been meaning to install it in a VM, but I don't know much about VMs so I haven't bothered. I'll probably have to get on it this weekend, I think the shitty directory tree is holding back proper package management. Programming languages are shipping with their own package managers, a sure sign that the directory tree and the system-level managers are not doing their job well enough.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

Oh I dunno. Maybe that "iPhone" and "iPad" I keep hearing about have something to do with their profit instead of OSX.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

C' ya

That's a real good attitude to have. /sarc I do find it common in the Linux World. The "we don't care if you use our product or not" argument is old and its the reason Linux is always being treated as 'low man on the totem pole'. Quite frankly it's an unprofessional response and as such, Linux will continue to be treated as unprofessional. Personally, I'd rather see folks in the Linux realm do well. But currently, the attitude and the aversion to anyone wishing to make a living with anything Linux is the problem. I always hear the argument that 'you can write and sell software on Linux' but I've seen the way the Linux or open source community turn on their own when they actually try to make a buck to feed their family.

Re:Or, maybe Linux is dying...

[msclrhd]

Wikipedia [http://en.wikipedia.org/wiki/Usage_share_of_operating_systems] lists GNU/Linux as having a 1.49% share on Wikimedia sites and the median from various sources is 1.11%.

Re:Or, maybe Linux is dying...

[Anonymous Coward]

No, Linux on the desktop isn't dead: the desktop itself is dead. Not entirely, of course, but the momentum is shifting toward other/more platforms. That is where SaaS shines: put all the logic in once place, and skin the app for different devices (including the desktop). The 'skin' could be html, or if you need native gui controls, you can go that route also.

I don't see how any of this presages a lack of developer interest in developing on Linux; it's just that folks are developing differently to reach a broader audience.

Of course, if your app doesn't use the network in any way, that's all mute and native is probably the way to go. SaaS is just a buzzword. I think it's the fact that folks are more and more interested in writing networked apps that is motivating the change in focus.

Re:Or, maybe Linux is dying...

[noahm]

I disagree with most of your comment, except the bit about OS X. Purely from my own experience and observations, I feel like lots of people who would otherwise be driving the state of the art in Linux desktop development have instead switched to Mac OS. These people aren't necessarily the folks who would be doing the development of the desktop apps, though many are. Many of them are simply power users or people with good ideas for how usability could be improved. They would be active contributors on mailing lists and forums, and they would contribute ideas, suggestions, and bug reports. But instead of working to improve the Linux desktop, they've become happy consumers of a Apple's desktop. They're perfectly content, and Apple continues to develop their software, so I suppose there's not a lot to complain about except from idealistic points of view.

In reality, though, I think the Linux desktop has made amazing strides over the years. Yes, I think things could be better, but there are a lot of really dedicated people doing very good work. Rather than lamenting what we could have had, we should celebrate the best of what we've got, as a community. The Linux desktop is certainly not dying.

Re:Or, maybe Linux is dying...

[jbolden]

Well if you have been along that long than how about:

Linux on the server (big battle)
Linux on mainframes (surprise win)
Linux dominating supercomputing (another big win)
Linux on embedded.

Re:Or, maybe Linux is dying...

[Alex Belits]

Most likely it's hairyfeet or some other Microsoft shill in here.

Re:Or, maybe Linux is dying...

[Zontar The Mindless]

I manage some very large general-interest sites...

IOW you're hosting many TB of straight and gay porn?

Broader concept

[bradgoodman]

It may be orthagonal to the point - or on the point - I can't tell - but one thought immediately came to mind.

---

If by "app" you are talking a "user application", meaning one with a whole user interface (i.e. GUI), etc. - I'd say that is eroding. I wouldn't say that "SaaS" is what's eroding it however. I've been using Linux for almost twenty years, and only have written a single GUI-based (Glade) application for it. I see Linux as a "back-end" system, and have always used Windows, or a Mobile platform as the "front end". When I need to write, let's say a "management" interface for something, or something else that requires something prettier - like a GUI, I've always implemented this as a web service.

This is more of an effect of the fact that Linux has never really made it big into the desktop arena. Conversely however, Android is making it mainstream in the tablet and smartphone space (albiet, under Java for user-apps) - so as tablets and phones eclipse desktops in the user-oriented space - Linux, ironically, stands to become the dominant force in "user" systems, they just won't be "desktop" systems as we think about them today.

Run the server locally

[jonsmirl]

HTML5 is just another GUI front-end library. In no way does it require you to write cloud based apps. If you want a native Linux application write the GUI in HTML5 and run the server on the same machine as your GUI. Hmmm.... something kind of like the Xserver model, but brought 30 years into the future?

When people whine about the ending of location transparency with the Xserver, what is going away is the Xserver as the primary GUI library, not location transparency in general. The Xserver needs to die, it is pass its prime and we need to move onto newer GUI technologies.

So stop writing native Linux applications and instead start writing HTML5 applications that ship with a built-in server. The cool thing about apps in this model is that the GUI works on Linux, Mac and Windows plus you can run the server locally or in the cloud - your choice. If you want to help out convert some native Linux apps into the HTML5 model.

Wayland is a key transition technology. It allows apps like Chrome/Firefox to be written directly to EGL. Plus you can run a user space Xserver as a legacy tool.

Re:Run the server locally

[tepples]

If you want a native Linux application write the GUI in HTML5 and run the server on the same machine as your GUI.

A few issues:

• That runs into all sorts of firewall issues when the operating system blocks applications from listening on localhost.
• On which port of localhost does your application listen that isn't already used by the other running applications?
• Pop-up blocking features make notifications more difficult.
• To what extent do all deployed web browsers support drag and drop of objects between the web browser and other windows?

Re:Run the server locally

[Anonymous Coward]

Anonymous post in case my employer reads this. I'm involved with implementing a new version of what is really a desktop application this way. And it sucks. You loose all the advantages of maintained state, plus you spend the usual amount of extra time fighting CSS and all those other irritations, because you haven't got a decent widget set to work with.

Actually, my belief is that the real way to go is "web" applications that are written as proper GUI applications in the browser, and use the server as a data API. But, you use a widget library that sits on top of HTML and CSS and hides it - and the library does decent layout management so you avoid the layout abortion that is CSS. Pity one is stuck with Javascript, maybe Dart is the answer.

Re:Run the server locally

[jonsmirl]

All of these problems can be solved. I believe the rest of the world is going down the browser GUI path, so it is up to the Linux community to decide if they want to follow. Gnome/KDE are almost certainly dead ends.

This is not an overnight switch. It will take a decade to fully transition to an HTML5 plus local server model.

One reason why so many people don't like SAAS is because the server side of the app has been implemented closed source. Providing open source servers is one way to address that problem. For example convert Open Office to a HTML5 plus local server model as a response to Google docs.

Re:Run the server locally

[0123456]

For example convert Open Office to a HTML5 plus local server model as a response to Google docs.

Why the fsck would I want to convert Open Office to HTML5 when it already works fine as C++?

Re:Run the server locally

[tepples]

All of these problems can be solved.

Let me know when they are.

Re:Run the server locally

[Anonymous Coward]

The first 3 are really just setup issues. The distro can configure the firewall properly, put the service on a port that isn't being used, and add the localhost to the 'allow-popups' setting in the browser. The 4th one is a huge problem (drag and drop).

I don't think anyone is going to replace the desktop with a browser though.. and that's the only thing that would be useful for.

I could definitely see a desktop, with applications that all load through a browser... that's definitely possible to implement right now.

But you would have to set a standard for what info apache (or whatever) will provide to all of the applications. And we all know how long standards take. (unless you want each app to install its own server.. which does not sound like a good idea.)

Re:Run the server locally

[JohnFen]

I believe the rest of the world is going down the browser GUI path

Oh, god, I hope not. I've yet to see a browser-based anything that doesn't suck more than a natively-implemented version of the same thing.

Re:Run the server locally

[Anonymous Coward]

Why would you ever want to block applications from listening on localhost?

Re:Run the server locally

[jonsmirl]

Google has ported Quake to run inside Chrome using HTML5. You just aren't seeing fully developed HTML5 apps yet.

The GUI libraries in older versions of HTML were not equivalent to their native counter parts. A lot of that had to do with Microsoft torpedoing the standardization and advancement of HTML. They seem to have stopped doing that now. Let's see how browser centric Win8 is now that the consent decree has expired.

Re:Run the server locally

[Xanny]

I agree with this, nothing stops us from writing html 5 / css / javascript apps that people download and execute as an html script and call that an app. You just have the website locally and render it in your browser. Same principle, and it works everywhere.

Re:Run the server locally

[Anonymous Coward]

Wow you sure told him, tepples.

Re:Run the server locally

[tepples]

The distro

Is the distro Fedora, SUSE, Debian, Ubuntu, Mint, or something else? Must the developer of an application make available a package for every distro with more than 1,000 users?

Re:Run the server locally

[JustSomeProgrammer]

I've seen plenty of old 3d shooters redone in HTML5. There's not a single one I could stand playing for more than 5 minutes. The performance is always horrible. The suggestion that all games should be done in HTML is silly until you can give HTML full native control over my system so the designers can really use the features of my system and at the same time I would never WANT HTML to have that much control over my system.

Just because something has been done doesn't mean it has been done well. Or should be the model for how things are in the future.

Re:Run the server locally

[Belial6]

Because you then only need to maintain one UI for every computer system present and future? Because you could then VPN into your server and the application would function on your tablet while hosted on your server/desktop?

Re:Run the server locally

[Belial6]

You make a very good point. I would take it one step further, and say that developers should consider not just running the server side locally, but build the server side into a VM image. This way not only does the UI become abstracted from the host OS and hardware, so does the back end.

Re:Run the server locally

[Anonymous Coward]

I think the only practical way to do it is to break it into a few parts:

1) a single web server, configured by the distro;
2) the CGI apps that get installed to {wwwroot}/{appname} or wherever
3) a launcher (stripped down web browser) (configured by the distro) that connects to localhost:{port}/{appname} (ex: you click on libreoffice... the launcher opens localhost:{port}/libreoffice)

Since a developer would only deal with #2, there would only need to be one package for all of the distros (or more like an RPM, DEB, etc... but that's how things already are).

Re:Run the server locally

[Flammon]

Absolutely. Chromium is already being ported and I would love to see Mozilla's Azure optimized for Wayland. I'm anxiously waiting for the speed boost.

Re:Run the server locally

[Anonymous Coward]

But why would you want to do that?

HTML5 (i.e. HTML + CSS + JavaScript) is a very crude development environment. An amateurish mark-up language with inconsistent, not-well-thought-of syntax and error correction rules combined with JavaScript, which is... well, JavaScript with it's fundamentally broken object relativity and other bat-shit insane conversion rules and loopholes.

The only reason one would want to use such combination is when one does not have local access to the computer. There are tons of better tools when doing stuff locally, and there's no reason to limit oneself to the compromise of legacy that HTML5 eventually only is.

(For example Trolltech once again overdid themselves with QML, which is nothing short of an amazing meta language.)

Re:Run the server locally

[shutdown -p now]

So stop writing native Linux applications and instead start writing HTML5 applications that ship with a built-in server.

Why would an HTML5/JS app even need any server? Just write a simple browser host - or use an existing one - and load the markup and the code directly from the filesystem.

Re:Run the server locally

[BenoitRen]

HTML5 is just another GUI front-end library.

No, it's not. It's just a mark-up language. There is nothing graphical about it. You should look at something like XUL instead.

Re:Run the server locally

[Anonymous Coward]

Any links to examples?

Re:Run the server locally

[jez9999]

Why use HTML5, and not XULRunner? XUL is designed more for client-side apps rather than web apps, and you still get a nice cross-platform app that you can use Javascript to code and not worry about memory management.

Re:Run the server locally

[seantide]

One problem with both XUL and HTML5 is all of this web stack is horribly fragile and slow.

Yea I'm sure we can eventually through enough horsepower at the problem to ignore much of the problem, but can't we come up with something better?

I will always prefer native applications until someone can make a web stack that doesn't suck.

Re:Quick question.

[Anrego]

Yup, they screwed it up big time!

Us geeks down here are happy with our favorite window managers (I like openbox personally).. but we've lost all the major desktop environments as champions for new users.

hmm

[buddyglass]

Could it be because the set of target environments is so large compared to Windows / OS X? You've got to support multiple distros (and versions of distros), multiple desktop environments, etc.

Re:hmm

[Anonymous Coward]

It's all about barriers to entry.

Gaining the expertise to develop any sort of application is a pain, but the promise of money has encourage many people to make native Windows applications.

The money is now pulling towards web applications, because that happens to be a half-usable platform that can be sold to the people running Windows and OS X (The entrenched end user, who has all the money and wants a fast solution like "web applications", not a good one like "Develop a free software ecosystem")

If you use Qt, you can make decent enough GUI applications for Windows, OS X, KDE, GNOME, raw X11, whatever.

But it still takes expertise, and there's not really any promise of money, so people are just learning the "web" framework, because hey, they already have a client, and there's public servers for testing, and there's supposed to be money in it.

Packaging for different distros is definitely a bitch (I still haven't learned how Arch Linux' package system works), but I think now that many developers are using Git, you can just host a git server or use github and you instantly have the target market of "everyone else who uses Git", which is a lot of developers.

tl;dr: Fragmentation is just a symptom of a general lack of a solution. Unifying the fragments is not a solution. Adding more fragments that fufill the same role is also not a solution.

If we can get people to understand the greater issues that cause these fragmentations (For example, there are many window managers for X11 because X11 is a stupid piece of crap), then we can make solutions that fix the problem directly and elegantly, without the risk of adding more fragments.

Re:hmm

[Anonymous Coward]

That's just not true. Write your code once, and distribute it in source form. Let the distros take care of making sure it fits into their repositories. As for "multiple desktop environments," GNOME can make Qt look native, and KDE can make Gtk+2 look native. That all requires zero effort on a developer's part.

Re:hmm

[Anonymous Coward]

Exactly. I've been saying it for years. Linux has no standard development environment. No standard gui. No standard development libraries and compiler. They're all different version numbers with every linux distro out there. It's why a linux distro is more akin to an embedded OS then a desktop OS.

This is also why android is succesful. Google put a standard on the development environment. These are the versions of each component that you're gonna get and that's it. If there's a bug, it will be patched against the current version, not a newer version that could break your already distributed application. It's also why they used java for programming the apps in. Works against any kind of cpu while keeping the things like graphics etc native but keeping the api calls the same so the java app appears to run fast. Only some of the code is in java that needs to be while the rest just call native c api calls that are built against that particular phone's cpu.

Re:hmm

[Anonymous Coward]

Actually, most serious applications don't have any trouble with that. When developers cut corners on packaging then it becomes a problem, but if you truly want your app to run on any distro, include the libraries with it instead of local versions. This bloats your download, but it gives you much better compatibility. Also, this is true on all platforms.

Re:hmm

[Rich0]

UGH!!!!! NO!!!!!!!

Do NOT bundle your libraries. Just list your dependencies, and the distros will supply them. And if one of those libraries is not backwards-compatible within the same SO versions pound on them mercilessly as there is no excuse for that in the unix world where you can have libfoo.3.1.so and libfoo.2.2.so on the same system.

Half of my job as a distro package maintainer is stripping out all the libraries that get bundled, and the other half is trying to guess what the real dependencies are. Oh, and the third half is fixing the security problems in the bundled code that I missed.

Bundling libraries doesn't just bloat downloads. It bloats RAM use since your personal version of libfoo can't be shared, and it tends to propagate security problems, or at least it forces the user to download all of chrome or whatever anytime libz has an exploit.

App to edit photos and make illustrations

[tepples]

What application do you use to edit photos and make illustrations? Or do you consider that application part of "the dev environment"?

Re:App to edit photos and make illustrations

[spike_gran]

That's a fair point. I guess I do think of a photo editor as part of the dev environment, but, that is really stretching the definition of "dev environment" too far.

But the point I was trying to make, I guess, is that the native apps I use are used in my role as a developer. Rarely do I use native apps in the role of end user.

I think that...

[Yaa 101]

I think that the majority of problems that were in the native environment are solved.
Most of the desktop applications are mature and complete and a big part of the commandline never had problems that werent solvable to begin with.
The last 10 years just caused a lot of problems to be solved on the web platform, now that HTML et al. are getting in mature state as well we will see coming 5 years that most problems on that platform are getting solved too.
Even scalability problems are getting solved with virtual computing.
This means attention is shifting to new platforms like mobile and pad until something new comes round the corner.

Right...

[SmallFurryCreature]

That is why everyone wants VLC to be ported from Linux to mobile devices to finally get a decent player out there. That is why mplayer despite having no interest in doing windows support has people working on turning it into a windows version to get an even better player out there?

What apps are we talking about exactly? Fart apps? Angry birds?

There is a ton of software available for free install. What more do you need?

Or maybe the author is talking about payed for apps? Maybe the genius that wrote synergy should start charging for it? Maybe pay a buck here and there and then it suddenly counts? Okay, my pc will also then cost me a few hundred bucks but hey, at least we got APPS instead of applications. And we can only search for them by the broadest terms and their are ranked by how much their owners spams them.

Seriously, where is the issue?

Re:Right...

[Beelzebud]

Umm Mplayer has been available on Windows for years, and they have windows binary downloads on their own website...

Re:Right...

[Anonymous Coward]

I was wondering this myself. Perhaps it is that so many linux apps do too much. People will download a CNN "app" that does nothing more than open a web browser to cnn.com. Yeah there are stupidly easy ways of doing the exact same thing yourself, but it isn't special because it is in a repository er I mean app market. There is also an insane amount of duplicate functionality in these app markets. Just think about how many twitter clients there are in addition to the stock app and the ability to just go to twitter.com.

But there is another thing. People will literally download apps to show off. One of the early Android apps I remember simulated a Star Trek tricorder. It used input from the magnetometer, accelerometer, microphone, etc to make graphs that couldn't be saved. It had NO purpose, but it was something someone could download and show a friend. Same with fart apps or sound boards. These things have no purpose and will probably only be used a handful of times. No developer for Linux would waste their time on something that will be used for 5 minutes by a small immature subset of the greater public.

Really, the only thing Linux lacks is commercial non-free apps with polish. Think GPS turn by turn navigation with offline maps. Or an offline bird guide featuring bird calls and high resolution pictures. Or blockbuster games developed by major studios.

just extend android app platform ...

[swframe]

I don't want to rewrite my app for a small market. Just make it easier to build android apps to run on desktop linux.

Well...good?

[Anonymous Coward]

As a developer and a Linux user, I'm not sure why I'd write a native app for any OS unless there was a good reason why it couldn't run in a browser. Easy discovery, no installation headaches, no local machine capabilities/libraries/etc. headaches, write once run anywhere (old versions of IE can suck it), easily extensible/mashable via API's, ad nauseam. How many "native" iOS and Android apps are thin wrappers around webkit to get around device API access limitations?

I don't see that as a problem for Linux. Just the opposite. The less you make the OS a central part of the equation, the easier it is for not-Windows OS's to prosper. Who needs to shell out money for Windows if all you need from the OS is to fire up a web browser? With the latest GTK apps you can do HTML5 rendering.

Of course, we're a long way from eliminating the need for native apps, but even in this area Linux is leading. With GTK 3.2 you can render GTK applications in decent (HTML5) browsers. I've seen demos of fairly intensive apps like GIMP running in Firefox.

Re:Well...good?

[Anonymous Coward]

Of course, we're a long way from eliminating the need for native apps, but even in this area Linux is leading. With GTK 3.2 you can render GTK applications in decent (HTML5) browsers. I've seen demos of fairly intensive apps like GIMP running in Firefox.

GIMP is already multiplatform. Why the hell would I want to run it in a browser of all things ?
And the performance, the performance of running shit in the browser is simply apalling compared to native apps, yes even to badly coded native apps.
Really its a non problem. Wxwidgets, QT and other lesser toolkits let you already develop multiplatform programs that work in linux, windows and os x.
What the fuck do you need a browser for ?

I finish saying that if the worst comes to pass, that in the near future it will become impossibile to install FOSS programs in windows or mac os because of stupid app store policies, then why do we have to cater to that demographic ? The same demographic that doesn't give a shit about linux ? Develop applications for linux and if people want to use them they'll have to use linux. You have no idea how many windows users rely on FOSS applications. If those programs were to vanish from the windows/mac world who is going to replace them ? Paid programs ? Yeah right.

Or maybe Linux is everywhere and you don't notice

[cozytom]

Lets see, I am typing this in Linux, on my desktop at work.

I first saw the story on my android phone, running Linux.

There was a WiFi router that was running linux that sent it to my android phone.

Nope, not the year of the desktop.

Slashdot madlibs

[liquidweaver]

Is _____ (Suggestions: Apple, Microsoft) KILLING Linux/Google/Bitcoin _____ ?

For example:
Is MICROSOFT BING KILLING Google SEARCH?
Is APPLE SANDBOXING KILLING Bitcoin MINING?

I think I have the formula fogured out, these articles are actually heuristically generated from statistically high word count topics, and kdawson and Soulskill are actually AIs.

Re:Slashdot madlibs

[sourcerror]

Is PORN killing BARREL ROLL?

Re:Slashdot madlibs

[Anonymous Coward]

This has been going on for years. I think this tool was written in 2003:

Slashdot Story Generator

Do try to keep up.

Re:Slashdot madlibs

[liquidweaver]

That's a fantastic!

Re:Slashdot madlibs

[Lieutenant_Dan]

I find those stories more fascinating than I can find here on /. .

The memes could use some updating. Maybe they make this open source?

Then what's the selling point?

[tepples]

If GNU/Linux doesn't work out of the box with a random sample of ten different users' respective favorite applications, then it isn't likely to work out of the box with other users' favorite applications either. If people have to buy a copy of Windows (retail) to run their needed applications in virtualization, which costs more than getting Windows (OEM) for nearly free with a new PC, then what's the advantage of running GNU/Linux on a home or small business desktop over sticking with Windows?

Why should you care?

[Anonymous Coward]

Usually native-apps are made only for some absolute requirements (performance, feature or otherwise) and then you are usually forced to write platform-native code.
If you can write cross-platform apps that function and perform well, why should you "tie-down" to some specific platform?
Qt is very good at helping make portable apps, once you learn it there's much less interest in learning yet-another platform-specific toolkit/API when one is enough.
Granted, Qt excels at most-used cases (as any toolkit ever will) and things outside of that you must write yourself, sometimes platform-specific.
Writing platform-native code in modern days is more case of "have to" than "prefer to": if there's alternative and/or willingness cross-platform is more preferred than platform-specific.

Just my thoughts..

Garbage, garbage, garbage

[tepples]

You have a point about Diablo and COD: MW2, which are rated Gold or better. But as for the rest:

Adobe Flash CS3 - wine

One of the toolbars doesn't show up according to the AppDB.

TurboTax - wine

Recent annual editions are rated Garbage.

Stone Edge Order Manager - wine

There is no specific entry in the AppDB for this product, but it runs on top of Access 2007. And according to the AppDB, forms in Access 2007 don't work without "an override for oleaut32.dll".

Sonic 3 & Knuckles - wine

Rated Silver for playable in a window. But it crashes in the full screen according to the AppDB because unlike Windows, Wine doesn't support upscaling LDTV resolutions such as 320x240 to SDTV.

Starcraft - wine

The Battle Chest version (bundle with Brood War) is rated Garbage because it can't find disc 2, and the downloadable version is also Garbage because clicking on a unit causes the game to crash.

Street Fighter IV - wine

Rated Garbage.

Re:Garbage, garbage, garbage

[masternerdguy]

As a wine app maintainer, let me say wine is always improving. You should contribute, contributing testing data is helpful in directing development. And using a native DLL to fix a compatibility isn't uncommon with wine. The point is that if everyone thought like you we'd still have punchcard storage for fear of migrating our old punchcard programs over to tape. But from your posts you don't think Linux has a chance, so whatever. The spirit of OSS involves developing then sharing solutions for ourselves and not waiting for everyone else to do it for us.

Re:Garbage, garbage, garbage

[tomhudson]

All very well and good, but the point is that it's a lot cheaper (unless you put a value of $0.00 on your time) to just use the "free" windows that comes with your new computer than it is to use Wine, which can't even run simpler games like SimCity2000 (released 1995) without crashing at some point.

Until Inux gets a single GUI that is used by...

[FlyingGuy]

At least 75% of the distro's it is going to remain a non RAD platform.

The ONLY project that comes close is Lazarus. They have it pretty well worked out. Yes is is Delphi and I know everyone just LOVES to hate Pascal but guess what people as a language it does 99% of what needs to be done and the other 1% is just esoteric stuff that can be done without.

If you want it to do C++ then get onto the project and write the C++ for it.

It is drag and drop GUI interface, right click of the component or double click on the component and write your logic.

It is free software and it works but until the Linux community gels around a single GUI nothing will change and the chances of that happening are about the same of a snowballs chance in hell.

Linux could rule the desktop but because there are so very many variants with all the cooks screaming "Mine is better" it never fucking will.

Re:Until Inux gets a single GUI that is used by...

[rubycodez]

Non-issue, for the major software and desktops can run apps intended for one on the others. Can run KDE apps under GNOME, versa vice. 75% of linux desktop users use GNOME or GNOME compatible thing anyway. Most of the world is going to go mobile anyway, Linux has huge edge there.

Re:Until Inux gets a single GUI that is used by...

[seantide]

It is not a non-issue. The more libraries and garbage you have sitting under your app, the harder it is to keep it all working properly.

The bloat on your typical Linux system makes OSX look lean and mean. Its a mess to write code for. I can run all kinds of things on top of all kinds of ugly layers, but why is that the best solution our industry can offer?

Re:Until Inux gets a single GUI that is used by...

[rubycodez]

strange, I've had zero gui admin issues. Write in gtk or qt, it'll run in lots of places.

There are other aggregators

[tepples]

I searched for Google Reader, and it appears to be an RSS/Atom feed aggregator. There are plenty of such aggregators that run locally, some with a sharing feature to create a new e-mail message with a selected link in the body using the MUA that you have selected in your operating system.

Correction

[IGnatius T Foobar]

Browser-accessed software does not lessen the incentive to work on native Linux apps. It lessens the incentive to work on native desktop apps in general. The number of native Linux apps that aren't getting written is dwarfed by the number of native Windows apps that aren't getting written. Think about it in that perspective.

The problem I have with cloud apps...

[Tastecicles]

...versus desktop-based, frivolously, is the penchant for cloud developers for the addition of payware components in their apps. But, that's the hook, isn't it? You get basic interface or a limited-time run with the advanced stuff, then the advanced components (let's say for online games, premium credits for better ingame gear) come at cost of real-world fiat.

In my world, outdated and quaint as it is, once you buy something you own it. It's yours and its disposition is your business. If you want to use a spoon to open a can, no "license agreement" is going to prevent that, is it?

Re:The problem I have with cloud apps...

[Anonymous Coward]

That voids the warranty!!!!

Re:The problem I have with cloud apps...

[Tastecicles]

I don't frankly care about the warranty - by the time you've shipped a piece of broken kit back and forth (at consumers' expense, usually), swapped out dead parts, spoken for a couple hours of technicians' time, you've bought a new one. If I break my laptop and go buy a new one I'll make sure it's as close to being the same model as possible so I got a heap of spare parts.

Re:Quick question.

[TheRealGrogan]

Just how is XFCE anything like Windows 95, because it has menus and a little x to close program windows? If anything, it's one of the least "Windows-like" clients. It can in fact be very different depending on how you set it up. The way it looks and behaves out of the box can be changed in about 2 minutes, through the GUI, such that you wouldn't recognize it. With Windows you would need third party shell replacements to even come close to the functionality of XFCE.

I have not tried Gnome 3 or Unity because I hate Gnome (and Ubuntu for that matter) but KDE 4 is getting to be pretty good.

Re:Or maybe Linux is everywhere and you don't noti

[Anonymous Coward]

Cool anecdote, bro.

Requiring new versions of Windows

[tepples]

A lot of that had to do with Microsoft torpedoing the standardization and advancement of HTML. They seem to have stopped doing that now.

IE 9 requires Windows Vista, meaning we can't rely on HTML features introduced in IE 9 until April 2014 when Windows versions prior to Windows Vista enter end of life. IE 10 will require Windows 7, meaning we can't rely on HTML features introduced in IE 10 for the foreseeable future, or at least until five years after Microsoft announces the end of mainstream support for Windows Vista.

Re:Requiring new versions of Windows

[jonsmirl]

Chrome runs fine on all versions of Windows.

Re:Requiring new versions of Windows

[shutdown -p now]

You don't need to wait until the official EOL, it's not like it's a contract that's binding anyone but MS itself. It few enough people use XP or Vista, just drop support for them. On the other side, if many keep using it after EOL, it's still not a ticket out for you.

In any case, for a HTML app that doesn't have to be a full-fledged client/server app - and, in case of desktop app replacements, that would be pretty much all of them - you don't need to rely on IE. Just embed Gecko or WebKit or whatever you fancy into a simple window with minimal chrome, and host your app in it. Or use one of the many existing solutions for that. That way, your app does not have any browser dependencies at all.

Re:Requiring new versions of Windows

[Anonymous Coward]

In any case, for a HTML app that doesn't have to be a full-fledged client/server app - and, in case of desktop app replacements, that would be pretty much all of them - you don't need to rely on IE. Just embed Gecko or WebKit or whatever you fancy into a simple window with minimal chrome, and host your app in it. Or use one of the many existing solutions for that. That way, your app does not have any browser dependencies at all.

That sounds fantastic, every app with its own rendering engine. Is that really a step forward?

Re:Requiring new versions of Windows

[shutdown -p now]

That sounds fantastic, every app with its own rendering engine.

It's not any different from every app coming with its widget toolkit, which is the reality for e.g. all commercial apps using Qt and running on Windows or OS X.

And no, it's not a step forward for anyone, really. It just seems like it's going to be the easiest - if not the only - way to get some semblance of cross-platform GUI compatibility in near future.

Linux Software Sucks (But There's Hope)

[vinn]

I'm all for SaaS. Here's the problem the free software community has run into for the last decade: as it's gotten more popular there's been a call for more an more apps. However, there's not enough developers or developer's time to create those apps. It's because a few guys are over here making PIM utility for KDE, a few are over there working on timeclock utility for GNOME, a bunch are working on packaging this and that for Ubuntu and a bunch are working on packaging things for Fedora, etc. All the while the real apps, like a great office suite, get neglected. Let's face it: Linux apps suck, the Linux desktop experience sucks and the Linux experience of managing / integrating with a complete enterprise (user accounts, desktops, laptops, Windows domain controllers, shared network filesystems) horribly, horribly sucks.

If we could take the energy spent on developing distribution specific things, desktop specific things or even entire classes of applications and instead concentrate that effort into shared things with less duplication of effort, we'll generate more developer time. That developer time can be used to make the Linux desktop much more enjoyable.

I want to have a single sign-on system with my Windows systems, shared network drives with the Macs and Windows systems, the same login scripts, shared applications or at least similar enough applications users can move between systems, and shared document formats. "Sure", you say, "all that exists right now." Yes, but I want it to work out of the box like when I turn on a Windows laptop for the first time or a even a Mac (to a slightly less degree.) I want to go from opening the box to putting it on a user's desk in 10 minutes regardless of the OS. SaaS helps immensely with that.

Buying proprietary apps that end up broken

[tepples]

As a wine app maintainer, let me say wine is always improving. You should contribute, contributing testing data is helpful in directing development.

I have seen several cases where the demo works but the full version doesn't. If I'm to be buying copies of proprietary software just to end up unable to use them because they're Garbage nor to exchange them due to retailers' return policies, how do I recoup this cost?

And using a native DLL to fix a compatibility isn't uncommon with wine.

As I understand it, obtaining a lawful copy of said native DLL costs $200 for a copy of Windows. Or what am I missing?

The spirit of OSS involves developing then sharing solutions for ourselves

I agree. So how does one fund the development and then sharing of viable substitutes for those applications?

qtcreator and monodevelop

[jopsen]

Have you tried recent versions of qtcreator or monodevelop? Actually pretty good, the only thing they lack is package generation.

Re:qtcreator and monodevelop

[Anrego]

I'll pass on mono, but qtcreator actually looks pretty damn encouraging.

Thanks!

Re:qtcreator and monodevelop

[thetoadwarrior]

mono, lol

SaaS for all OS's

[jellomizer]

In really SaaS is not cutting Linux native apps, It is cutting all OS native Apps.
This is good.
Think about it, you can be running Windows, Linux, Solaris, Mac, and Android Device or an iOS device and still access all your information... From anywhere... And up-times are excellent, I mean a 1 hour outage every 3 years isn't really that bad, Most companies will not be able to keep their services running that well. This is Huge. We just never realize it as it came up quietly.
Yes it does come with trade offs. But a lot of these trade offs are not technical it just requires the end user to be a little smart.
1. Get a good contract. Just don't use the canned one that they use for everyone. If you need it and it is critical to your application you better make sure the contract that you sigh for the service is in your favor.

2. Have a backup plan... Just like you need with your own in house stuff you should have a backup plan. Down load your data or transfer it to an other SaaS. You can always be sure you can get your data due to #1, make sure that you have a contract entitle you to your own data, in a format that is exportable.

3. Keep an eye on your SaaS company health. Just like you need to do with your normal software. How many people are still using Word Perfect for Dos just because they can... The same with your SaaS company, If the service is getting out of date you should plan a switch to a new one.

SaaS is a good model, it frees the user from having to choose a particular OS and gives them global access to their data.

Ubuntu .. and Linux in General

[AdmV0rl0n]

Far as I can see - Linux spent a few years in desktop terms - with video and people pushing it as a nice computer platform. And how did they do this? Was it by presenting broken, limited, badly designed, poorly thought out and new desktop window managers?

No, the whole thing was not based on Gnome2 or KDE, but it was often driven by Compiz. I know this is a gentle over use of generalising something - but its also true. If anyone thinks that people were being exited by Gnome 2 or KDE 4 - then they are nuts. And none of the others are anything new really. They are rehashes on what was around in various flavours.

Unity (ha, I laugh at that name, when they chose it they must have been in the pub and knew using the largest dose of sarcasm possible was apt) - was just someone senior - drawing crap up on a napkin, And the stories of 13 windows users, one OSX mac user and three monkeys called Tom, Dick and Harry just about fit. Its for tablets!! Yeah - how many tablets are in your eco-system. Well.. er... none right now, BUT there soon will be. Yeah. That was what - two years ago. So how many now Mark. Er.. well, we have some ARM stuff floating around. here.. there. Yeah where can I buy them mark. Er.. well you can't really. But soon!!

And the moment I start reading stuff like we won't put stuff in that MS is pulling out of 8,9,10 is the moment where I think insanity has properly struck. If others are cutting back their offering, reducing choice, you should be looking to fill the gap. And no, this is not a rant against tablets. Its an absolute rant against the theory that people are taking their very good sound desktop and laptop computing platforms - and are decimating them because of 'Tablets'.

I could apply the same to the lunacy of Gnome 3. At least Gnome 3 seems to have some loose idea of being something that is building up. Unity just seems to be broken. we won't fix it, and here, have what you are given. A desktop being presented with more and more cut down. And all for tablets. And seemingly by default no user testing from its own users in the offing.

It would almost seem as if Compiz became a pinnacle - and having seen that, devs are running away with 'simplify' as a goal. Compiz drew in more people to Linux than anything Unity has done.

Personally - I think shuttleworth should stop pushing Unity as the global desktop of his choice. And beyond that, while Linux offers choice - which is fine, choice implies that you have a wide variance from one end of a spectrum to the other. Choice ceases to exist if everyone produces poorer desktops, reduces function and capability, and cuts features. So everyone has to do what, Move to XFCE or E17 or whatever? Are you kidding me? Thats the answer? All the mainstream ones have gone wrong, so into the lifeboats everyone jumps. I don't think the current desktop situation in Linux really is offering true choice. Yes, you can choose one of 12 desktops, and guess what, every single one of them is heading for a mess, is unpleasant, is heading for being a tablet, or is a retrograde step backwards. Not one of them is actually a true step forwards. Most are a side step to give the users back what Shuttleworth is screwing up.

People love to take something like a computer. And they spend time in it. They reach in and put their ID there, and its an extension of their persona. The web is filled with people who have taken a desktop and they amend it and customise it, and make it their own. The idea of uniformity to a group of developers is a sound one, I get it. You want very sound ground rules, and APIs and fundamentals. But thats what you build a structure on. Its what you put a customising set of APIs on top and its where you hand over something that the users then wish to customise. But you are not building a desktop - or indeed any real product where you want others to use it for THE DEVS. You're supposed to be building something that people can step into, use, and then make their own.

The moment you get an over arching commentry from the devs

Re:Ubuntu .. and Linux in General

[Anonymous Coward]

relax and enjoy one of the most important things that linux gives you for free ... freedom.
and ubuntu is not the only distro, take a look at debian if you don't like ubuntu's "defaults".

so lean back, smoke some weed and focus on the tools that are given to you as a gift to do something creative.

cheers
/ vx

Nope

[MBGMorden]

I think what's killing native Linux app development is that most of what needs to exist already does.

Seriously - while some programs could use some tweaking (IE, GIMP isn't quite as robust and capable as Photoshop, but it does similar things and is good enough for most casual users), just about anything that you'd want to do for day-to-day stuff there's already a native "app" for that.

The only time I find Linux lacking is for video games, which as an entertainment medium follow a different model than utility/productivity software. Aside from those, I sit down to my Linux box every evening and never does the thought cross my mind that I need something that isn't there.

That's not specific to Linux either - I choose to use Linux at home but the same is mostly true for Windows and Mac too. The desktop computer platform has matured to the point where there's simply not a lot that needs to be done anymore. Eventually you stop trying to make a better hammer or wrench and just start using them without care to their improvement.

IE is installed. Chrome isn't.

[tepples]

For one thing, if your application requires Chrome or Chrome Frame, and organizations' IT departments have made it a policy to block Chrome and Chrome Frame, organizations will not adopt your application.

For another, even a user unrestricted by IT-department-imposed application restrictions doesn't want to install your application and then suddenly find that it doesn't work in IE and have to wait until the next opportunity for an Internet connection to download Chrome.

there's whole universes between 0 and 1

[Thud457]

Ya know, maybe Linux doesn't need to take over the world.

It's mere viability is enough to keep the close source folks moving forward (?).
(Parts of) It have moved pretty well into the embedded and phone spaces.
What Linux needs is hardware support and interoperability.

Linux doesn't need the Eloi. And they don't need(?) Linux -- they have Winders, Mac, IOS and Android (lol, wat?)

Key facts

[Anonymous Coward]

1. This could happen to Android if all programs fail to work on all devices.... screen size, CPU, GPU, UI overlay, etc;
2. Does anyone see the appeal of a terminal in today's world?

You have all your company's data in the cloud to be able to access it and have "redundant" copies. Hurricane/Tornado/earthquake/etc hits, and there is no access to the Internet/cloud. I'm working, you are not. While it may be "rare," there really is no need to have your data stored off-site to be secure or redundant.

I don't want a terminal on my desk or in my hand (Siri/Google's voice search). I want my own workstation :)

mike mcgrath

[nimbius]

from the fedora project gave this speech at fudcon tempe 2011, and left most of the room thoroughly trolled. I think he titled it "how the cloud is killing linux" or something to that nature.

"the cloud" is just another way of saying "closed source." if we fought it before we will fight it again, and i predict for alot of the same reasons as well as emerging concerns like privacy. the strength of the open source community is the borg-like adaptability to change and our sterling motto of 'hell hath no fury like a hacker scorned' has shocked everyone from RAID vendors to Wireless companies. im not saying we're getting a desktop linux renaissance tomorrow, or the next year, but linux has always stood for freedom in most respects and for many its not something traded so easily. for example, had facebooks image storage service been so wonderful, no one would have developed the importer tool for digikam that sucks down all the images for a particular user. there will always be a bashpodder user, and that user will always be empowered to become a developer if and when she desires.

Re:mike mcgrath

[Anonymous Coward]

"the cloud" does not have to be closed source. See LibreOffice going web-based.

Re:mike mcgrath

[jon3k]

Cloud doesn't necessarily mean closed source[1][2]. We should be working towards open source and standards based "cloud" platforms. Fighting against the concept of "cloud computing" is a misguided, losing battle.

[1] http://www.cloud.com/
[2] http://www.openstack.org/

linux is in trouble, desktop and server

[sloth jr]

[note: linux puts bread on my table and has for 10+ years and I've been soaking in it since the early 90s]
What I see in Linux is not good; the problem is ironically the bit often touted as its success: freedom and flexibility.

IF the linux community would like to see mainstream adoption (it's not clear that it does), then a lot has to change; distros have to die and die hard. The multitude of desktops and package managers (RPM/YUM versus APT) need to die. There needs to be ONE way of doing things. Linux is often touted on the server side, and deservedly so. Unfortunately, increasingly, more and more of the server applications are being created non-native, in shit like Java. Yay, it works on any compliant JVM! So why does that JVM have to be running on Linux? There are fewer and fewer native server applications. It's all Java or some stupid shit horked up in Python or Ruby or PHP. There's nothing that inherently drives any of those applications to Linux.

Consider OpenStack; great, cloud-controller software, abstract means of firing up and provisioning VMs, storage, and networking. If you're running on Ubuntu - hey, super, first-class citizen. If RedHat/Fedora/CentOS - fuggedaboutit. nova works well-enough in diablo, RPM packages out there. Glance and Swift? Good luck. I'm sure you can get it working, despite shitty documentation that barely admits other distros besides Ubuntu even exist. But why the hell should I have to fight these battles every time I want to install anything new? It's a disaster that mirrors exactly the sort of diversification issues that helped put a spike in commercial Unices of the 90s.

There's no way Linux will "die", it's still a great OS on its own merits and the price is definitely right. Hobbyists and hackers will always run it. Just don't expect that it'll ever become mainstream without brutal community choices. http://xkcd.com/927/

I work on Linux Apps

[AdamJS]

Because I frequently encounter a situation where every single implementation of a given software idea thus far on Linux is horrendously dreadful, suffers from feature-creep or is over-engineered to the point of actively working against its users.

A good example is comic book readers.

i don't care about os's any more

[Anonymous Coward]

i write apps to do a specific function. this code can execute on any system that has an interpreter.

why would i care about writing the same things for a specific OS?

I don't see the problem

[Anonymous Coward]

Is it a bad thing if linux never becomes mainstream if doing so means losing the freedom that we so cherish ?
People are obsessing around the "one true way". Well guess what, there isn't a one true way and thats why I love linux. I can't be the only one to think so.
If people want to be shackled (one desktop, one gui toolkit, one way to install apps, no freedom etc...), hey they can use os x or future windows 8. Less choice the better right ? Is this the new motto ?
If yes, then fuck the new motto.

Linux has survived long before becoming mainstream, and will continue to survive long after mainstream stops giving a shit (if it ever did in the first place). The only people affected by this are the artists/designers that have infected KDE and Gnome to think that their goal is somehow to cater to the windows users with glitz/blink all the while throwing in the drain important concepts as usability and productivity.

Re:Right...makes me realize I'm out of touch

[shoor]

I honestly don't know if SmallFurryCreature is being sarcastic or serious. Does everyone in fact want VLC ported from Linux to mobile devices to finally get a decent player?

I don't have a mobile device, nor do I feel a need for one. But then, I don't even like telephones (and I haven't like them since I had a job in the Navy many many years ago when I had to be on the phone all the time.)

I do like being able to get on the internet from the convenience of home with a nice big LCD screen. I use Linux and haven't used Microsoft since the Windows 3.1 days, but then I was programming in a Unix (mostly BSD 4.2) environment in the 1980s. I realize everybody is not like me and I'm not saying everybody should be like me; but, is SmallFurryCreature being sarcastic or not?

Plenty of us use native apps

[freeweed]

Native apps that I use in Linux every day:

Clementine (audio player)
Xine (video player)
Musicbrainz (mp3 tagger)
Google Earth
Pidgin (IM client)
Firefox
Geeqie (photo browsing/basic editor)
Minecraft (duh)
Open Office
Kate (text editor)
K3B (burning software)

And this is just the stuff I can think of off the top of my head while at work. On top of this, there are dozens upon dozens of other apps I use less frequently, but regularly. About the only web app I use that's replaced a native desktop app is Gmail.

I suspect you simply do a lot less "user" type stuff than most people. Pretty much none of this could be replaced with web apps, at least not yet. Maybe Google Docs/Picassa could take out one or two things, if I hosted everything I did on the web. Google's storage limits severely curtail that type of activity in my case.

Without native Linux apps, I'd be back to Windows in a second. Not by choice, but due to lack of it. Or maybe I'd buy a Mac.

Re:Quick question.

[jbolden]

I think you forgot Android. There is a successful Linux based UI now.

Unity and Linux

[OldposDells]

We'll see what happens. When Windows 8 releases there might be a push on Linux. Unity will hopefully get better. Someone should really just fork Gnome 2.X.

Dump Ubuntu Day - Next Saturday!

[Anonymous Coward]

This Saturday go out and move your money!

So?

[jon3k]

Good riddance. Web browsers and web based applications provided something compiled languages only ever really promised - real world portability between platforms. I'm tired of maintaining software locally, can you imagine the collective hours we've all spent updating applications on millions of desktop computers?

Not to mention the fact that I don't want to be locked into a set of *nix applications anymore than I want to be locked into a set of M$ or Apple applications. I should be able to move seamlessly between platforms at my convenience. This is a GOOD THING not a BAD THING. Web based applications means there's no more "I can't use Linux because it doesn't have application Y" - as long as you have a web browser you can access any application. If anything the new paradigm should increase Linux adoption.

crap

[l3v1]

The ignorance of the lines of thought in the article are the same crap that pops up every once in a while from people who can't fathom [at all] that "apps" can be more than twitter, facebook and gmail clients. There is more you can do with a computer you know, and there are a million tasks for which the oh so precious "cloud" and saas apps are not suitable and can't ever be - and it's not just about trusting your data to the "cloud", it's about having your data handy with fast access locally all the time with enough local processing power and memory to handle computationally expensive tasks, not having to pay for storage, access time, processing power, and so on and so forth.

I know, I know, this is the online generation. Shove it. The two concepts (having extensive saas ecosystem and a normal local hw, os and app environment) should not be exclusive, they should exist in parallel. Making design decisions solely based on the "saas rules" stupidity even for a local environment (think ubuntu or win8) is crazy. People can adapt, of course, if all options are taken away (luckily it's not the case right now), but that doesn't mean the choices were good.

SaaS?

[Spugglefink]

I've spent almost all of my free time today doing development work on a native Linux application, and I have no clue what "SaaS" is without going to look it up.

I'd say that means the idea that this SaaS thing is killing native Linux app development is probably just a wee bit overstated. What on earth does it stand for anyway? Space alien anal Sex? Special anchovy asparagus Sauce? Sending apps across Saturn? No, wait, don't tell me, I'll guess eventually.

Of course it's just as fair to say my app isn't really all that relevant in the modern world, and it will die eventually. My team is a shadow of what it once was, and so am I, in terms of my devotion to development work. Frankly I spend entirely too much time goofing off here on /. to get any volunteering done, and that's largely because the audience has gone from a cheering mass to a few sporadic crickets chirping in the dark.

Linux on the desktop is dead. Long live Linux on the desktop.

We could port to Windows, but who wants to work for free on a platform where people get paid lots of good money to work?

CLI for home users

[unixisc]

Hairyfeet

Between the 2 OSs - Windows 98 and Windows NT, which one do you think was targeted at home users? 98, right? Under your reasoning, MS should have done the reverse - positioned NT towards home users, and 98 towards Corporate users, where one was more likely to have admins familiar w/ DOS commands. Although to be fair, DOS had an extremely limited set of commands and no choice of shells - like from ash to zsh, and so it wasn't difficult to get a quick handle on stuff. Also, installing anything from DOS was just a question of typing 'Setup' or 'Install'.

But even under XP, there are times you're forced to go into the command shell. Here's an experiment - the next time your internet connection is not working and you call up their tech support, note what they tell you. Everytime I've called, they'd ask me to run cmd, and then under the DOS prompt, do a ping and give me an address to ping. If one were to get rid of the DOS subsystem in NT/XP/7, how would one check for such connectivity? For whatever reason, MS doesn't have an option where you click Control Panel -> Connect to -> and then enter an IP address.

As to the question of whether a Linux could be done without CLI, it absolutely can. Go to distrowatch, and download the GNUSTEP OS, which is a Debian with GNUSTEP as the only UI. I dunno whether you were familiar w/ NEXT, but in college, when I struggled w/ SunOS, Ultrix, AIX and the like, I found in our computer lab NEXT workstations which were absolutely neat. Go to the Workspace Manager, and you'll see the entire directory tree. Clicking was all one needed to do (I'm not sure whether the install required a text driven instructions). And if one had to, one could edit the configuration files sitting in the GUI. There was no CLI mode - you had to open a terminal within the GUI.

I think that once GNUSTEP comes as an option on most distros (like it is on all BSDs), it's more possible for there to be a fully functional GUI where CLI is not needed. If Apple can do it, any distro can. Why they choose not to is another story altogether.

Re:CLI for home users

[hairyfeet]

Wow I hadn't thought of NeXT (yes I am old enough to remember who it was written, man I wanted one of those cubes back in the day!) since jobs went to Apple. and while i have no idea how it handles (don't have any spare PCs ATM) All I want is a nice, simple, GUI based, pretty OS that I can guarantee will actually update for the next 7 years without crapping over its own drivers and choking, is that so much to ask for? you mentioned XP but I have XP boxes in the field that's been there since 2001, that is 11 years, THREE service packs, and about 1500 updates and NO driver failures, and NO CLI fixes required. the most I had to do to them was upgrade the RAM when it got cheap. Hell I have a customer that just NOW retired the Win2k box i built him in 2000!

Oh and finally as for your XP instruction given by the ISP? That is caused from something going badly wrong NOT from the OS simply updating itself! There is a BIG difference from needing CLI because you had a network outage and needing CLI because you dared to apply updates, totally different things. if the only time I needed CLI was when the network went down in Linux? I'd have NO problem with that. But when Dell has to run their own repos because even THEY can't get decent enough QA in Linux to keep their machines running without custom building their own distro? I'd say there be trouble in paradise friend.

All I want is a distro that will work and KEEP working without my customers having to drop into term for the life of the box or 7 years, whichever comes first. Is that REALLY too much to ask for? And I don't care if they have to upgrade once a year or two as long as the upgrade process is as simple as "click button, let it do its thing, reboot". But at least in my experience it has been "run upgrade, reboot, find broken hardware, enjoy three days of trying to get it to work again, oh and enjoy the forum hunts".

The reason I'm against CLI is because frankly i think its become a crutch. The GUI is a second class citizen compared to "open up bash and type" which is treated as the number 42 in Linux. So I truly hope that somebody proves me wrong, I truly hope that by 2014 when me and hundreds of thousands of retailers are sitting on pallets full of XP boxes and laptops they'll be a version of Linux we can just slap on and sit in the window. But so far I've gone through over a half a dozen and have yet to find one that would fit the bill. I might give GNUSTEP a spin when the next load comes through in a week or two, but it'll have to have Flash support and be able to work with all the major sites like FB for it to work for my customers.

Driver support w/o CLI

[unixisc]

Honestly, there needs to be a central repository of open source device drivers (needn't be fully optimized) for all the major pieces of hardware - wi-fi cards, network cards, graphics, printers, audio and so on. Oh, and list it by manufacturer & model number - how on earth does one know which wi-fi chipset is being used in one's laptop? At least this way, any distro that wants to be popular would have absolutely no excuse not to support something critical, like networking or wi-fi, or DVD. And yeah, it should have ways to be installed from all the major package managers - apt-get, yumm, et al. No doing a make after extracting a tarball.

I believe that there the Linux Foundation has such a project, in which case, it should be a central rallying point for hardware vendors and Distro authors to come together and ensure that all products listed are so supported.

Re:Driver support w/o CLI

[hairyfeet]

I'll give you one better, put the "80%" as I call them, AMD, Intel and Nvidia chipsets and GPUs, Broadcom and Atheros wireless, Realtek and SiS networking, Realtek and Simatel sound, and put them on a disc similar to what we have with Windows. I have a pair of discs, one 32bit and one 64bit, that I can put into ANY box or laptop and it WILL have the driver. There REALLY needs to be something similar for Linux. like you said they don't have to be optimized, hell don't have to be the most current, just something that will "just work" and allow someone to get ANY machine up and functional. After that as you said a central driver repo could take care of the rest.

The sad part is despite all the experience Linux has with repos MSFT is actually gotten ahead of them on this. When I left my dad's new PC at his home and told him I'd be back on the weekend to install Win 7 he got impatient and decided to do it himself, thanks to the new Windows Update which runs during install and downloads any and all drivers you need (it comes with just about every generic networking driver, along with a basic driver for most others, the rest is hosted at WU) when I got there the machine was running beautifully, didn't even need the driver disc. it all "just worked".

I truly want Linux to succeed and carve out a decent niche, I really do. I'm old enough to remember when the market was plumb full of choices, we had Atari and Commodore, Tandy and Apple and IBM clones up the butt, plenty to choose from. Hell I even think it would be better for Apple and MSFT as they'd have to stay on their toes. But until the driver problem is fixed or there is a hell of a lot longer support schedule I just can't carry it, the updates/upgrades just break too much stuff. I need an OS that once i hand it to the customer they don't need me anymore except for hardware upgrades or replacement, with a good AV I can get that with windows. With Linux I just end up spending all my time dealing with forums and looking for fixes. Surely there has to be a better way!

The solution to Our Division(tm)...

[ameen.ross]

...is ironically NOT Unity.

Then what is? Debian. Most distros are based on Debian to begin with, including Ubuntu. Debian stable is still on Gnome 2 by default, which suits those who are more conservative i.r.t. GDM's. As for me, I'll be switching to Debian Sid, which is on Gnome 3 by default (and I like it). Gnome 3 even has gnome-panels in fallback mode, which is awesome.
The best thing of all, is that this single distro offers about as much freedom of choice as all distros combined. Most users stay with the default though, so that gives developers a steady target.

No

[TheSkepticalOptimist]

Linux is killing native app development.