If I write free software in my spare time but my employer finds out about it and disapproves, what can they do about it? Would using e-mail for the project constitute some kind of violation of corporate resources? Could I be fired? Would I have to stop working on it even if it's during my own time? Could I be sued?
This thing is huge. I was thinking of something around the size of a 5 port hub for $50. If it's as big as a computer why not do it in software a la Linux style? Hell, you could probably get away with writing a couple of shell scripts to generate traffic and some deliberately bungled iptables mods. Put that on a Linux Router Project floppy disk and your done. Oh, wait a minute. I wonder selling a big box with little chips inside is more profitable? Hey!
I wonder if its even possible for an alternat office suite to gain a foothold in the market place.... but also its mindshare and what people are used to using.
I don't think so. Take MS Word for example. This is a pretty sophisticated product. There's nothing that even comes close. Eveything else is a cheap imitation that when examined for more than 5 minutes reveils just how shollow the features are. I do not believe the mind share argument. I believe if your product is truely good it will eventually succeed. There are many features that by themselves seem trite but together they make the difference that leaves competitors in the dust. Someone could write a word processor that could beat out MS Word. It may be a pretty sophisticated program but in many ways it's very broken.
The problem with open source office products is that something like a word processor demands a very serious design and great code to make it happen and there just isn't enough critical mass to get a serious product out.
In general there are two ways open source projects develop. One is when one person quietly writes it alone. This can result in a pretty good program because they aren't distracted with "talkers" and they have a good focus and singular vision. The other way is when you have a loosely associated bunch of people working on the same codebase. This is a pretty good way to write crappy software. No one knows how the whole thing works so they add their little bit and try not to change anything they don't have to. This doesn't work because sometimes refactoring of something that transcends the entire codebase is required.
I'm speaking in general terms of course and there are exceptions to every rule(the Linux kernel has a pretty decent development model going but notice everyone is totally dedicated and sobordinate to Linus and it's being funneled into his codebase; there's no CVS).
So given the above, the problem with developing an open source office product is that no one person is willing to put the tremendous amount of thought and time(years) into something that could be competitive with something like word.
What I'm hoping will happen is people will develop libraries with well defined interfaces that abstract the components of such applications in a way that will allow the constructions of such enourmously complicated programs. But that's another area of weakness in open source projects. No one thinks too much about librarifying thier code and when they do other developers always think that re-doing it themselves from scratch would be better; how many ftp clients are out there?
Exactly. Well said. Use the vendors. It will cost you an extra 5%(I don't really know for sure what the markup is but it can't be that much). What happends if you build the box yourself and find theres some wierd chipset incompatabily? I think that 5% is pretty good knowing you'll get a tested box and have someone to call.
Re:Bill Joy - Open Source curmudgeon
on
Sun Launches JXTA
·
· Score: 2
The code for all of these is clean? Mozilla works well? You wanna tell me where the documentation for Samba is please? Guess which company John Ousterhout worked for while doing some of his most significant work on the Tcl language? You guessed it, Sun Microsystems.
Re:Bill Joy - Open Source curmudgeon
on
Sun Launches JXTA
·
· Score: 1
He said that while everyone has been talking about this "Open Source" thing, he just wanted software components to work well consistently, something that he doesn't believe Open Source does particularly well.
Sun will advocate Open Source when it in convenient. Do not be mislead as to their intentions.
So to advocate Open Source he must lie and say that all Open Source works great? I'd say it's a pretty accurate statement on his part. In fact I'd say, Open Source code is a flea-market of mostly crap. And even when it does work the code is ugly. Please name one open source project that works well, has clean code with decent comments, is documented even a little, and written in a way that is extensible.
Re:Marketing mindset a little strong.
on
Sun Launches JXTA
·
· Score: 1
>> Bill Joy isn't a "developer" - he is %100 manager.
> Not true. His code is out there.
Such as...? Give an example of some recent code by Bill Joy's then.
I'll be *very* suprised if he has written anything non-trivial for years.
Bill Joy has written some pretty important stuff; like the original vi. And at this point he doesn't have to code. He can just sit back , dream stuff up, and have other people actually write it.
You, too, can write Silicon Valley press releases!
[Thing], originally a research project spearheaded by [famous employee] and [semi-famous employee or soon-to-be famous employee], is now an [buzzword], with a vision to enable developers to create innovative services and applications. It addresses the need for an open, generalized [CS term] that interoperates with any peer on the network including PCs, servers and other connected devices.
Ok, lemmy try,
[My dog], originally a research project spearheaded by [The President of the United States] and [myself], is now an [Global Enterprise], with a vision to enable developers to create innovative services and applications. It addresses the need for an open, generalized [hardware interrupt] that interoperates with any peer on the network including PCs, servers and other connected devices.
Mmm....
Re:This could be very interesting...
on
Sun Launches JXTA
·
· Score: 1
Using XML as the protocol works when people have relatively high bandwidth (XML metadata can start to approach the size of the actual data in the document. This doesn't matter on a T1, but it sure does at 33.6).
But the docs say XML is not required. With a little recoding you can use something else down the line. Also I didn't get the impression that the XML stuff was elemental, meaning using the shell to find files and such involves no XML. The actual messaging is binary.
Hey Sun! Get lost!...And take the marketroids with you!
Ok, I know this ones a troll but the sentiment of other comments doesn't sound good. What's the problem? So Sun is a Mega-Corp. But the're giving us stuff! They did the whole Open Office thing. And now they give us JXTA which is not some worthless bone thrown out of Bill Joy's garbage bin but something that could potentially make Gnutella, Napster, and Freenet look like tin cans with some string. They've got specs. Thats the kind of organization OSS projects need. When was the last time you saw an OSS project with a spec. Never. Invariably, you just start typing and half the time it turns out to be crap. Whatever the truth is, you should look at the software before posting fooling comments. I suspect the normal people are off looking at the white-paper.
Ahh, well then you're connecting to your neighbors machine. Excuse me. Like I said it won't make it past your broadcast address. BTW, I would be carefull about doing this in the first place. Your hostname is passes during nbt session establishment. So if by chance it really is your neighbor(I'm not joking about this, they could be very close by), they could theoretically track you. In fact if they happen to be running Samba, a log file with your hostname will be written in/var/log/samba with debugging info!
why would it be hard for them[(MS) to change the protocol and break Samba or other CIFS servers and clients]?
Because the packet formats and server/client behavior has been documented by MS themselves. It's not great documentation but it's good enough that if they changed anything fundamental they would invalidate the dozens of server and client implementations based on their own instructions. Do you think companies like Network Appliance, EMC, IBM, HP wouldn't have a total freakn' canniption? They have enough legal troubles as it is. The only thing I could see them changing is paripheral stuff like RPC calls and domain authentication/management stuff. Also, it would be contradictory to their recent(since ~97) behavior of documenting the protocol and answering questions. The Storage Network Industry Association(SNIA) recently released an enhanced version of the current spec based on addition information from MS. Yes, their actually playing ball on this one(the irony is that other vendors have been stingy with what they have learned about CIFS over the years(this is what the SNIA is trying to remedy)).
Besides CIFS/SMB is such a complete and utter mess that they run the risk of someone coming up with a MUCH better protocol all together. You could do all of what CIFS/SMB does and more in half the traffic. For example to issue an an RPC command you have the RPC layer, which is encoded in a Transaction2, which is encoded as an SMB, which is passed over the NetBIOS session service. It's as if every time they came up with somthing new(e.g. RAP) they just slapped it right on top of that last thing. What a nightmare! It makes NFS look like ftp. Beleive me, they don't care if you can interoperate with it. I doubt they fully understand the dynamics of this rediculous protocol. If they so much as looked at the code funny they would seriously affect the stability of the Windows platform(this has actually been the source of a great deal of instability in Windows; the server, as well as quite a bit of other sillyness runs in the Kernel).
Actually I don't think that will work. You need a netbios name in the CalledName field during session establishment. To get that name you need to do a node status request(or know the name of the machine your connecting to in advance) and then connect by ip. Theoretically it can be done but not with smbmount. It will try to do a broadcast lookup and stop right there because(as the other poster pointed out) it will get blocked at the router.
I do admit that there is one thing that could be a lot easier. By default Samba is not configured to use encrypted authentication but it is required by Win98 and up. I would say this little dude stops most newbies in there jammies. And yet they don't make setting that up trivial. Here's how you do it anyway (see/usr/doc/samba-2.0.6/docs/textdocs/ENCRYPTION.txt or equivalent for details).
Run cat/etc/passwd | mksmbpasswd.sh >/etc/smbpasswd and put the directives:
in the [global] section of/etc/smb.conf. Now run the smbpasswd program like you would the normal passwd program to hash the password and update the/etc/smbpasswd file. Finally restart smbd like/etc/rc.d/init.d/smb restart and try to access the share. It should work now!
Re:a question for Jeremy Allison (or anybody)
on
Samba 2.2.0 Released
·
· Score: 1
The bug that bothers me most is when a windows box goes down (can you imagine that?) only root can unmount the [smbmount] share.
Use autofs. I believe it can recover from this condition but if you actively try to access the share while the Windows machine is down it might take a while to reset.
To do this, add a line like this to your/etc/auto.misc (in the case of Red Hat at least). Notice in this example I specify the uid as the user to be given access. That's because I only wanted jimbo to have access to that share. I believe you can use the mount 'user' option to allow any ordinary user to mount the volume. Now start/restart with/etc/rc.d/init.d/autofs restart and simply do like ls/misc/jimbo1 and viola. You don't even have to use the mount command either.
Make sure the permissions are strick in this file though!
Re:Stupid question about netbios naming resolution
on
Samba 2.2.0 Released
·
· Score: 1
am I the only one who thinks it would be very useful to have Linux's hostname resolving scheme support Netbios name resolution?
Well, as Jeremy has pointed out, you may use the "wins" module for nsswitch. I have never used it but I can imaging mixing DNS lookups with WINS lookups might be a little tricky.
like Windows Network Neighbourhood - how does the new KDE do this?). Seems to me "Linux as a workstation" could benefit seriously from this. You don't really want to be going around explaining the "smbclient" command parameters to every employee
Richard Sharpe is woking on libsmbclient which is designed with this sort of stuff specifically in mind. He's also working on a GTK+ app to do just what you suggest.
Incedentally an interesting thing that came out of this was the development of the smb URL syntax which so far is:
Scheduled release for Windows XP is in the near future, but I don't belive it has went gold yet, I say M$FT breaks everything again, or in a friendly windowsupdate patch soon after release.
I'll take that bet. Microsoft has actually been quite cooperative with the Storage Network Industry Association(SNIA) in providing more detailed information on how the Common Internet FileSystem(CIFS) (yes, that's MS's offical name for their networking protocol) servers and clients are supposed to work. It would be very hard for them to change something and not have it affect a lot of implementations. Now authentication software is a different story. I'm not sure what the status of their Kerberos business is.
Slashdot Mirror
how about just a measly Sky Scraper first? If you can build a 47000km elevator I suspect you could make one heck of an office building.
Has it occured to anyone that MS might be just propping up the penguin to show the Justice Department that they do have competition?
If I write free software in my spare time but my employer finds out about it and disapproves, what can they do about it? Would using e-mail for the project constitute some kind of violation of corporate resources? Could I be fired? Would I have to stop working on it even if it's during my own time? Could I be sued?
This thing is huge. I was thinking of something around the size of a 5 port hub for $50. If it's as big as a computer why not do it in software a la Linux style? Hell, you could probably get away with writing a couple of shell scripts to generate traffic and some deliberately bungled iptables mods. Put that on a Linux Router Project floppy disk and your done. Oh, wait a minute. I wonder selling a big box with little chips inside is more profitable? Hey!
I wonder if its even possible for an alternat office suite to gain a foothold in the market place.
I don't think so. Take MS Word for example. This is a pretty sophisticated product. There's nothing that even comes close. Eveything else is a cheap imitation that when examined for more than 5 minutes reveils just how shollow the features are. I do not believe the mind share argument. I believe if your product is truely good it will eventually succeed. There are many features that by themselves seem trite but together they make the difference that leaves competitors in the dust. Someone could write a word processor that could beat out MS Word. It may be a pretty sophisticated program but in many ways it's very broken.
The problem with open source office products is that something like a word processor demands a very serious design and great code to make it happen and there just isn't enough critical mass to get a serious product out.
In general there are two ways open source projects develop. One is when one person quietly writes it alone. This can result in a pretty good program because they aren't distracted with "talkers" and they have a good focus and singular vision. The other way is when you have a loosely associated bunch of people working on the same codebase. This is a pretty good way to write crappy software. No one knows how the whole thing works so they add their little bit and try not to change anything they don't have to. This doesn't work because sometimes refactoring of something that transcends the entire codebase is required.
I'm speaking in general terms of course and there are exceptions to every rule(the Linux kernel has a pretty decent development model going but notice everyone is totally dedicated and sobordinate to Linus and it's being funneled into his codebase; there's no CVS).
So given the above, the problem with developing an open source office product is that no one person is willing to put the tremendous amount of thought and time(years) into something that could be competitive with something like word.
What I'm hoping will happen is people will develop libraries with well defined interfaces that abstract the components of such applications in a way that will allow the constructions of such enourmously complicated programs. But that's another area of weakness in open source projects. No one thinks too much about librarifying thier code and when they do other developers always think that re-doing it themselves from scratch would be better; how many ftp clients are out there?
Look at the moderation on this:
Moderation Totals:Offtopic=3, Troll=1, Insightful=1, Interesting=1, Funny=2, Overrated=1, Underrated=1, Total=10
I really screwed up with the F i r s t P o s t subject.
A long time ago I wrote:
I had trouble getting XFree86 3.3.1 to run on my new IBM Thinkpad 560E with the Trident Cyber 9382 running the latest version of Slackware (3.4)
Hmm, X really hasn't come that far has it? That new 560E is a 150 and cost me more than $3000! And Slackware was at 3.4? Great. I feel old now.
Exactly. Well said. Use the vendors. It will cost you an extra 5%(I don't really know for sure what the markup is but it can't be that much). What happends if you build the box yourself and find theres some wierd chipset incompatabily? I think that 5% is pretty good knowing you'll get a tested box and have someone to call.
The code for all of these is clean? Mozilla works well? You wanna tell me where the documentation for Samba is please? Guess which company John Ousterhout worked for while doing some of his most significant work on the Tcl language? You guessed it, Sun Microsystems.
He said that while everyone has been talking about this "Open Source" thing, he just wanted software components to work well consistently, something that he doesn't believe Open Source does particularly well.
Sun will advocate Open Source when it in convenient. Do not be mislead as to their intentions.
So to advocate Open Source he must lie and say that all Open Source works great? I'd say it's a pretty accurate statement on his part. In fact I'd say, Open Source code is a flea-market of mostly crap. And even when it does work the code is ugly. Please name one open source project that works well, has clean code with decent comments, is documented even a little, and written in a way that is extensible.
> Not true. His code is out there.
Such as...? Give an example of some recent code by Bill Joy's then.
I'll be *very* suprised if he has written anything non-trivial for years.
Bill Joy has written some pretty important stuff; like the original vi. And at this point he doesn't have to code. He can just sit back , dream stuff up, and have other people actually write it.
You, too, can write Silicon Valley press releases!
[Thing], originally a research project spearheaded by [famous employee] and [semi-famous employee or soon-to-be famous employee], is now an [buzzword], with a vision to enable developers to create innovative services and applications. It addresses the need for an open, generalized [CS term] that interoperates with any peer on the network including PCs, servers and other connected devices.
Ok, lemmy try,
Mmm....
Using XML as the protocol works when people have relatively high bandwidth (XML metadata can start to approach the size of the actual data in the document. This doesn't matter on a T1, but it sure does at 33.6).
But the docs say XML is not required. With a little recoding you can use something else down the line. Also I didn't get the impression that the XML stuff was elemental, meaning using the shell to find files and such involves no XML. The actual messaging is binary.
Hey Sun! Get lost!
Ok, I know this ones a troll but the sentiment of other comments doesn't sound good. What's the problem? So Sun is a Mega-Corp. But the're giving us stuff! They did the whole Open Office thing. And now they give us JXTA which is not some worthless bone thrown out of Bill Joy's garbage bin but something that could potentially make Gnutella, Napster, and Freenet look like tin cans with some string. They've got specs. Thats the kind of organization OSS projects need. When was the last time you saw an OSS project with a spec. Never. Invariably, you just start typing and half the time it turns out to be crap. Whatever the truth is, you should look at the software before posting fooling comments. I suspect the normal people are off looking at the white-paper.
the caption reads:
6 Structure to rest on 3,000ft wide concrete base with lake to absorb earthquake shockwaves.
How can that be when the building itself is 3,700ft tall?
And what principle of Engineering permits the structure to be wider at the middle than it is at the bottom!?
atomic bomb! Could you take one of these little reactors slap a detonator to it and make a A-bomb!?
If so I'm moving.
their ad compaign is working pretty well.
Peace, Love, and Linux :~)
Ahh, well then you're connecting to your neighbors machine. Excuse me. Like I said it won't make it past your broadcast address. BTW, I would be carefull about doing this in the first place. Your hostname is passes during nbt session establishment. So if by chance it really is your neighbor(I'm not joking about this, they could be very close by), they could theoretically track you. In fact if they happen to be running Samba, a log file with your hostname will be written in
why would it be hard for them[(MS) to change the protocol and break Samba or other CIFS servers and clients]?
Because the packet formats and server/client behavior has been documented by MS themselves. It's not great documentation but it's good enough that if they changed anything fundamental they would invalidate the dozens of server and client implementations based on their own instructions. Do you think companies like Network Appliance, EMC, IBM, HP wouldn't have a total freakn' canniption? They have enough legal troubles as it is. The only thing I could see them changing is paripheral stuff like RPC calls and domain authentication/management stuff. Also, it would be contradictory to their recent(since ~97) behavior of documenting the protocol and answering questions. The Storage Network Industry Association(SNIA) recently released an enhanced version of the current spec based on addition information from MS. Yes, their actually playing ball on this one(the irony is that other vendors have been stingy with what they have learned about CIFS over the years(this is what the SNIA is trying to remedy)).
Besides CIFS/SMB is such a complete and utter mess that they run the risk of someone coming up with a MUCH better protocol all together. You could do all of what CIFS/SMB does and more in half the traffic. For example to issue an an RPC command you have the RPC layer, which is encoded in a Transaction2, which is encoded as an SMB, which is passed over the NetBIOS session service. It's as if every time they came up with somthing new(e.g. RAP) they just slapped it right on top of that last thing. What a nightmare! It makes NFS look like ftp. Beleive me, they don't care if you can interoperate with it. I doubt they fully understand the dynamics of this rediculous protocol. If they so much as looked at the code funny they would seriously affect the stability of the Windows platform(this has actually been the source of a great deal of instability in Windows; the server, as well as quite a bit of other sillyness runs in the Kernel).
Probably was your machine in the other room.
yeah but NetBIOS over TCP/IP does, dumbass.
Actually I don't think that will work. You need a netbios name in the CalledName field during session establishment. To get that name you need to do a node status request(or know the name of the machine your connecting to in advance) and then connect by ip. Theoretically it can be done but not with smbmount. It will try to do a broadcast lookup and stop right there because(as the other poster pointed out) it will get blocked at the router.
Run cat /etc/passwd | mksmbpasswd.sh > /etc/smbpasswd and put the directives:
encrypt passwords = yes /etc/smbpasswd
smb passwd file =
in the [global] section of /etc/smb.conf. Now run the smbpasswd program like you would the normal passwd program to hash the password and update the /etc/smbpasswd file. Finally restart smbd like /etc/rc.d/init.d/smb restart and try to access the share. It should work now!
The bug that bothers me most is when a windows box goes down (can you imagine that?) only root can unmount the [smbmount] share.
Use autofs. I believe it can recover from this condition but if you actively try to access the share while the Windows machine is down it might take a while to reset.
To do this, add a line like this to your /etc/auto.misc (in the case of Red Hat at least). Notice in this example I specify the uid as the user to be given access. That's because I only wanted jimbo to have access to that share. I believe you can use the mount 'user' option to allow any ordinary user to mount the volume. Now start/restart with /etc/rc.d/init.d/autofs restart and simply do like ls /misc/jimbo1 and viola. You don't even have to use the mount command either.
Example /etc/auto.misc entry:
jimbo1 -fstype=smbfs,username=jimbo,uid=500,fmask=770,dma sk=770,password=mypass,workgroup=myntdomain ://jimbo1/myshare
Make sure the permissions are strick in this file though!
am I the only one who thinks it would be very useful to have Linux's hostname resolving scheme support Netbios name resolution?
Well, as Jeremy has pointed out, you may use the "wins" module for nsswitch. I have never used it but I can imaging mixing DNS lookups with WINS lookups might be a little tricky.
like Windows Network Neighbourhood - how does the new KDE do this?). Seems to me "Linux as a workstation" could benefit seriously from this. You don't really want to be going around explaining the "smbclient" command parameters to every employee
Richard Sharpe is woking on libsmbclient which is designed with this sort of stuff specifically in mind. He's also working on a GTK+ app to do just what you suggest.
Incedentally an interesting thing that came out of this was the development of the smb URL syntax which so far is:
smb://domain;user:pass@server/share/path/to/file
or to get a Network Neighborhood list of servers:
smb://workgroup
Scheduled release for Windows XP is in the near future, but I don't belive it has went gold yet, I say M$FT breaks everything again, or in a friendly windowsupdate patch soon after release.
I'll take that bet. Microsoft has actually been quite cooperative with the Storage Network Industry Association(SNIA) in providing more detailed information on how the Common Internet FileSystem(CIFS) (yes, that's MS's offical name for their networking protocol) servers and clients are supposed to work. It would be very hard for them to change something and not have it affect a lot of implementations. Now authentication software is a different story. I'm not sure what the status of their Kerberos business is.