top | item 10198484

What the IBM Acquisition of StrongLoop Means for the Node.js Community

244 points| ijroth | 10 years ago |strongloop.com

130 comments

order
[+] rynop|10 years ago|reply
former IBM dev here (6 years in different areas). I hope they let you work on an island and don't force you to use their tooling/approve listed of open source software.

IBM has long embraced the OSS community (I loved LTC), but the process to release/use anything OSS was not good (to put it nicely) and was one of the many reasons I left.

I worked on a project that was on an island, and while it was more efficient than normal IBM dev, it was still 10x slower and tons of red tape then what was required to make a competitive product (and stay competitive). IBM is big and ultimately has to protect themselves - IMO ultimately it gets in your way of making a good/competitive product.

I like what StrongLoop does for the node community, and I do indeed hope you are successful and nothing changes. But to be honest I am very skeptical.

[+] kodablah|10 years ago|reply
Current IBM dev here (but only speaking anecdotally for myself and my project), can confirm that when we were acquired they let our product and policies remain and did not "blue wash" the project. I would assume the same will occur for StrongLoop and their software practices will probably continue mostly unfettered.
[+] rilut|10 years ago|reply
Sorry, what is work on an island?
[+] tracker1|10 years ago|reply
As much as I fear this will drag StrongLoop, I feel that this is an excellent fit. StrongLoop has been focused on building "Enterprise" tools and module around node, as well as working on whitelisting/reviewing various modules... This is a good fit for IBM and their clients who are inclined to also want to have "approved" or "certified" modules to be able to use separately from the rest of npm.

As much as I like the power/openness of npm, a lot of corporate environments move more slowly... modules need to be cleared by legal and usually limited to specific versions. Having more resources to do this is a good thing and can only help people who are working for financial institutions (as an example).

[+] mbesto|10 years ago|reply
> And it’s not just Node.js. Maybe you haven’t looked at what IBM is doing with open source lately. I was surprised when I dug in. For example, did you know they are leading contributors to:

Linux OpenStack Cloud Foundry Docker plus many Apache projects like Spark, Cordova and Hadoop and of course, Node

They are? (keyword => "leading")

https://github.com/docker/docker/graphs/contributors

https://github.com/apache/spark/graphs/contributors

I find this claim dubious. Maybe I'm not digging hard enough.

[+] nickstinemates|10 years ago|reply
Speaking just about Docker and not commenting on this post in particular: IBM employs at least 5 people who dedicate their time to working upstream. Duglin, as an example is #11 on the contribution graph in your list

disclosure: I run partnerships for Docker.

[+] kitd|10 years ago|reply
IBM are working on the runtimes for things like Node, Spark, Python, Docker, and augmenting them to perform well on IBM platforms and integrate with IBMs enterprise monitoring and diagnostic tools.

Some of these augmentations have been contributed back to the main projects.

[+] jacques_chester|10 years ago|reply
On Cloud Foundry they are a major contributor; only Pivotal donates more engineering effort (disclaimer: I work for Pivotal Labs).

IBM has a lot of working engineers at Cloud Foundry HQ in San Francisco. Last I checked IBM-CFers were the whole CLI team and they're doing great.

[+] draw_down|10 years ago|reply
What is it about these posts that bring out the HN contingent who think themselves firebrand skeptics? This is basically a marketing piece, what good is fact-checking it
[+] mbubb|10 years ago|reply
From the customer side - I dealt with Netezza and Softlayer before and after IBM acquisition. In both cases an almost palpable drain in support and intelligence. I hate to say it like that as i know that there are brilliant folks in IBM (Watson, etc)

Softlayer and Netezza in different ways were smart, nimble and fearless companies. You had real relationships with the engineers. I got to know Netezza folks in Massachusetts, Poland and Australia - some of the smartest folks I've met. They shared scripts and passionate expertise.

IBM took it over and the bureaucracy set in. The term "TAM" brings tears of rage to my eyes... Opening a support ticket is about as hard as applying for a mortgage online. And they want to have these endless conference calls with 7 or 8 folks from their side. And nothing gets done.

I am embarrassed about the way I have acted on these calls. I have called folks out and out liars. I have screamed at and bashed conference phones.

Maybe it would have been better if I hadn't known the Netezza folks - they were good.

And Softlayer...

I used to be able to call a guy down in texas and after a 20 minute phone call have a cluster of servers ordered. Once did a hadoop cluster this way. Go off and have lunch come back and the servers would we ready by the afternoon.

And now: 2 major outages in the past 2 months. No communication - in both cases entirely their fault. Power failure and network misconfig causing an arp storm. Ignored for hours while we submitted tickets and called support... Nightmare.

And an absurd situation where their security dept threatened us with taking an haproxy server offline due to a clean-mx false positive - even after the tireless guy running clean-mx emailed to that effect...

It became apparent in the discussions following this event that the TAM and sales support which has had our account for years, knew nothing about our business.

Just horrible bureacracy and bad service.

So I have had really negative experiences with 2 IBM acquired companies. Hopefully it will be better for StrongLoop.

For anyone affected - watch for the good folks shedding off.

[+] beginpanic|10 years ago|reply
I hear you on the support side. I used to work for a company that used a product that was bought by IBM. Before that, you could pick up the phone and talk to the lead developer. After, you opened a support ticket online and waited 24-48 hours for a response from someone in Costa Rica or India. Now I work at IBM with this same product, and even I as an employee who supports the product, even I have to go through the same Level 1 tech support crap.

However, the cash injection made the core product worlds better, and it was good to begin with. Support got worse, the product got better. It's almost an even balance.

[+] benjarrell|10 years ago|reply
I completely agree and add that I always find support saying, "Oh, I only support 'X', you need to call for support on 'Y'."

It is a race to pass the buck.

[+] osullivj|10 years ago|reply
I used to work with an old IBMer who used to say "it's no accident that IBM nearly collapsed at the same time as the Soviet Union did". Lou Gerstner famously rescued IBM from oblivion by turning it into a consultancy led company, and moving the focus away from mainframes. This is just a case of IBM snaffling up the latest cool thing so they can sell consulting services.
[+] aus_|10 years ago|reply
> ...moving the focus away from mainframes.

Actually, System z (mainframes) still accounts for a quarter of IBM's revenue and about half of its profits. [1] At least, that was the case in 2012. I suspect that hasn't changed much with IBM's recent announcement of the LinuxONE. In fact, IBM contributed significant effort get Node.js (and thereby V8) ported to s390x. [2]

[1]: http://www.economist.com/blogs/schumpeter/2012/09/ibms-mainf... [2]: http://www.economist.com/blogs/schumpeter/2012/09/ibms-mainf...

[+] draw_down|10 years ago|reply
Which is fine, people who build things need an exit after all.
[+] jsprogrammer|10 years ago|reply
What is the connection to the Soviet Union? Were they a large IBM customer?
[+] johnward|10 years ago|reply
Mainframes still bring in a lot of money. I actually think of IBM as being led by finance people.
[+] funkysquid|10 years ago|reply
If you're trying to relieve fears, the first paragraph does not help.

"IBM has identified Node.js as an important part of the future of enterprise middleware and StrongLoop’s technology and expertise as pivotal to their strategy to help companies fully unlock the value of their existing IT investments and legacy data with APIs."

[+] gshx|10 years ago|reply
Seems like too much fluff - both the acquirer generating it and the acquisition itself. It would be helpful to see why/where one would use node.js when other platforms support many more workloads well than just pure I/O bound ones. In the "enterprise middleware", going native (C++) or managed (JVM/CLR) brings with it a much richer ecosystem alongwith a more robust runtime that can better support other workloads that entail for example cpu binding. It's hard to see a real need for node.js as it seems to not do anything significantly better than other platform options.
[+] oneweekwonder|10 years ago|reply
Recently I got heavily into node-red[0] after playing with a ti sensortag[1], and I'm really amazed at what it can do, and that it is open source and actively used by IBM in bluemix.

Now they acquired another heavy weight in the js world, I wonder what is their next step.

On a side note, to the audience: have you looked at node-red, what do you think of it?

[0]: http://nodered.org/ [1]: http://www.ti.com/sensortag

ps. if you run node-red locally note by default it is insecure, you need to setup the config. But really it is a must try!

[+] OldSchoolJohnny|10 years ago|reply
IBM recently bought out Softlayer hosting and support went from great to glacially slow and all but uncaring in about 28 seconds.
[+] nickpsecurity|10 years ago|reply
Damnit! That's one of the companies I found searching for a physical version of cloud hosting. Looked promising. Now they're IBM. (sighs) Going to have to delete them from my bookmarks...

Sucks that IBM has this effect. If they didn't, esp if had opposite effect, they could be the biggest and most awesome tech company in existence.

[+] rogerthis|10 years ago|reply
Once I worked in a project which IBM was also involved. And everything in IBM related to that project was slow, very slow.
[+] NDizzle|10 years ago|reply
Yep. Same here. IBM contract in one of our huge data centers. (This was for a large accounting firm.)

The process to get a new blade installed in one of our chassis took 6 months.

[+] astrodust|10 years ago|reply
It's a gigantic organization that makes hardware, software, and does consulting. Does one anecdotal story really characterize it?
[+] aikah|10 years ago|reply
Congrats , I know nothing about "modern" IBM but it seems like they are trying interesting things with Bluemix. Let's see what happens next.
[+] athenot|10 years ago|reply
This announcement fits along with Node.js 4.0's release containing a commitment to an 18-month roadmap in giving Node.js a greater maturity.

As much as I like to think I have chosen Node.js purely for its merits, having a wide community who adopts a platform brings a few perks:

- Hiring developers requires less of a gamble on their part. Elixir looks very promising and probably has a lot of advantages over Node.js thanks to it's Erlang foundation but it makes it harder to assemble a team—I hope this changes as I wish the Erlang/Elixir folks all the best.

- There is a great amount of wealth contained in the package repository NPM. Before Node.js, this was a great strength of Perl's CPAN (and TeX' CTAN before that).

- Having large organizations adopt a platform will eventually increase OSS contributions.

I could be mistaken but I don't see Node.js following the bureaucratic of Java's JSR if it continues to adopt a lean and mean approach akin to UNIX tools (do 1 thing well).

[+] cdnsteve|10 years ago|reply
If anything this is a strong validation of Node.js in the enterprise space. This is actually good news for anyone working within the enterprise trying to use a different technology set. It makes selling Node to a CTO a bit easier (even if you choose not to use StrongLoop).
[+] rodgerd|10 years ago|reply
"Enterprise" is already using node.js This is IBM tryng to stay in touch with their customers.
[+] mkozlows|10 years ago|reply
I realize this blog post is full of optimistic "nothing will change!" statements, but that doesn't seem super-IBM-like.
[+] kordless|10 years ago|reply
Congratulations Issac! Happy for you and the team.
[+] ijroth|10 years ago|reply
Woot thanks Kord. I'm excited for Node.
[+] Aldo_MX|10 years ago|reply
If IBM screws ExpressJS up (wishfully not), Hapi[1] is a good alternative.

https://github.com/hapijs/hapi

[+] pluma|10 years ago|reply
Considering the development history of Express mostly consisted of cutting down on features and streamlining the core, it should be easy to notice whether the acquisition has any effect.
[+] codingdave|10 years ago|reply
It is hard to say what this means at this stage...

IBM develops products that it acquires based on where IBM needs the product to go. Sometimes that matches where the user community wants it to go, sometimes it does not. But they will put resources behind it, and it will change - we just need to give it time to see what direction that change takes things.

[+] ramigb|10 years ago|reply
Just a question -might be dumb- is the merger of io.js and node.js which happened a couple of months ago related to this news in anyway? Anyways IMHO this good news and I hope it will make node.js much better, who knows maybe replace V8 with something even better.
[+] cag_ii|10 years ago|reply
I can't imagine it wouldn't be. This kind of decisions aren't made very quickly. I'm more curious if the IO team was aware of this...
[+] amelius|10 years ago|reply
What node.js needs is a better way to manage packages. Right now, upgrading can fail with no way to roll back. Also, reproducibility is missing. E.g, it is not easy to download sources, and deploy those exact same sources on different machines.
[+] ijroth|10 years ago|reply
Please check out slc build and slc deploy from StrongLoop which attempt to solve exactly this issue and are being used by major deployments.
[+] tg3|10 years ago|reply
npm is now its own company, so I think there is hope that serious work is being done to fix those issues.