Ask HN: 'Enterprise' hater | Hacker News Reader

Ask HN: 'Enterprise' hater

15 pointsalifaziz21 comments

Define and share what is 'enterprise application' in your words.

Rules: Simple. No marketing terms.

I like Steve Jobs' definition: With enterprise software, the people who are making the purchase decision / order (whether it's in-house or externally developed) are typically not the people who're going to use the software. There are exceptions, of course.

I like this definition. It is short and zesty and feels true. However I don't think it is a sufficient condition for something to be enterprise software. Some tools for software developers or other highly skilled employees are too expensive to be bought by the employees themselves, but I still don't consider it enterprise software.

I think for it to be real enterprise software it must also be dictated from the powers that be that the minions should use it, wether they like it or not.

- Costs 5 or 6 figures

- Has often unnecessary hardware and software requirements to reduce sticker shock. For example, requires Solaris on SPARC and the database must be Oracle or MS SQL Server when it's just a glorified CRUD web app.

- Is buzzword-compliant (extra points for more esoteric buzzwords (like JSR-170 compliant)).

- Is often written in Java to allow your typical IT drone to write the necessary glue to integrate it with other enterprise software.

- Is excessively complex at all levels (UI/UX, configuration, extension).

- Is poorly documented due in part to the excessive complexity and more in part to the fat support contracts the vendors rely on.

- Makes your IT boss happy, because it continually uses up their yearly budget allotment, generates a lifetime of status meetings, and gives more of their subordinates something to do.

- Makes you, the IT worker, miserable because you have to continually fight it in order to keep it working.

- Makes you, the office worker, miserable because you have to continually fight it just to get anything done.

I worked 5y at an Enterprise Search vendor. For us there, Enterprise application was:

* to be used "within" a large org or corporations (vs web facing apps)

* which deployment/usage is through the overall organization (not just departmental (=specialized app) this is why often the 2 types of license sold were "corporate or Enterprise license" with 1 important parameter: the number of employees (per seats) of the corp

* Most of the time, installed within the firewall on the corps own servers (vs cloud)

* if modern architecture the app to be deployed globally should have "connectors"/connectivity so that it can communicate with other pre existing systems

* very often deployed by system integrators or at 3 to 4 x the price by the engineers of the Enterprise App vendor

An internal piece of software built by so many people that no one can agree on a feature set. Usually way over engineered using the latest technology buzz to enhance developers resumes.

An application is a software product which fulfills a useful purpose (performing a task, storing data, etc.)

An enterprise application is an application which is suitable for use in a wide variety of environments seen in companies. (for example, as a point of difference, an enterprise application might support a wider variety of platforms or data, or at greater scale than a non-enterprise application)

Thanks d4rt.

What about collection of different applications which is inter-dependent with each other. Each application must be communicable and extensible with less pain. That's what in my mind for enterprise right now..

Enterprise extensibility and inter-operativity or inter-dependency is extremely complex to get right.

Multiple standards exist for just about anything, and most implementations of something are broken in one way or another. I've seen obvious and subtle breakage in things that ought to be simple: * timestamps (not unambigious, not including timezone data) * mac addresses (within SNMP, in a variety of formats in different places on the same kit) * integration with LDAP (group membership only looking at primary membership) * etc.

We live in hope. Postel's Law applies.

"One size fits no-one" software

Sold by people who promise everything and deliver very little after a lot of high priced consultants customise it to what they think your requirements should be. Costs a fortune to buy and several more in 'maintenance' fees. New, incompatible version, is released a month after you go live.

* Has known bugs that will not be resolved until the newer version is purchased.

* Is developed and delivered in big-bang style rather than incrementally.

* End users and support staff are not involved in purchase or evaluation.

* Primarily sold by offering decision makers (people with budget) tickets to the British Open, Premier League football, great meals, etc.

A software system intended to store and manage a large portion of the data created and used by the members of an organization.

Most of the comments I see here so far are funny, pejorative, and spot-on accurate but the original post asked for a definition.

- a software system: an enterprise application is like a library (as in books) except that instead of books it stores smaller pieces of information - employee names, timecards, purchase orders, etc.

- intended to: because they don't always actually do that

- store and manage: not just store but make provisions for users to add to and look up the data stored there.

- a large portion: because a single spreadsheet shouldn't count. Postings on TheDailyWTF notwithstanding.

- by the members: enterprise apps may involve public-facing information services but if that's all they do then they don't count as enterprise

- of a an organization: not just businesses but schools, governments, and any other groups with a defined membership and purpose.

An environment that combines the most intractable features of the La Brea tar pits and an industrial-scale cash furnace, with attendant performance and maintenance issues, dependencies on custom kudzu code, and all of which is in aggregate entirely critical to on-going business operations.

If you can afford to rip it all out and replace it - without seriously risking blowing up your business, your entire revenue stream, or risking having to issue a mea culpa to a legal entity, or having to jack-hammer congealed product out of your mongo-million dollar manufacturing line - then it's not Enterprise code.

And yes, stuff that might not be traditionally thought of as an enterprise installation can fit this definition. The large-scale deployments of Microsoft Windows and Microsoft Exchange and SitePoint can fit this definition, as can your own large-scale deployment of your own customized software platform, as can anything to do with financial verifications or controlled substances, too.

A large application trying to do many things at once. Has a very long sales cycle and is expensive. It requires sales and marketing effort teams to get sales. Support services offered because it is a pain to install and/or run.

Total cost > $500K

A site license is > 500 seats

Long, intensive sales cycles

Heavy support and uptime requirements

And I'm not against it or anything. It can be more work than small or med businesses but it's also significantly more money. In fact, the amount of money is probably inordinately more. So you can actually become more financially successful doing enterprise software.

This is the version of the software with one or two additional features, and one or two additional zeros on the license cost.

I am building one right now ~ its a system to track piece pay for factory employees...

it takes all parts, processes and prices for those processes and generates reports.

Its enterprise to me because it could be valuable to many other companies.

My friend Kragen wrote a good post about enterprise software:

http://www.mail-archive.com/kragen-tol@canonical.org/msg0010...

Software that aims to reduce the costs for a business and does not aim to increase its revenue.

an application that is large, redundant, complex, and build to do lots of different things; costs a lot of money/resources; and is subsequently configured and customized down to fit the needs of the customer.

I'll add that selection is often the result of a top-down decision, with an attendant disconnect between the needs and interests at the top as compared to those in the trenches. The greater the separation of those mindsets, the worse the result, speaking generally.

Reply on news.ycombinator.com