>For a start, one of the many serious flaws I have previously observed with OpenOffice Writer is its complete inability to generate more than a single table of contents with the page numbers updating properly. If you read the post carefully you'll noticed I said "multiple indexes" not multiple tocs. And yes, it does work, every single time. I have an ODM that links in 40 writer files and correctly generates a TOC and three indexes using separate concordances. It works, you can call it bull... or anything else you like.
If it really is unable to generate multiple TOCs (specicially) I'd call that a minor bug since multiple TOCs would be pretty weird.
Heck, most of the things I (and others) have pointed out are logged as confirmed bugs in the OO.o tracker. The PDF/OpenType fiasco I mentioned gets several highly voted bugs just on that one topic alone. I didn't mention the PDF output of OpenType fonts. My documents generate to PDF fine, but I'm only using the "standard" fonts provided by the distro & OOo.
Even if that is so, your anecdotal data point doesn't counter the mass of evidence that OO is full of serious bugs, starting with their own bug database. I'm a developer on a couple of largish code bases (not OOo); and I know very well that a bugzilla needs to be read with a great deal of context. Only an active OOo developer is qualified to interpret the contents of the OOo bugzilla.
The "good enough" argument is a fair point, but for these specific examples, I respectfully disagree that they are even "good enough". Sure, if you're literally only writing a trivially formatted letter or resizing an image, they can do it, but of course, so can much simpler programs. The big problems come when you want to do things a little bit more advanced
Like maintain multiple hundreds-of-pages documents with chapters, table-of-contents, multiple indexes, etc... Like that? Because I do this with Open Office and it works very well.
unfortunate reality is that OpenOffice has always been horribly bug-ridden, often in quite fundamental ways, and worse, Well, I'm one satisfied user who pushes it pretty hard every day.
Spreadsheets that can't sort data when the cells contain simple calculations are pretty broken, too. I sort spreadsheets by formula results multiple times every day.
In general though, the command line is very rarely used on Ubuntu, which is a good thing; Or on openSUSE, or any of the mainstream distributions.
if you tell a normal Windows user they'd have to use the DOS prompt to accomplish something, their eyes would glaze over. But you do need to go to the command line on Windows to do things. "ipconfig/release" and "ipconfig/renew" being the best and most common example, but there are others. Being in charge of 200+ Win32 desktops and laptops I here the help desk steer Win32 people to the command line fairly frequently, often because the GUI is too difficult or just can't be trusted.
Neither distro you mention, IMHO, is targeting home users in the way that Ubuntu is. You don't see friendly smiling people holding hands, one or two clicks to download You've got to be kidding?
Opensuse.org: Nice front page, three options - I clicked download - then I look at a complex table and it fails the WifeTest(TM) dismally. And why bother. Truth is NOBODY'S WIFE IS DOWNLOADING AND INSTALLING LINUX! If Ubuntu wants to chase some fantasy market, then fine, let them. LINUX is installed by either professionals who know what it is or by geeks. A download page that provides sensible options isn't going to intimidate either of these groups.
The WifeTest is silly. They don't go to Best Buy and but a copy of XP / Vista, and they don't download LINUX distributions.
Ubuntu.com: she guessed the right domain, clicked download after looking at the screen for a few minutes, then figured "I must have a standard computer" What she didn't complain about the lack of "laptop" as an option here?
and started downloading. WifeTest(TM) said she would have bought or requested free CD's except she knew I could burn an ISO for her. Then Ubuntu failed the WifeTest as well, because without you she wouldn't have been able to download and install it.
... with windows so long because of the inaccessibility of installing and putting together a distro the way you wanted.... Of course the "desktop user" has no interest at all in "putting together a distro".
I provided LINUX desktops to user in various places starting years ago... they worked and people were productive.
When did Word Perfect first become available for LINUX? Mid 90s...? I don't recall. But LINUX has been serviceable as a desktop since then.
I'd like to coin the term "Ready for my mom's desktop." Meaning after a few hours training she can use the platform without too much hassle. I don't think this is true at all. It is the generic desktop that LINUX is currently most suited to; as vertical apps are generally not available.
The UIs are extremely poorly designed on Linux and worse still they're often inconsistent with half a dozen ways to do the same operation. Are you using KDE? Because GNOME has a very detail HIG that is ruthlessly enforced - enough to spark the occasional war on the mailing lists. GNOME is a very clean and consistent interface. Via the control panel an end-user can adjust anything they need with items organized in a very orderly fashion.
And don't even get me started on the continued use of the terminal for/any/ normal user operations. It isn't required for any normal operation.
You could potentially write something for Mono and port it over to.net, but to have that intermediate step would probably slow you down. There is *no* intermediate step. Mono implements the CLR; if I create an app in Monodevelop I can drag the exe/dll over to a Windows machine with.NET... and it runs. There is no porting required.
What gives? Why are 860 people out of 1,000 reporting the desire to leave their current employers? Despite popular belief here there really are a great number of employers who respect/appreciate their employees and reward them accordingly. Generally, if you contribute you will be rewarded.
I anticipate a salary increase this year, like almost every previous year. Like everyone else in my department. And like most of the developers and sys-admins I know.
This is why, unless your application is so complex you can sell support services, GNU apps will almost always be immature and buggy. This certainly seems to born out in reality often enough. Most of the project sites are filled to overflowing with abandoned, broken, and incomplete projects. Part of this is the resource problem and, I believe, part of this is the reinvent-the-wheel syndrom that Open Source seems to suffer from - rather than finding a decent application and building on it everyone starts their own, realizes the problem is more complex than they thought, and either moves on or tries to bill their project/app as simple-and-lean (a euphamism for incomplete).
To make apps that are solid/stable and feature-ful requires time. Yep.
Maybe when you are young and motivated from the idealism found in universities, one can afford to spend the time it takes to contribute. But after, life intrudes. The only other alternative is to make an app that competes with MS This is a bit boolean. I'm in my thirties and still have time to work on my projects (and I even get paid a very little bit from time-to-time). Modern life has lots of leasure time, hacking can be a nice hobby. And most of my projects relate to scratching my own itch and I use them at my paying job - many hackers I know have the same configuration. However, I really don't expect anyone to help me out, I've been doing this too long. I realize that developers, at least good ones, are scarce.
But if anyone wants to try their hand at a.NET / Gtk# application take a look at Consonance. There are lots of interesting avenues to explore, you can do db4o, threading, Quartz, Cairo, etc... backend / frontend, whatever!
> Quite often the userbase is not aware of the > programmer shortage.
Agree, there is nothing wrong with being VERY up front with the projects needs.
> If you have a user list then quite often a plea > for programmers/testers will achieve results. I > have done this a few times for my major project > and it has always worked.
I'm less optimistic than always, it depends a great deal upon the type of project, but it certainly it acceptable for a Open project to occasionally ping the user base for potential developers.
> I also disagree wiht parent that you should > have posted the url on slashdot.
Agree, a project needing developers isn't really news anyway. If Slashdot started carring every project's call for developers....
> chances of finding interested developers is low.
Make sure you are listed on the main index sites, especially freshmeat, will certainly help people find your project. It may or may not help you find developers (probably not, from experience) but the more users you find the more likely you are to pick up a developer.
I find Google's project hosting to be sufficient for my smallish projects; and a couple of larger ones I work with have switched too. It doesn't offer all the bells and whistles but it is reliable and fast (unlike sourceforge).
On the other hand I don't think just using a hosting service will get you any developers; Open Source developers are actually pretty scarce (count how many projects on any of these sites are actually abandoned).
I've read about this. Supposedly you can also control OOo from.NET (C#). But I've seen a total of one trivial example and nothing else.
Controlling M$-Office is easy and *VERY* well documented. The only thing I've really seen for OOo concerns Java developers and is an enormous tome of API docs (API docs seem to be considered "documentation" in Java land). I'd really like to use OOo in my apps but there just isn't much information readily available.
If there is documentation, articles, and examples I'd seriously like pointers to them (substantive ones - not "Hello World" noise).
No, I think the article puts forth an incorrect premise, makes a big deal of that premise, and that the [incorrect] premise is based on old much-hashed gripes. Many of which can be leveled against any development environment (UNIX & glibc have their own share of cruft and then there is the ugly mess that is things like the openssl and/or gnutls APIs).
Posting this on Slashdot was just elevating a typical and dreary fanboy article to an inappropriate level.
His premise being that the.NET API refelects old & krufty Win32 [it doesn't], that Objective-C [with which I'm very familiar] is more advanced [despite no equivalent to generic, LINQ, or namespaces, etc..], and that he can't make pretty applications on Windows. Among the various incorrect, or at least substanceless statements, is that there is no ribbon library for.NET - which isn't true.
I read the article. His arguments about.NET are non-specific and in context rather humorous. I do about 50% of my development in Objective-C so I'm well aware of his proposed alternative. You want to talk about simplistic? The Objective-C toolkits are pretty darn simplistic..NET isn't perfect but it is a highly productive environment - and this is if we don't include generics, LINQ, and namespaces - none of which Obj-C provides (or only extremely primitive alternatives). I would suggest the developer in the article look into C# a little deeper.
And yes, there is a ribbon library for.NET; and it is Open Source! (although I personally do not like the whole ribbon movement)
No, I'm just saying this article doesn't offer anything new and is (a) primarily a rehash and (b) the arguements against.NET and for Cocoa don't offer much substance.
There are lots of legitimate critiques of various technology - this one isn't and doesn't merit being carried on Slashdot.
Should Slashdot carry every rehash published on the web that supports its political/technological agenda? How boring.
And the same thing happens in glibc - very non-optimal behavior gets carried forward for backwards compatibility. The calls used by NSS are a prime example of this. Every platforms suffers from this issue.
On the other hand, I've got lots of 12+ year old UNIX code that works just fine today.
I read the article. His arguments about.NET are non-specific and in context rather humorous. I do about 50% of my development in Objective-C so I'm well aware of his proposed alternative. You want to talk about simplistic? The Objective-C toolkits are pretty darn simplistic..NET isn't perfect but it is a highly productive environment - and this is if we don't include generics, LINQ, and namespaces - none of which Obj-C provides (or only extremely primitive alternatives). I would suggest the developer in the article look into C# a little deeper.
The argument about old krufty code in Windows and the Win32 API has been around since.... the Win16 API! It didn't really seem to slow down Win32.
On the flip side is the argument that the need for backwards compatibility is holding back Windows - yet developers complain about the migration from XP to Vista?
All smells like we-will-find-anyway-to-condemn-Windows to me. Note: I do all of my development on LINUX, so I'm not a Windows booster. I think lots about Windows just stinks but there is an issue of credibility here.
If you want a clean new coherent API and you want to develop on Windows Microsoft has provided an option:.NET
It doesn't officially have "extensible authentication methods" but an authentication method could be added easily enough (code is modular and very object-oriented).
Slashdot Mirror
If it really is unable to generate multiple TOCs (specicially) I'd call that a minor bug since multiple TOCs would be pretty weird. Heck, most of the things I (and others) have pointed out are logged as confirmed bugs in the OO.o tracker. The PDF/OpenType fiasco I mentioned gets several highly voted bugs just on that one topic alone. I didn't mention the PDF output of OpenType fonts. My documents generate to PDF fine, but I'm only using the "standard" fonts provided by the distro & OOo. Even if that is so, your anecdotal data point doesn't counter the mass of evidence that OO is full of serious bugs, starting with their own bug database. I'm a developer on a couple of largish code bases (not OOo); and I know very well that a bugzilla needs to be read with a great deal of context. Only an active OOo developer is qualified to interpret the contents of the OOo bugzilla.
The "good enough" argument is a fair point, but for these specific examples, I respectfully disagree that they are even "good enough". Sure, if you're literally only writing a trivially formatted letter or resizing an image, they can do it, but of course, so can much simpler programs. The big problems come when you want to do things a little bit more advanced
Like maintain multiple hundreds-of-pages documents with chapters, table-of-contents, multiple indexes, etc... Like that? Because I do this with Open Office and it works very well. unfortunate reality is that OpenOffice has always been horribly bug-ridden, often in quite fundamental ways, and worse, Well, I'm one satisfied user who pushes it pretty hard every day. Spreadsheets that can't sort data when the cells contain simple calculations are pretty broken, too. I sort spreadsheets by formula results multiple times every day.The WifeTest is silly. They don't go to Best Buy and but a copy of XP / Vista, and they don't download LINUX distributions. Ubuntu.com: she guessed the right domain, clicked download after looking at the screen for a few minutes, then figured "I must have a standard computer" What she didn't complain about the lack of "laptop" as an option here? and started downloading. WifeTest(TM) said she would have bought or requested free CD's except she knew I could burn an ISO for her. Then Ubuntu failed the WifeTest as well, because without you she wouldn't have been able to download and install it.
... with windows so long because of the inaccessibility of installing and putting together a distro the way you wanted.I provided LINUX desktops to user in various places starting years ago... they worked and people were productive.
When did Word Perfect first become available for LINUX? Mid 90s...? I don't recall. But LINUX has been serviceable as a desktop since then.
Yep; nothing in this list excited me at all.
I anticipate a salary increase this year, like almost every previous year. Like everyone else in my department. And like most of the developers and sys-admins I know.
But if anyone wants to try their hand at a
> Quite often the userbase is not aware of the
> programmer shortage.
Agree, there is nothing wrong with being VERY up front with the projects needs.
> If you have a user list then quite often a plea
> for programmers/testers will achieve results. I
> have done this a few times for my major project
> and it has always worked.
I'm less optimistic than always, it depends a great deal upon the type of project, but it certainly it acceptable for a Open project to occasionally ping the user base for potential developers.
> I also disagree wiht parent that you should
> have posted the url on slashdot.
Agree, a project needing developers isn't really news anyway. If Slashdot started carring every project's call for developers....
> chances of finding interested developers is low.
Yep.
> Most would have just been idle browsers.
Or worse, hordes of malcontents.
Make sure you are listed on the main index sites, especially freshmeat, will certainly help people find your project. It may or may not help you find developers (probably not, from experience) but the more users you find the more likely you are to pick up a developer.
Maybe the question was made generic so it would be of general interest? It certainly is a legitimate question for lots of projects.
I find Google's project hosting to be sufficient for my smallish projects; and a couple of larger ones I work with have switched too. It doesn't offer all the bells and whistles but it is reliable and fast (unlike sourceforge).
On the other hand I don't think just using a hosting service will get you any developers; Open Source developers are actually pretty scarce (count how many projects on any of these sites are actually abandoned).
No, you can use M$-Office from Java just as well. I think the issue here is: documentation.
I've read about this. Supposedly you can also control OOo from .NET (C#). But I've seen a total of one trivial example and nothing else.
Controlling M$-Office is easy and *VERY* well documented. The only thing I've really seen for OOo concerns Java developers and is an enormous tome of API docs (API docs seem to be considered "documentation" in Java land). I'd really like to use OOo in my apps but there just isn't much information readily available.
If there is documentation, articles, and examples I'd seriously like pointers to them (substantive ones - not "Hello World" noise).
No, I think the article puts forth an incorrect premise, makes a big deal of that premise, and that the [incorrect] premise is based on old much-hashed gripes. Many of which can be leveled against any development environment (UNIX & glibc have their own share of cruft and then there is the ugly mess that is things like the openssl and/or gnutls APIs).
.NET API refelects old & krufty Win32 [it doesn't], that Objective-C [with which I'm very familiar] is more advanced [despite no equivalent to generic, LINQ, or namespaces, etc..], and that he can't make pretty applications on Windows. Among the various incorrect, or at least substanceless statements, is that there is no ribbon library for .NET - which isn't true.
Posting this on Slashdot was just elevating a typical and dreary fanboy article to an inappropriate level.
His premise being that the
I read the article. His arguments about .NET are non-specific and in context rather humorous. I do about 50% of my development in Objective-C so I'm well aware of his proposed alternative. You want to talk about simplistic? The Objective-C toolkits are pretty darn simplistic. .NET isn't perfect but it is a highly productive environment - and this is if we don't include generics, LINQ, and namespaces - none of which Obj-C provides (or only extremely primitive alternatives). I would suggest the developer in the article look into C# a little deeper.
.NET; and it is Open Source! (although I personally do not like the whole ribbon movement)
And yes, there is a ribbon library for
No, I'm just saying this article doesn't offer anything new and is (a) primarily a rehash and (b) the arguements against .NET and for Cocoa don't offer much substance.
There are lots of legitimate critiques of various technology - this one isn't and doesn't merit being carried on Slashdot.
Should Slashdot carry every rehash published on the web that supports its political/technological agenda? How boring.
And the same thing happens in glibc - very non-optimal behavior gets carried forward for backwards compatibility. The calls used by NSS are a prime example of this. Every platforms suffers from this issue.
On the other hand, I've got lots of 12+ year old UNIX code that works just fine today.
Other than the way the slashes point in file paths there is nothing DOS in .NET but the fanboys are incapable of surrendering one their favorite cards.
And fanboys booster "Open Source" in any and all arguments for free in order to stoke their desperate need for an indentity.
I read the article. His arguments about .NET are non-specific and in context rather humorous. I do about 50% of my development in Objective-C so I'm well aware of his proposed alternative. You want to talk about simplistic? The Objective-C toolkits are pretty darn simplistic. .NET isn't perfect but it is a highly productive environment - and this is if we don't include generics, LINQ, and namespaces - none of which Obj-C provides (or only extremely primitive alternatives). I would suggest the developer in the article look into C# a little deeper.
"how one developer migrated from Windows to OS X"
.NET
That pretty much says it all: "one developer"
The argument about old krufty code in Windows and the Win32 API has been around since.... the Win16 API! It didn't really seem to slow down Win32.
On the flip side is the argument that the need for backwards compatibility is holding back Windows - yet developers complain about the migration from XP to Vista?
All smells like we-will-find-anyway-to-condemn-Windows to me. Note: I do all of my development on LINUX, so I'm not a Windows booster. I think lots about Windows just stinks but there is an issue of credibility here.
If you want a clean new coherent API and you want to develop on Windows Microsoft has provided an option:
It doesn't officially have "extensible authentication methods" but an authentication method could be added easily enough (code is modular and very object-oriented).