Survey: Techies reckon open sourcery has better prospects than familiarity with a single vendor's cloud wares

Getting skilled up in open source could be a better career bet than focusing on a specific vendor's cloud technology. Or so states IBM sponsored research from O'Reilly Media, which paints a picture of open-source technologies proliferating as cloud adoption continues (be it public or hybrid). Earlier research from the IBM …

[AMBxx]

Sounds desperate

No longer relevant company desperately clutching at straws to validate own view of world while laying off all experienced employees.

[jake]

One thing people tend to forget about FOSS.

By it's very nature, FOSS will be around as close to forever as makes no nevermind. Corporate closed source software, on the other hand, is just as ephemeral as the company in question. IBM is doomed to die, eventually. So are Amazon, Goophabet, Apple and Microsoft. Shirley the proverbial Thinking Man should throw their lot in with the obvious long-term winners and eschew the obvious losers?

Before you poo-poo this, think about it. Where are Burroughs, Sperry, Allied Signal, Philco, Amdahl, Remington Rand, DEC and ROLM? We won't mention the likes of HP, the poor mewling thing, so senile it doesn't know it's dead. And that's just for a start.

[Pascal Monett]

Re: IBM is doomed to die, eventually

Agreed, but its intellectual property will be recovered by someone else, or maybe many someone elses.

So there may be hope for support in the long run.

[jake]

Re: IBM is doomed to die, eventually

There are many patents from now defunct companies that stop anybody else from using their technology, often even in a support roll.

Note that I'm just the reporter, I don't necessarily agree with it.

[max allan]

Re: One thing people tend to forget about FOSS.

FOSS as a thing will live on. Companies as a thing will live on.

Specific FOSS or comapnies will die. You only have to spend 5 minutes.looking around github at the number of abandoned repos, often for once working software that are now stale and won't even build with supported (eg no glaring security holes) versions of compilers/interpreters/operating systems.

That doesn't even touch the surface of stuff that was never completed or got removed for whatever reason (Truecrypt for example)

[Anonymous Coward]

Re: One thing people tend to forget about FOSS.

True, but at least if a FOSS project is abandoned its source code is there for someone else to pick up.

For example, a client of mine recently migrated an old system to a newer OS. One of the dependencies hasn't been updated since 2005, but it is open source and still on GitHub. It wouldn't compile in a newer build environment but as the source was available, I was able to clone it and fix it. The client now has a modern patched version of what was a defunct project that will compile on a new OS.

I got paid sweet, sweet money for the time and effort too.

I can't do that with dead proprietary software.

[EmilPer.]

single vendor cloud wares ...

... had to deal with $title, they were extensively documented, every minor version had its own page, every option named etc., but no story to tie them togeter, no clue about how to actually use in a real life scenario

got the paid help and got it working, turns out some concepts were called using two different terms, and some different concepts were called using the same term in different contexts

on top it off there was very little information about failures on the net, even on stackoverflow

opensource might not be perfect but at least I can read the source when I don't understand the docs

[Anonymous Coward]

Re: single vendor cloud wares ...

I read the start of that comment thinking you were talking about open source.

Having worked on one open source, all the new shiny web project i can confidently say that problem is not unique to closed source. We had a load of different libraries, all with completely different opinions about how enterprise code should work. It was a mess and being able to read the source code was no help. If the concepts don't align the code won't play nice.

[EmilPer.]

Re: single vendor cloud wares ...

you are correct, open source in itself is no guarantee, but at least you have a chance if you have access to the sources

[Pascal Monett]

"94 per cent reckoned FOSS code was as good if not better than the proprietary option"

Seems obvious. FOSS code is public, therefor if you code like shit, you're going to have it up for the world to see and the world will tell you you code like shit.

Proprietary, on the hand, allows a bad coder to do his worst, secure in the knowledge that nobody will see his code so it doesn't matter.

Of course, there are very good coders that write proprietary software, but hiding the code means you can hide mistakes and bad practices - and with the amount of bugs due to buffer overflows that still happen, well let's just admit that there are more bad coders on the proprietary side.

[jake]

Re: "94 per cent reckoned FOSS code was as good if not better than the proprietary option"

I think they were talking about the functionality of the code, not the beauty of the functions and algorithms (etc.).

[A.P. Veening]

Re: "94 per cent reckoned FOSS code was as good if not better than the proprietary option"

Wrong, they were talking about the amount of errors in the code.

[jake]

Re: "94 per cent reckoned FOSS code was as good if not better than the proprietary option"

How do you figure? Re-reading the article, I see nothing at all about errors (or lack thereof) in the code, just commentary on usability.

[A.P. Veening]

Re: "94 per cent reckoned FOSS code was as good if not better than the proprietary option"

"Better" implies some way to measure. One easy way to do so is count the number of bugs per amount of code.

[jake]

Re: "94 per cent reckoned FOSS code was as good if not better than the proprietary option"

Nah. "Better" is subjective, no measurement required.

[sabroni]

Re: secure in the knowledge that nobody will see his code

Have you really never worked in closed source? You think we don't do code reviews or share code with each other?

"I can't fix that, Gary wrote it so i don't have the code!"

[max allan]

Re: secure in the knowledge that nobody will see his code

I have worked in closed source. And all the really gnarly bits of code get left alone.

"Gary wrote that, he understands it, I don't, it passes some tests : approved"

I spent about 3 days trying to empirically demonstrate a bug. Turned out to be an off by one error. And while they were in fixing it, because it was the first time anyone but the original author had checked it, properly they found a calculation that had been optimised out of the loop that needed to be in the loop.

This on very specialist software selling at 100s of £k per seat. We just released the new version as "optimised and improved accuracy" and never told anyone what was in it.

[A.P. Veening]

Re: secure in the knowledge that nobody will see his code

In similar circumstances (and before anybody complained about the error), I used "Preventive maintenance", which was even absolutely true, it prevented complaints from customers.

[ecofeco]

Here's a wild thought

How about no cloud at all for anything critical?

Yeah, crazy talk, I know. After all, what could be more secure than using some one else's computer that you have no control over for your important data? What could possibly go wrong?

Oh wait...

[jake]

Re: Here's a wild thought

That is crazy talk! I mean, if you were even partially correct it would mean that hundreds of billions of dollars (perhaps tens of trillions) world-wide are tied up in something that is intrinsically not secure! Shirley all those CEOs can't possible have made a mistake by jumping on the Cloud bandwagon, could they?