Rebid?
I bet IBM would like to bid.
Fujitsu has become the latest service provider to be associated with a failed SAP project in Australia, after a project it led saw the buyer underwhelmed despite going over time and over budget. The project in question was an asset management system for the government of Australia's Northern Territory. The Territory is huge – …
Having worked on projects up there myself and having no intention to defend the Ibmentureitsu style conglomerates, the "natives", i.e. Darwinites, frequently aren't a help either.
Not sure I can blame 'em. After a week something up there, the heat, the humidity, the beer, just saps the life out of you. I wondered how they can get anything done - and maybe the locals have given up wondering and just drink beer.
I like heat and sun, but not when it involved putting in major IT systems.
A thought occurred to me: how do the Canadians deal with their remote, isolated, sparsely populated areas? Does Ottawa take care of all of their IT projects?
In that case you have seen teeny weeny little projectettes.
No reasonable scale SAP project could possibly be carried out or even led by a single admin so that is a ridiculous comparison. Not that I am sticking up for SAP, which is an absurd behemoth that has a tendency to grow like a cancer. It is also hideously ugly and user-unfriendly, producing utterly incomprehensible error messages, often in German (although they aren't any more comprehensible in English).
Single admin project - Full system update including hardware refresh and SAN implementation.
Consultants - Adding a single module.
You tell me which was the projectette. You must be management to under estimate the abilities of a competent, dedicated systems admin*. Don't like my assumption you are an inbred waste of obstructive space? You shouldn't have made an assumption about the capabilities about peoples skills. Our industry is full of people who are capable of the most astounding work, unfortunately it's also full of idiots who belittle them too.
*I would like to point out I don't class myself in that league
because SAP "empowers the users" and makes them think they can "do IT". And then they get their butts handed to them and the techies have to take over.
The users are always right,... about what they THINK! they want, not about what they actually want, nor about what is best technical practice.
Done a system install there. Nothing hard about working in Darwin, _if_ you dress right and do not take yourself seriously. Which so many PHB/suit types who are drawn to SAP do not seem to. Affordable accomodation is a problem but this is becoming so across all of Oz. The weak minded do get drawn into the local alcoholic culture. I suggest the BOFH might meet his match in a few pubs there. :-) <offtopic> I would love to see monkeyboy do a presentation there in the Wet to general public And take bets on whether he would make it offstage. </offtopic>.
IMHO, the turnover in PMs is an indicator they saw a badly drafted design, bad or no measurable specs and implementation, you get the idea. Very sensible to run. Why has not IT got the equivalent of a construction quantity surveyor ? In the construction industry I saw them save taxpayers and bidders millions by investigating doubtful bids in very early contract evaluation. Bidder could and did gracefully withdraw and no harm done. If QS figures were out, an investigation was done as to why, to keep QS costs knowledge current.
Now places like Kunnunurra and Katherine in pre Wet are unlivable. Dunno how they do it, even with A/C.
Because in IT there is no such thing - it's all replaced by marketing sharks before project signing, then by promotion-searching sharks once the project is signed.
Nobody in that bunch is interested in making sure that the initial specs are "good", only that the project is "a challenge" that will result in good marks and a reference on their CV.
@ Robert Grant
I jsut Wikipedia'ed and there were 6 Paul Grahams. Did you mean the Computer Programmer? Or am I just showing myself to be a BOF? I don't know which features of SAP would be required in an equivalent: which ones are necessary? BTW Compiere & Adempiere both have modern UIs. But I suspect, that they don't have the necessary functionality.
OK, I'll chip in a bit on that thought...
My primary professional work is on PeopleSoft, but I keep myself happy by writing tons of utility scripts in Python, mostly to do with PeopleSoft automation.
Writing a simple ERP system is well... probably simple, assuming basic requirements, good functional experts, dedicated coders and a well-thought out general architecture using state of the art programming techniques. If the requirements are simple then SAP is likely overkill.
However, large multinational-company-capable ERPs have two things which make them a little bit less easy to copy successfully.
First, the customers': as they scale up in size, tend to have all sorts of rather bizarre organizational structures and requirements.
Just because you can run a 200-person company in one country with a simple ERP doesn't mean that you can run say Toyota's business in 30 countries on that same ERP. Yet, that is what multinationals want, single-point-of-truth systems that accommodate configuration of their organizational structures and in-house business practices. Public sector customers are usually the worse of the lot given that they have all sorts of extra rules, like union seniority plans and a general refusal to adapt their business processes.
Second, if the possible complexities in the business weren't enough, you then hit legislated requirements. Running a French payroll is very different from US payroll. The US will have hefty tax requirements, per state, while France will require very specific itemized lines on its payroll slips.
As another example, working out a new ERP engine to calculate VAT for all possible combinations of buyer country vs vendor country vs service/goods permutations kept a team of coders busy for a year, at one vendor. As well as one extremely knowledgeable VAT functional expert who could just rattle off the VAT rules applicable to all the countries involved in her sleep. And that VAT engine needed to be available online and in batch mode, in modules as different as AP, AR, PO, etc... I have since then rarely worked with people as clever as those involved in that project.
These legislated requirements can change frequently and most ERP vendors accommodate that by driving as much business logic as possible through configuration tables. The end user screens are only a very small part of the iceberg.
Vendors like SAP, for all its faults, will deliver products that, once certified for given country, will kinda meet big customer requirements and legislative requirements.
IF implemented correctly and in a domain for which they are applicable. (oddly enough, ERP salesfolk will always tell you their stuff is applicable to your business and easy & cheap to implement).
SAP and the like have evolved through decades of adhoc adjustments to rules that have nothing to do with programming logic. Rewriting them is possible, for sure, but it requires hefty _global_ functional knowledge and ongoing support. Not just coding virtuosity.
I'd say this is a special case of Joel Spolky's rewrites-are-not-that-easy arguments.
http://www.joelonsoftware.com/articles/fog0000000069.html
Or remember that the Extreme Programming craze got launched on the back of a failed payroll implementation @ Chrysler.
http://en.wikipedia.org/wiki/Chrysler_Comprehensive_Compensation_System
Having said all that, I really like the idea of Open source ERPs, esp. on Python.
Anyone care to share success stories on _using_ open source ERPs in companies? Especially for companies w. > 500 employees?
Surely, these have been done before? 250k people can't require too many assets, can they? And how much managing can these assets require? Geez! Any medium sized town probably has software that could be converted or used directly.
SAP seems to have done an epic fail here. If only because they didn't check their files to see if they had recently done something similar for someone else.