Slashdot Mirror
Ask Robert Merkel About GnuCash Development
GnuCash is probably the best-known Linux accounting program for home and small business users. GnuCash development is now sponsored by The Linux Developers Group, Inc., a company formed this May by the core GnuCash developers, including Robert Merkel, who originally got involved with GnuCash because, he says, "I was sick of my father nagging me about the dud accounting program he used." Please feel free to ask Robert about GnuCash, guile, shared libraries or almost anything else, even cricket. (He's Australian and a major cricket fan.) We'll send Robert 10 of the highest-moderated questions, and post his answers as soon as we get them back.
I use Quicken, and have never tried out GnuCash (hope to do so soon). My complaint about Quicken is that it is written at a very basic level, its logical model based around the checkbook and the bank statement. I understand a bit of double entry bookeeping, and have a hard time telling Quicken to handle, say, liabilities properly. I'd really love a personal finance product that allows one to use a debit-credit-ledger model in addition to the checkbook-statement model (so you can handle payables and receivables properly, etc.). I'd like the flexibility to set up arbitrary accounts as proper ledgers. Does GnuCash do this? If not, is it planned? Sorry about the AC post. Will find my /. password shortly...
Wouldn't you consider 'GNUCash' to be a contradiction?
Cheers
Congradulations on starting Linux Developers Group, Inc. Despite the nay-sayers, I've been increasingly interested in starting my own company around a product idea, and of course I want to GPL it. My question is, did you create a business plan and sales forcasts that so many books say to, or do you think that it doesn't really apply to new-style GPL based business since everything comes down to a user's choice? What did you identify as the main pitfalls? Have you identified marketing opportunities? Thx.
uhh... since "default=gnucash" you need to have an image with "label=gnucash". You just need to change "old_linux" to "gnucash", rerun /sbin/lilo, and it will start.
:-)
How far it will get is another story.
---
LWN had a feature on tax software on Linux a little while ago. There are a few projects out there.
---
I saw in the gnucash source tree that there have been some attempts at OFX (XML-based Online Financial eXchange format), are they still alive and is anyone active on them? Is there any work being done on online banking and how important are they to the gnucash developers? What/When should we expect to see something come from these efforts and how far along are you?
This has actually already been done. You can print to either Quicken/Quickbook style checks or Microsoft money style checks (or you can create your own custom format).
Gnucash 1.6.0 really is getting to be a pretty full featured program. It's not much like the old X-Accountant program that you probably tried way back when.
UGH! I withdraw my question. That's what I get for just skimming the damn headlines.
-l
Help cure AIDS, cancer, and more. Donate your unused computer time to worldcommunitygrid.org. Join Team Slashdot!
Seeing as how Microsoft has purchased Great Plains and how Peachtree is an awful hack, our small business has a great interest in finding an affordable suite under a libre OS. We have looked into a few options, but honestly, I think GnuCash for small businesses is worth waiting for.
So, here are my questions:
. Do you guys have any timeframe on the small business release?
. Have you considered creating a startup to fund development of the small business suite? There is a *lot* of potential here for selling services unlike Nautilus... tax updates, training, dead-tree manuals, etc.
Thanks for all the good code!
-l
Help cure AIDS, cancer, and more. Donate your unused computer time to worldcommunitygrid.org. Join Team Slashdot!
No, that's not correct.
As an experiment, try this:
volatile float point_one, point_seven;
point_one = 0.1;
point_seven = point_one + point_one + point_one + point_one + point_one + point_one + point_one;
printf ("point_seven %s 0.7\n", point_seven == 0.7 ? "equal to" : "not equal to");
The program will print "not equal to", because of accumulated rounding errors in the LSBs of the float. This kind of miniscule inaccuracy is unavoidable when converting from base 10 to base 2, and it's absolutely unacceptable in heavyweight financial software. So the previous poster's question is quite valid.
What you really want for financial apps is string-based (bignum) or BCD (binary-coded decimal) math routines, where addition and subtraction always yield precise results. Intuit/Quicken has such math routines, refined and bullet-proofed over many years, which is one reason they have such a good reputation among financial institutions.
For personal finance, the requirements are less stringent. "Insignificant" rounding errors (are supposed to) get chopped before the user ever sees them. But bignum/BCD arithmetic prevents them from ever appearing in the first place, completely eliminating them as a possible source of error.
Actually, I may as well make that my interview question: Since the GnuCash engine is supposed to be general-purpose, what steps have been taken to assure potential big-ticket users that the unavoidable rounding errors present in binary floating point math will never have an unexpected effect on GnuCash's results?
Schwab
Editor, A1-AAA AmeriCaptions
One thing it doesn't have,though, is the ability to print your checks directly onto blank check paper - you have to have a pre-printed (but un-filled-out) check.
I have a GPL'd program at http://lager.dyndns.org/freecheck/index.html that does this first part (make a blank check from REALLY blank check paper) - I'd love to see it integrated, but it's a perl/postscript mix, and GnuCash uses gnome-print, I believe...
If it still fails, it would be good to submit some bug reports.. I know its a pain, but since it's a free package think of it as cost to using it. Or if you know a bit of programming do some constructive bug fixes.
I wish all open source packages the best, but I know a lot are still in development so I don't get as annoyed when it doesn't work, unlike when I pay a $100's of dollars for commercial software. I know this is not a good excuse, but I am sure with time GnuCash will be an excellent program. Such as DDD is excellent free debugger, GnuCash is will one day be an excellent account program (*we live in hope*).
I'm likely not alone in my need to be able to sync my finance software with my Visor. Are there any plans to link GnuCash to a Palm-based money manager program, a la Quicken and PocketQuicken? Or are there any plans to develop a new, GPLed "PocketGnuCash"?
1.6 is based on an integral type (gnc_numeric). This is one of the big changes since 1.4. I'm not sure if a round-off error was ever demonstrated to have occurred with the old code, though.
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
I know there has been expressed interest on the part of the GnuCash developers to have GnuCash be right there in Gnome Office with the other applications. What specific plans do you have along those lines? Any plans for integration with Gnumeric, e.g. to allow quick importing of financial data into a spreadsheet or vice-versa? Or being able to pull in GnuCash reports into an AbiWord or Open Office document?
Celebrate the finer things in life
Why would I ask about the sound the total number of GnuCash users make?
BilldaCat
Currently there are a number of XML billing initiatives such as XMLPay, VisaXML, etc... How do you see these trends influencing the design and development of GNUcash ... in particular do you think that your users are only looking for a simple view/format/verify client (think IMAP server + XML extensions) while GNUcash moves towards being the agent of record (aka a specialised ASP?) Of you do see GNUcash as a drop-in replacement for existing monolithic accounting packages? In summary, given that the market for electronic businesses and purchases (greater use of international currency, direct attachment to electronic settlement, near real-time risk management) is going to change, how do you think GNUcash should respond?
LL
A while ago I was hesitant to use GnuCash because I discovered that it uses floating-point numbers instead of integers to store currency values. What have you done to work around the round-off errors inherent in such a system, and do you plan to migrate to an integer based system? If so, when?
I assume you're asking this because of that little blurb the other day about it requiring around 60 different libs. That's actually misleading as that's the count of all the .so files it needs, and they are all available in about only 5 packages total. Sadly, the journalist that wrote that story was not very forthcoming about this fact, I assume he was mad because he was having trouble making it work.
Need Free Juniper/NetScreen Support? JuniperForum
Again, do all of those libraries really do something? How did it get to require so many packages? Did you start out with a Gnome system with a few other odds and ends, and just picked up from there, or was there a concious decision?
Jesus was all right but his disciples were thick and ordinary. -John Lennon
What is the future of multi-user gnucash? Is there a working group or something similar? How much of a priority is this?
It would seem that a corporate offering would best be accomplished with this feature, and ensure some cash flow for your company.
On a similar vein, are the multiple dependencies meant to create a decision where there will be healthy installation contracts for you? (It's a joke, BTW)
Jesus was all right but his disciples were thick and ordinary. -John Lennon
GPLed programs come with - I quote - "ABSOLUTELY NO WARRANTY." Check out sections 11 and 12 of the the file COPYING in the GNUCash source distribution.
--- Hot Shot City is particularly good.
Is there any possiblity or chance that custom reporting mechnisms can be added into GNUCash? (I.E. A number of rules can be set up to group certain transactions into groups and generate different reports based on other sets of rules.) It is the one thing in the 1.4 release that I thought was needed. The reports that were there seemed limited and I resorted to copying data into gnumeric to generate the reports I wanted. Does 1.6 correct this, or am I still limited to a number of default reports? Is there currently work on this area, or if not where in the code can I best look to start to add this capablity.
The specification can be downloaded here, bypassing the annoying requirement to provide information for their database before downloading.
I want to use GNUCash to replace Checkfree's old Windows software - that software is the only thing keeping me using a Windows box - in fact, this is very nearly the _only_ thing I boot my Windows box for (maybe the rarely occasional game).
I have the old Checkfree software (the have since switched to MSMoney as being the default application) for bill payment for several years now. It has everything I need - simple reports, a check register, and bill payment. I have tried to think of a different system to use, but all such systems come down to a double entry style affair (since most involve some web-based bill payment system), which can be fraught with errors during data entry.
I want a system where I can enter all my transactions, electronic bill payments, checks, deposits - and then at the end of the month reconcile my bank statement/balance - like I do with the Checkfree software. GNUCash comes close (it does most everything, except the electronic bill payment)...
Worldcom - Generation Duh!
Reason is the Path to God - Anon
Then I roll around in piles of it, naked, while laughing maniacially.
Best Slashdot Co
No that there will be corporate backing, will there ever be bank integration like we see from quickbooks? Will I be able to tell gnucash to login to my bank and allow me to make transactions? How about the ability to tie into online investment?
"If you love someone, set them free. If they come home, set them on fire." - George Carlin
Money Laundering through Casinos
How to Launder Money in the Futures Market
Money Laundering 101
Want Root?
Will the next major release feature some decent documentation? Extensive documentation is the only thing I miss from GNUCash. Having only recently come to live in the US, I have enough trouble understanding the local economics, so it would be good to have a nice tutorial for my personal finance management program as a starting point.
I've rerun lilo, but now my computer won't boot. Is this a known bug, or should I post it to your World Wide Web site?
Thank you
XML causes global warming.
You will at least need a skin that makes it look like Excel. (Just like MS did to WP, etc.) Are there any plans for this?
Currently I use Wine to run Quicken Deluxe on my RedHat box, and it almost works. I've never been able to get the printer to do anything from Wine, and although I've read floppies to generate backups, I can't get Quicken/Wine to write them.
I have looked at GnuCash several times in the past, and it has yet to impress me enough to change due to several features:
- Reporting facilities: perhaps it's changed in more recent releases, but I could not find any way to customize or organize reports. I prepare reports for session meetings, and it would be nice to be able to customize the reports.
- User Interface: GnuCash seems to treat each budget item as a seperate account from which transactions are made. While this is probably the "correct" way to do it, I find it extremely difficult to handle the mountian of split checks I write each month. I have come to enjoy Quicken's much simpler handling of split checks.
- Data Interoperability Although GnuCash can read QIF files, it can't write them. I can imagine that one day I'll hand over the checkbook to somebody else; I would be remiss in locking them into a particular checkbook manager. By giving GnuCash the ability to interoperate with other checkbook managers, it makes the data more available to everyone. Let's not forget: the data is always more important than the application that it was created/edited in.
I do understand and appreciate the effort that goes in to writing a full-strength application like GnuCash. I just have philosophical differences with the choices that the authors have made.The opposite development paradigm (championed by klunky kitchen-sink packages like Microsoft Office) tends to be less reliable, slower, more expensive, more difficult to maintain.... you know the rest.
Applications like GnuCash seem to follow the monolithic paradigm. Would it be easier (or otherwise benificial) to break GnuCash up into a number of smaller, simpler applications? For instance, a database and a set of GUIs, with data analysis tools implemented as console programs. That way, with a little knowledge of shell scripting, you could put together a tool to generate virtually any report you wanted.
Many of the previous questions asked questions about the possibility of adding certain features to GnuCash (PalmPilot integration, online banking, XML interfaces for finacial tools...). If GnuCash were more granular, it would be very easy to make the system work with Other applications - easy enough that most Linux users could probably hack together what they needed to make it work.
The problem, of course, is that more granular architechures are often harder to use and configure for new users. UNIX mail, X, and network configuration were all things I agonized over as a new user, mainly because they are hybrids cludged together from very different bits of software.
My question is this - can you have the best of both? Can you make a usable application as large and complex as GnuCash, but still preserve the UNIX "one-function-per-program" design that makes it so extensible?
--
In spite of the suggestions and all the tests that I have made, I have not cavato a spider from the hole.
I heard that GnuCash is the Accounting program of choice for the Mafia and Illuminati (the whole Trilateral Commission, I believe). My question is: Do you plan on adding features for efficent money laundering to help your best customers?
Good quote, too many chars. Seriously, the slashdot 120 char limit sucks!
Cool, I haven't checked out GNUcash lately, I'll definately give it another look!
Conformity is the jailer of freedom and enemy of growth. -JFK
Is there any plans to integrate check printing support on standardized forms?
Conformity is the jailer of freedom and enemy of growth. -JFK
In this litigious society, have you spent much time thinking about possible legal issues around GnuCash? A lot of people get weird when money's the issue, and I can imagine someone trying to sue because:
- A bug in GnuCash caused an incorrect balance and thus overdrafts on the account.
- A bug caused private information (such as a bank balance) to be exposed to people it shouldn't be.
- A bug in online banking caused money to be lost, mistransfered, etc.
Obviously, bugs are possible in all software, but it seems like the intended audience for GnuCash is slightly less technical and perhaps thus less tolerant of bugs.
These same problems are faced by the commercial venders too, but they have armies of lawyers to help them out. I don't know how an open-source project would handle a lawsuit against it, and I'm wondering what you guys have thought about on this issue.
Moreso than simple home accounting software, I'm interested in electronic tax filing. Any plans to integrate GnuCash with any existing filing services, or to integrate tax features into the program (if there are any, I apologize for overlooking them)? As tax preparation and filing are services many of us are comfortable paying a fee for (especially online), I don't wonder if there isn't a revenue stream in there somewhere, as well.
I do not have a signature
do you know of any plans for an os x port that uses quartz/aqua?
the animal doesnt even have opposable thumbs, focker!
Though my question is about GnuCash specifically, I guess it does address the larger issue about how Open Source projects compete for market share and mind share with the products of well-capitalized corporations that can form strategic partnerships that (despite the shortcomings of Closed Source software generally) offer real value to consumers. Any thoughts?
Thanks for taking the time to respond,
- Brad Heintz
--
http://www.bradheintz.com/
- updated
Online banking is a great feature that I would love to see incorporated into GnuCash. I realize that there are probably a ton of obsticles to making this a reality, so here is my question:
Would it be possible to incorporate some sort of perl plugin to GnuCash that would go out to a bank's web site, access the banking account information on the site, access the relevant data and then format it in a way that could be pasted into the existing GnuCash data?
My bank, Region Bank, allows customers to access their checking account via a web browser. I hate to see this resource going to waste.
G'day Robert!!
(aka Robertspionage**)
I would like to say that the name GnuCash sounds really cool for a finance program! It sounds so much better than OldCash!
Unfortunately I have no money, therefore I have no real questions to ask you and any questions I did ask would make me look stupid.
However, I feel that I need a question to avoid being moderated offtopic so - "Does GnuCash have turbo and come with a subwoofer?"
Okay there, now onto the important stuff, being a cricket type guru/tragic yourself, I thought I would pass on my helpful tips learnt over the years when explaining cricket to Americans:
I call it the "5 Easy Steps to Learning Cricket and Changing your Life". The movie is due out later this year.
1) Use fuzzy abstract concepts like "ball" and "bat". Explain that two people from one team have the "bat" at any time and there are 11 people on the field from the *other* team who are allowed have the "ball". These people with the "ball" want to eliminate the people with the "bat".
2) In normal play, the "ball" is thrown at the "bat", similar to this game you may have heard of called baseball - except in cricket you are allowed to hit the person - in fact it is encourage and looked upon as *good tactics* ;-).
The people with the "bat" try to hit the ball to the edge of the ground (a 4) or out of the ground (a 6).
3) Then explain the nitty gritty, that "silly mid on" is actually a fielding position etc, the importance of the box ;-)
4) Emphasise the fact that England are the worst cricket team on the planet, just to dramatise it. Emphasise the fact that New Zealand are the second worst team on the planet, just to add humour.
5) Finally, explain the joys of a cricket game that lasts 5 days and still gets no result, the joys of no more than 30 seconds for ad breaks while a cricket game is on, the joys of waking up a 2am to watch international cricket cause no other country in the world is in your damn time zone, the joys of spending a whole day watching cricket and feeling like your doing something, the joys of trying to drink one beer for each 4 or 6 hit in a limited overs match......
Well, gotta cruise, good luck with that cash thingy I've been hearing so many good things about!
Cheers,
Benspionage
** Aussies have found a glitch in the english language where you can add "spionage" to any name in the world and it still makes sense, try it at home kids!! Bobspionage, Davespionage, Brittany Spearspionage)
Do you know of any accounting firms who support this system? I used to consult for a few companies, and one of the most frequent complaints I'd hear around the office was that their accountants were forcing them to use Quickbooks/Great Plains/Peachtree/whatever and they didn't like it/it didn't have flexibility to work with their business model/etc. From the quick glance I've given GnuCash, it seems to be a dream to setup for most small to mid-size businesses, but if their accountants are stuck in the closed-minded "use my software or I won't help you" mentality, this would keep a substantial user-base down.
What are the five packages GnuCash needs?
The comparison you want is "to_digits(point_seven,2)==to_digits(0.7,2)", where "to_digits" is defined as float to_digits(float x,int digits) { return float(pow(10.0,-digits) * rint(x * pow(10.0,digits))); } . If you run that, you will find that it gives you the expected answer.
BCD is actually less accurate than using floating point numbers with equivalent numbers of bits. People use it because it's easy to define and understand.
What differentiates The Linux Developers Group from Eazel? Eazel's only product of value was given away for free, as is GnuCash, which from the LDG's website seems to be its only product. I absolutely applaud your efforts for Free Software, but what is your plan to be profitable?
-Brendan
-- Azaroth