What do you mean that's not how open source is supposed to work? There's nothing wrong with open source working this way. The FREEDOM is what's important, not necessarily the actual contributions.
I've got a PII 200 w/64M RAM, and Mozilla works fine for me. Startup is a little slow, as well as opening new windows. However, Galeon takes care of most of that. I especially like it's tab-mode, where every Internet "window" is actually just a tab in the main window, so I don't clutter up my screen.
Re:custom distributed component model...
on
Mozilla 0.9 Out
·
· Score: 2
Well, they could have used CORBA, since it does exist everywhere. Or they could have done it the AbiWord way, and use each platform's own component model.
Re:What's holding it back speed-wise?
on
Mozilla 0.9 Out
·
· Score: 2
Part of the problem is that Mozilla chose to write a good amount of the browser in Javascript. Another part of the problem is that they basically implemented their own distributed component model from scratch, as well as their own widgets, instead of using platform libraries and widgets. I understand the decision to use custom widgets (especially for the HTML), but the custom distributed component model was quite silly. Mozilla is basically GNOME re-implemented running one application. Personally, I find running Mozilla embedded in Galeon to be much, much better. This way Galeon handles all of the non-HTML stuff through GNOME, while Mozilla just does the HTML rendering.
The interesting thing is that this is exactly what Intel does, too. The only difference is that Crusoe's translations are softcoded, while Intel's are more-or-less hardcoded. Intel chips are actually RISC at the core, with a translation layer between IA32 and whatever they've got underneath. However, I'm willing to bet Intel's RISC machine is optimized directly for IA32 while Crusoe's is probably more general.
Also, it's interesting to note that having the RISC core give the Intel chip a boost, not a setback.
Actually, data loading/backend-type stuff is easier with an RDBMS, because data-entry is almost alway tabular anyway. However, once entered, it is usually easier to process it through an OO layer. The best way to accomplish this is to have a CORBA layer that the applications always use for talking to the database that actually incorporates all the business logic, but have the CORBA layer talk to an RDBMS.
I don't have a lot of experience with OODBMSs - I'd be curious exactly how they work. The closest I've worked with is PostgreSQL which is Object-relational. Are there any intro guides, especially to schema definition and stuff like that.
I haven't gotten into 3-2-1 penguins, yet. You wouldn't happen to have any insights as to what happened to Bob on the Esther video? I couldn't find him anywhere.
Well, you said it. RedHat is not profitable, period. - What's your point? Cygnus certainly was for a long period of time. The fact that RedHat hasn't gotten there yet doesn't mean that noone else has.
IBM _does_ make money on open source. They make money selling hardware that runs open source. People are buying and adding to mainframes to run Linux. That's making money. They developed the software, and people bought the hardware to run it - that _is_ an open source business model.
Apple has released a lot back to the community, including an open-source streaming server.
I alsoaven't yet haven't heard about any of the embedded Linux people going under.
The people who didn't make money on Linux are those who were banking on Linux taking over in less than 5 to 10 years, especially those counting on it taking over the desktop, and grew their companies too fast. For example, you have Eazel, which may have had a good idea, but they hired marketers and PR people and had an office over a year before they released 1.0. In addition, their business model was based on the rapid ascension of Linux on the desktop, which hasn't happened yet. VA Linux Systems tried to be the IBM of Linux, doing just about everything, and found out that it wasn't a profitable method. They tried to grow too fast and do too much. I'm guessing Penguin Computing is doing much better because they offer most of the same services as VA, but through partner arrangements. They are focused, and they know that they serve a specialized market.
Other companies weren't making any money before Linux, and still aren't, like Corel and SGI, although SGI has the potential to do so in a year or two, if Intel ever ships Itanium.
So, money can be made in Linux, but the key is to have good, sound business practices, and to know where you stand. If you understand you serve a specialized market, you can make money by being a free software builder. If your business is based on the assumption that everybody is going to switch their desktops to Linux next year, you're hosed.
Yes, but that has nothing to do with forking. That has to do with code being available at zero cost. Given that, code forking is a non-issue for the given reason. Not given that, why even talk about code forking.
This is a common misconception. The good/badness of forking has always been debated. I am of the opinion that forks are very good. I have always hotly debated anyone who thinks they are evil. Code forks brought us egcs, OpenBSD, Samba 2.2, Apache (it's actually a fork from NCSA), PCMCIA kernel support (it got merged in in 2.4, but it existed as a "popular" fork of 2.2 for a while), Real-Time Linux, and probably a lot of other things I can't think of right now.
The _ability_ to fork also brings good things. For example, many people produce kernel "forks" which are small, but useful, until the given functionality gets rolled into the mainstream kernel. These mini-forks are really what give free software a competitive edge.
1) MS was really on the right path to begin with, and Eazel's experience & testing has independently confirmed it.
-or-
2) Eazel's UI expertise & usability testing wasn't worth a hill of beans because they ended up copying most of MS Explorer.
Actually, you missed an option. Windows, by virtue of being what everybody uses, is now the standard for usability, because that's what everyone already knows. Eazel, when they started, tried do rethink a lot of assumptions, but found that they couldn't because everyone was already used to the way MS did it. They found out that if you write a filemanager, you have to copy MS, or everyone will be confused.
So what you're saying is that you haven't _used_ the filemanager, but you can comment on where it is compared to others? If you _have_ Nautilus, right-click on a file and go to "show properties". Click on the "emblems" tab. This allows you to put any emblem you want on a file (you can configure the available emblems elsewhere). This way, you can "tag" files with such things as "draft", "important", or whatever else. I can't currently find the searching feature, but I know it's either there or planned, so you can find all of your "need to work on" files, no matter what directory they are in.
* Previewing of text files in the manager (the icons for text files include text from the file itself)
* Use of emblems to mark files and to search for files
* Iconic representation of file permissions
* New configuration mechanism using drag-n-drop
* Use of user modes - Beginning/Intermediate/Advanced
And that's just from casual use. There's probably more.
Companies drain money. If it was just paying programmers, it wouldn't have cost this much. But when you add a corporate infrastructure (net access, company databases, backups, CFOs, CIOs, SysAdmins, marketers, a building, computers for everyone, routers, switches, hubs) it all adds up. Eazel's problem was that they started the "company" part _way_ before there was a product. If the VCs had just made them be 6 to 10 hackers in a garage until Nautilus 2.0 came out, they would have had a chance. However, having the overhead of a whole company for over a year before there is anything to sell is what can easily drain $13 million in a year or two.
Who's not making money? VA - because they can't focus, and were overly optimistic. Corel - because they haven't made money in a while even before Linux. Eazel - because their business plan was pretty stupid anyway. Ximian may come out of this all right, if they play their cards well enough.
Add all of the consultants to that, and you've got a picture of whose making money.
Companies don't spend $13 million on a file browser. If you think that's all Eazel was doing, you're wrong. Now, they were definitely over-extravagant in their spending. However, let's take a look at what they might have spent it on:
$2000-$3000/month on Internet access - for one year that's $36,000
$200,000 for their infrastructure - backups, routers, gateways, plus licenses (this could actually have been more. You can really spend up to $2 million easily to make a scalable infrastructure - like if you use Oracle Apps to manage all your stuff).
Let's say they had 10 programmers (I don't know how many they had) on Nautilus - for good programmers, that's about a million per year.
Let's say they had another 10 programmers working on Eazel services, including their packaging and online disk storage, we've got another million there.
Then you have to pay the execs. I'm not going to guess at a figure. Then you've got another twenty to thirty people doing all sorts of marketing/reception/etc.
On top of this, you have office space. If they went for their own building, this could be a few million.
Then you have computers for everyone, and that can get expensive real fast.
So, as you see, $13 million can go pretty fast, especially if you're trying to start-up fast. Most of the dot-coms failed trying to start-up fast. Most companies do. Venture capital makes you think you can do anything because you have all that money, but then you end up wasting it buying the high-end of everything. The thing is that with $13 million, if the investors were willing to wait a little while, _could_ have been spread out over a decade, with the programmers all sitting in a basement, a 28.8 line to the 'net, and not bothered to even hire the marketing guys until the product was out the door and at revision #2. However, most VC places probably don't like that idea, so they try to get a full company in swing before a product is released, which, as you can see, really drains money.
So, of course a file manager doesn't cost that much money, but a company does. The problem is that they formed the company before it was ready, and thus the company drained them of their money. However, they probably wouldn't have gotten VC money doing that. The whole company infrastructure is a bigger drain than any or all projects put together.
Yes, but _someone_ has to make the RPMS. Let's face it, our mothers/grandmothers will never download the latest GNOME, no matter how easy it is. They will only get it when it's released on the RH CD.
Also, Ximian _did_ provide binaries. So what if you had to wait a month. Who really cares? Did your life improve _that_ much after getting GNOME 1.4 that the previous month seemed wasted? I doubt it.
Why wait for the RHs to provide binaries? Well, we have to wait for someone, or we'd be building from source, wouldn't we?
The Linux machine they were using was only for a single experiment, studying plant growth, I believe.
Re:Consulting for open source is less lucrative
on
Eazel On The Ropes
·
· Score: 2
And those developing code can provide something that noone else provides - experience with the code. Just having access to it means nothing. It takes time and energy to get involved and learn how it works. If I have a kernel question, would I rather ask a kid next door or Alan Cox?
The company who actually ran it is called SourceGear. They also are the developers of CVS (well, they bought Cyclic Software, who was the developer of CVS). They still provide server space, but they don't provide full-time developers anymore. I live in Champaign, and actually had a friend doing QA on AbiWord at one point in time, so I'm pretty familiar with the situation. If this has changed, it has been so recently. It used to say so on either AbiSource or SourceGear's site, but I can't find it. The only good references to this I have found are:
Slashdot Mirror
What do you mean that's not how open source is supposed to work? There's nothing wrong with open source working this way. The FREEDOM is what's important, not necessarily the actual contributions.
I've got a PII 200 w/64M RAM, and Mozilla works fine for me. Startup is a little slow, as well as opening new windows. However, Galeon takes care of most of that. I especially like it's tab-mode, where every Internet "window" is actually just a tab in the main window, so I don't clutter up my screen.
Well, they could have used CORBA, since it does exist everywhere. Or they could have done it the AbiWord way, and use each platform's own component model.
Part of the problem is that Mozilla chose to write a good amount of the browser in Javascript. Another part of the problem is that they basically implemented their own distributed component model from scratch, as well as their own widgets, instead of using platform libraries and widgets. I understand the decision to use custom widgets (especially for the HTML), but the custom distributed component model was quite silly. Mozilla is basically GNOME re-implemented running one application. Personally, I find running Mozilla embedded in Galeon to be much, much better. This way Galeon handles all of the non-HTML stuff through GNOME, while Mozilla just does the HTML rendering.
The interesting thing is that this is exactly what Intel does, too. The only difference is that Crusoe's translations are softcoded, while Intel's are more-or-less hardcoded. Intel chips are actually RISC at the core, with a translation layer between IA32 and whatever they've got underneath. However, I'm willing to bet Intel's RISC machine is optimized directly for IA32 while Crusoe's is probably more general.
Also, it's interesting to note that having the RISC core give the Intel chip a boost, not a setback.
Actually, data loading/backend-type stuff is easier with an RDBMS, because data-entry is almost alway tabular anyway. However, once entered, it is usually easier to process it through an OO layer. The best way to accomplish this is to have a CORBA layer that the applications always use for talking to the database that actually incorporates all the business logic, but have the CORBA layer talk to an RDBMS.
I don't have a lot of experience with OODBMSs - I'd be curious exactly how they work. The closest I've worked with is PostgreSQL which is Object-relational. Are there any intro guides, especially to schema definition and stuff like that.
Is there a free software OODBMS?
I haven't gotten into 3-2-1 penguins, yet. You wouldn't happen to have any insights as to what happened to Bob on the Esther video? I couldn't find him anywhere.
ADA Core Technologies - http://www.gnat.com/
Well, you said it. RedHat is not profitable, period. - What's your point? Cygnus certainly was for a long period of time. The fact that RedHat hasn't gotten there yet doesn't mean that noone else has.
IBM _does_ make money on open source. They make money selling hardware that runs open source. People are buying and adding to mainframes to run Linux. That's making money. They developed the software, and people bought the hardware to run it - that _is_ an open source business model.
Apple has released a lot back to the community, including an open-source streaming server.
I alsoaven't yet haven't heard about any of the embedded Linux people going under.
The people who didn't make money on Linux are those who were banking on Linux taking over in less than 5 to 10 years, especially those counting on it taking over the desktop, and grew their companies too fast. For example, you have Eazel, which may have had a good idea, but they hired marketers and PR people and had an office over a year before they released 1.0. In addition, their business model was based on the rapid ascension of Linux on the desktop, which hasn't happened yet. VA Linux Systems tried to be the IBM of Linux, doing just about everything, and found out that it wasn't a profitable method. They tried to grow too fast and do too much. I'm guessing Penguin Computing is doing much better because they offer most of the same services as VA, but through partner arrangements. They are focused, and they know that they serve a specialized market.
Other companies weren't making any money before Linux, and still aren't, like Corel and SGI, although SGI has the potential to do so in a year or two, if Intel ever ships Itanium.
So, money can be made in Linux, but the key is to have good, sound business practices, and to know where you stand. If you understand you serve a specialized market, you can make money by being a free software builder. If your business is based on the assumption that everybody is going to switch their desktops to Linux next year, you're hosed.
ADA Core Technologies
Cygnus (well, now they are bought out by RedHat)
I'm guessing MandrakeSoft is
Many, many consultants
Apple
IBM
Yes, but that has nothing to do with forking. That has to do with code being available at zero cost. Given that, code forking is a non-issue for the given reason. Not given that, why even talk about code forking.
This is a common misconception. The good/badness of forking has always been debated. I am of the opinion that forks are very good. I have always hotly debated anyone who thinks they are evil. Code forks brought us egcs, OpenBSD, Samba 2.2, Apache (it's actually a fork from NCSA), PCMCIA kernel support (it got merged in in 2.4, but it existed as a "popular" fork of 2.2 for a while), Real-Time Linux, and probably a lot of other things I can't think of right now.
The _ability_ to fork also brings good things. For example, many people produce kernel "forks" which are small, but useful, until the given functionality gets rolled into the mainstream kernel. These mini-forks are really what give free software a competitive edge.
-or-
2) Eazel's UI expertise & usability testing wasn't worth a hill of beans because they ended up copying most of MS Explorer.
Actually, you missed an option. Windows, by virtue of being what everybody uses, is now the standard for usability, because that's what everyone already knows. Eazel, when they started, tried do rethink a lot of assumptions, but found that they couldn't because everyone was already used to the way MS did it. They found out that if you write a filemanager, you have to copy MS, or everyone will be confused.
So what you're saying is that you haven't _used_ the filemanager, but you can comment on where it is compared to others? If you _have_ Nautilus, right-click on a file and go to "show properties". Click on the "emblems" tab. This allows you to put any emblem you want on a file (you can configure the available emblems elsewhere). This way, you can "tag" files with such things as "draft", "important", or whatever else. I can't currently find the searching feature, but I know it's either there or planned, so you can find all of your "need to work on" files, no matter what directory they are in.
Features in Nautilus not in Windows
* Previewing of text files in the manager (the icons for text files include text from the file itself)
* Use of emblems to mark files and to search for files
* Iconic representation of file permissions
* New configuration mechanism using drag-n-drop
* Use of user modes - Beginning/Intermediate/Advanced
And that's just from casual use. There's probably more.
Even if you couldn't predict the difference, they should not have built the infrastructure until it was needed. That was a waste of money.
Is that email address like Larry the Cucumber bigidea? Do you program for them? That would be the ultimate way-cool job.
Companies drain money. If it was just paying programmers, it wouldn't have cost this much. But when you add a corporate infrastructure (net access, company databases, backups, CFOs, CIOs, SysAdmins, marketers, a building, computers for everyone, routers, switches, hubs) it all adds up. Eazel's problem was that they started the "company" part _way_ before there was a product. If the VCs had just made them be 6 to 10 hackers in a garage until Nautilus 2.0 came out, they would have had a chance. However, having the overhead of a whole company for over a year before there is anything to sell is what can easily drain $13 million in a year or two.
RedHat is now making money.
ADA Core Technologies makes money.
I'm guessing Mandrake makes money.
IBM has made a butload of money.
Cobalt made money.
Tivo made money.
Who's not making money? VA - because they can't focus, and were overly optimistic. Corel - because they haven't made money in a while even before Linux. Eazel - because their business plan was pretty stupid anyway. Ximian may come out of this all right, if they play their cards well enough.
Add all of the consultants to that, and you've got a picture of whose making money.
Companies don't spend $13 million on a file browser. If you think that's all Eazel was doing, you're wrong. Now, they were definitely over-extravagant in their spending. However, let's take a look at what they might have spent it on:
$2000-$3000/month on Internet access - for one year that's $36,000
$200,000 for their infrastructure - backups, routers, gateways, plus licenses (this could actually have been more. You can really spend up to $2 million easily to make a scalable infrastructure - like if you use Oracle Apps to manage all your stuff).
Let's say they had 10 programmers (I don't know how many they had) on Nautilus - for good programmers, that's about a million per year.
Let's say they had another 10 programmers working on Eazel services, including their packaging and online disk storage, we've got another million there.
Then you have to pay the execs. I'm not going to guess at a figure. Then you've got another twenty to thirty people doing all sorts of marketing/reception/etc.
On top of this, you have office space. If they went for their own building, this could be a few million.
Then you have computers for everyone, and that can get expensive real fast.
So, as you see, $13 million can go pretty fast, especially if you're trying to start-up fast. Most of the dot-coms failed trying to start-up fast. Most companies do. Venture capital makes you think you can do anything because you have all that money, but then you end up wasting it buying the high-end of everything. The thing is that with $13 million, if the investors were willing to wait a little while, _could_ have been spread out over a decade, with the programmers all sitting in a basement, a 28.8 line to the 'net, and not bothered to even hire the marketing guys until the product was out the door and at revision #2. However, most VC places probably don't like that idea, so they try to get a full company in swing before a product is released, which, as you can see, really drains money.
So, of course a file manager doesn't cost that much money, but a company does. The problem is that they formed the company before it was ready, and thus the company drained them of their money. However, they probably wouldn't have gotten VC money doing that. The whole company infrastructure is a bigger drain than any or all projects put together.
Yes, but _someone_ has to make the RPMS. Let's face it, our mothers/grandmothers will never download the latest GNOME, no matter how easy it is. They will only get it when it's released on the RH CD.
Also, Ximian _did_ provide binaries. So what if you had to wait a month. Who really cares? Did your life improve _that_ much after getting GNOME 1.4 that the previous month seemed wasted? I doubt it.
Why wait for the RHs to provide binaries? Well, we have to wait for someone, or we'd be building from source, wouldn't we?
Why do you say it is far better than PostgreSQL? Please add additional information.
Also, getting the source is not the same as open-source. Oracle has also shipped all the source for their database applications for years.
The Linux machine they were using was only for a single experiment, studying plant growth, I believe.
And those developing code can provide something that noone else provides - experience with the code. Just having access to it means nothing. It takes time and energy to get involved and learn how it works. If I have a kernel question, would I rather ask a kid next door or Alan Cox?
The company who actually ran it is called SourceGear. They also are the developers of CVS (well, they bought Cyclic Software, who was the developer of CVS). They still provide server space, but they don't provide full-time developers anymore. I live in Champaign, and actually had a friend doing QA on AbiWord at one point in time, so I'm pretty familiar with the situation. If this has changed, it has been so recently. It used to say so on either AbiSource or SourceGear's site, but I can't find it. The only good references to this I have found are:
9 66 358018/966372740/966389964/index_html
9 66 358018/966372740/966389964/index_html
http://gnotices.gnome.org/gnome-news/966342954/
and
http://gnotices.gnome.org/gnome-news/966342954/
Yes, that is an underscore before html, but slashdot also may put in spaces in the URL that you should take out.
Or, more simply, you could just donate to the Free Software Foundation.