> there are hardly any valuable commercial software without an equivalent open source nephew
Sorry to disagree, but there are a LOT. Perhaps you mean "high-volume off-the-shelf shrinkwrap software"? Here are a few that don't have anything vaguely approaching an alternative, some of which are shrinkwrap:
- QuarkXPress/Adobe Indesign (No scribus doesn't count) - Adobe Photoshop (GIMP is good, but nowhere near pshop. No ICC profiles, no decent CMYK, lack of previews, slow). - Pongrass Classified Pagination (newspaper classified advertising booking & pagination system) - AutoCAD - Adobe Illustrator - Adobe Acrobat (no, I don't mean just distiller) - An AppleTalk server that's reliable and works properly with all mac apps - oodles of small-volume "solutions"-based customised software for industry-specific needs (ATEX, etc).
When I start seeing/ports/ of those apps, I'll be more enthusiastic. We're using linux terminals at work for our sales staff already, and other than some word doc issues with OO.o it works very well. However, our prepress and production department would be simply impossible.
That wouldn't work as well as you might expect. For one thing, some people may actually knowingly install spyware-ridden software in the understanding that "that's the deal" for the $0 software, warts & all. I'm not one of them, but I know a couple of people who are - people who would be quite annoyed that someone else was messing with their computers.
"friendly" viri may be written with good intentions, but they can still break things, and they're still MESSING WITH MY SYSTEMS WITHOUT MY KNOWLEDGE OR PERMISSION.
Hmm... if you're running a decent 'doze variant (win2k, XP if you consider it decent) you can always create her a 'restricted user' account. Unfortunately this breaks most games (which are apallingly coded) but will prevent program installation etc.
They'll be detecting that they're running under 3dmark and changing how they render. The most interesting way to show that this was a benchmark cheat not a coding fuckup would be to
'grep -i 3dmark <nvdriverfiles>' However they will have learned from ATI's q3 mistake and obfuscated any dodgy code if its there - oh well.
The point is that unless you run specific programs, the "optimisations" don't show up at all. If, in fact, they are benchmark cheats not driver bugs.
ExtremeTech never actually claimed that NVidia made specifc tweaks to the driver for 3dmark, and were in fact charitable enough to say they'd wait for a driver that "fixed" these problems before drawing any solid conclusions.
Samba DOES support UNIX permissions. Use the "cifs" client module not the "smbfs" one, and enable UNIX extensions on smbd. Its not hard, and works well.
That said, I haven't tried it in real production yet. I do find it scary that a reverse-engineered MS protocol is now an option for UNIX<->UNIX network file access because NFS is so obsolete and crap that anything looks good in comparison.
... and the gaping wide security hole that is NFS.
"Hello, I'm user ID 500 and I'd like my home directoy... thanks... my accounting data now!".
NFS doesn't actually have security anymore, never has since IP-capable machines became physically portable but more importantly since the assumption that every box would have a trusted admin became invalid... about 15 years ago.
This is only really true of things that were widely used before being superceded. Something like an _experimental_ distributed FS could be dropped from the kernel because nobody could be stuffed updating it to track changes in 2.5 that had broken the code. Unless you had the skills to fix it and add it yourself, or could afford to pay somebody to do it, you'd be stuffed.
from my ISP. Looks like I'll be waiting a/long/ time, since the ISP appears to have no intention of implementing IPv6 until their customers absolutely demand it as a mass.
And so the cycle continues: "nobody is using it, so we won't bother yet." Ipv6 tunneling/sucks/, we need native ipv6 across the networks to make it really useful - and that doesn't look like happening anytime soon.
At least windows and most linux and bsd variants now support ipv6 to a useful, if far from complete, extent.
NAT/sucks/. You have to bounce things through the firewall to talk to an internal host at all, requiring all sorts of nasty kludges. Lots of things don't work incoming through NAT, and lots more (H.323 VoIP, IPSec for example) don't work at all on NATed networks.
I'm waiting with great enthusiasm for my ISP to support ipv6.
You think OpenServer is bad... try OpenServer running Microsoft Xenix binaries using its compatability layer? Our Accts and bookings system doesn't read TERMCAP, and was written before Terminfo existed. Sheer hell.
I for one, do rather like the book. I vaguely remember hearing about it once before but had never even seen it 'till recently when you folk released it.
I admin a SCO OpenServer 5.0.5 box running Microsoft Xenix binaries in binary compatability mode. Need I say more?
Of course, I'm also running a RH9 terminal server, and only about 1/2 to 1/3 of the criticisms apply to it (no sendmail! yay!). Grumble... rm... grumble.
I happen to volunteer for my faviourite charity already, so I'm sure I can spare ten bucks:-)
I must agree w.r.t to the issue with rm. Its a problem - not that you/can/ instantly and silently delete files, but that there's no option not to.
What's wrong with "srm" (safe-rm) or "trash" or "del" (pick a name) grafted on to *n*x, and a "purge" command to ditch files from ~/.trash/ or whatever?
It is, indeed, too bad that it'd break a lot of things to change the behaviour of unlink(1) for consistency with the shell etc.
Basically - its fixable, but it'll break backward compatability. I'm increasingly of the opinion that we need a UNIX2 - that is, STUFF backward compatability and fix the problems once and for all. There are lots of issues much like this one.
Unfortunately, a "compatable mode" would still be needed to allow for legacy apps - but it should be possible, even if slower. Just imagine if UNIX vendors could be made to agree to a standard for the changes...
I've since found out a bit more about the issue, courtesy of a bit of extra reading on both sides - and I can definitely see your position.
There seems to be a very interesting attitude held by the moz folks. Not friendly at all - and some of the stuff that's been going on has been rather underhanded, especially the news reporting.
One question I have, for one, is - how did this name change get through in the first place? I, for one, was surprised to hear of it, as I'd heard of the Firebird DB before.
I seem to remember someone mentioning "newtzilla" - hell, at least that'd be a safe choice.
The things that puzzle me are (a) were NS's corporate lawyers the main decision makers here, as it sounds, (b) why didn't they get in touch with the firebird db folks and find out what/they/ thought, and (c) why firebird - what's so special about the name?
yeah... I was familiar with the firebird project and thought the Phoenix project had made a rather poor name choice. My sympathy (for all it matters) has been very quickly eroded by the reactions of the Firebird folk. Not what you'd call appropriate behaviour.
... if you like giving up half a square kilometer of desk space and the output of a small nuclear powerplant, the Mitsubisihi 2070U is just amazing, with the 2060U a close second.
I work with a lot of quality 21" monitors (newspaper , so we have layout staff) and the new mitsubishi monitors take the cake. Sony might compare (haven't worked with them) but Philips can't even touch them.
Our new machine is a G4/2x866 with a Mitsubishi 2060U - nice, too bad its an (os9) mac. My housemate liked the quality so much he went and bought a 2070U and a Radeon 9700 card - the results are incredible. I've never seen colour like it, and its so crisp its incredible. Almost matches a mid-range LCD for sharpness, but is better than the colour quality of the very best. BLACK IS BLACK WHEN WHITE IS WHITE. Wow.
Sounds like a right pickle. I'm lucky in that I am all roles from IT manager to desktop support, and I set the schedule for deploymens as well as do the deployments. Aah, the IT department of one.
We're using Mozilla with great success, but only on a win9x network doing domain logons against an NT4 server. Since the users' always use the same machines and its a simple network, we have no problems.
Your post could be an important heads-up for others considering a roll-out though - so thanks for that. Its one of the important features of public forums - learn and share (and ignore the trolls and spoilt children).
I guess you know for next time: tell your senior types that "I won't be able to make this deadline X unless you guarantee me uninterrupted, dedicated time. If you can't, factor in wasted time and move the deadline". If you can get that though at the beginning, maybe it'll go better next time.
I'm presently deploying LTSP-based thin clients here at work, and have the pleasure of no deadline - "we go live when its finished and solidly tested." . That's how big stuff should be done, when possible. A development phase with no deadlines, and only once things are ready to roll should dates be set. So, it takes 6 months longer - it also works, every time.
I have to agree in part, but also remember that unified interfaces mean less flexibility for tasks that aren't "what everybody does". For example, KDE/GNOME are all well and good but on a network booted single-task P100 thin client you probably don't want them when a customized version of IceWM does the same job infintely faster and more efficiently.
The "windows-like desktop" is not the only thing people need and want to use linux for. Be aware of that when you make your comments about WMs and desktop environments. I'm all for standardisation on one for normal desktop users, but that doesn't mean that other environments don't have uses for more specialized roles.
I admin a 10 use linux thin client network, so I'm speaking from experience here.
MS Word is a disaster zone like this. There are millions of obscure keyboard shortcuts, and I've run into a couple of problem that make word utterly unusable until its run with a command-line flag to reset the gui appearance to defaults. You know, full-screen on startup but not really fullscreen so you can't just go back to normal mode, etc.
I support 15 word users on a varied hetrogenous (mac, win9x and linux clients) and its a significant part of my job.
If for some reason MS were to come out with an actual WMP/lin release not just license the codecs to, say, Real, that'd be the way to make sure it wasn't modifying libc,/boot/vmlinuz*, ls, ld.so,/etc/ld.so.conf, etc.
The key issue is that MS are considering licensing the WMP9 DRM platform for use on other platforms, including linux.
This does not mean that there will be a Windows Media Player release for linux. Hell, there might be - but why would we care, when it may be possible to get (eg) a Xine plugin?
Personally, I don't like it one bit. I quite definitely want to steer clear of DRM, its a slippery slope indeed. First, pay-per-view downloadable movies. Before you realise what's happened, its demanding a licence or auth number to play your CDs, etc.
I don't want to be someone that people like MS, Real can use in their stats to say "we have xx BlahDRM enabled users, so clearly DRM isn't actually excluding anybody or otherwise flawed - come use it".
since nt4 and cousins (eg XP tablet ed) are not network-boot friendly unless you can throw a REALLY large amount of money at Microsoft.
citrix connections can be encrypted, prevent the user from storing local data, are vaguely efficient, come up really fast, and have the facility for proper user authentication that you couldn't really do with network booting.
However, I presume you have a reason not to use citrix/termserv, most likely since you'll need "tablet pc" OS features or similar that won't work over citrix.
This being the case, I'd put a preliminary comment at "no can do" with the basic OS and tools. Microsoft might well have the "XP remote boot tools" hidden away somewhere - good luck finding them though. I'd be budgeting for tens/hundreds of thousands of dollars in customization by MS or a big sol'n provider with an inside channel to MS.
Slashdot Mirror
> there are hardly any valuable commercial software without an equivalent open source nephew
/ports/ of those apps, I'll be more enthusiastic. We're using linux terminals at work for our sales staff already, and other than some word doc issues with OO.o it works very well. However, our prepress and production department would be simply impossible.
Sorry to disagree, but there are a LOT. Perhaps you mean "high-volume off-the-shelf shrinkwrap software"? Here are a few that don't have anything vaguely approaching an alternative, some of which are shrinkwrap:
- QuarkXPress/Adobe Indesign (No scribus doesn't count)
- Adobe Photoshop (GIMP is good, but nowhere near pshop. No ICC profiles, no decent CMYK, lack of previews, slow).
- Pongrass Classified Pagination (newspaper classified advertising booking & pagination system)
- AutoCAD
- Adobe Illustrator
- Adobe Acrobat (no, I don't mean just distiller)
- An AppleTalk server that's reliable and works properly with all mac apps
- oodles of small-volume "solutions"-based customised software for industry-specific needs (ATEX, etc).
When I start seeing
That wouldn't work as well as you might expect. For one thing, some people may actually knowingly install spyware-ridden software in the understanding that "that's the deal" for the $0 software, warts & all. I'm not one of them, but I know a couple of people who are - people who would be quite annoyed that someone else was messing with their computers.
"friendly" viri may be written with good intentions, but they can still break things, and they're still MESSING WITH MY SYSTEMS WITHOUT MY KNOWLEDGE OR PERMISSION.
Hmm... if you're running a decent 'doze variant (win2k, XP if you consider it decent) you can always create her a 'restricted user' account. Unfortunately this breaks most games (which are apallingly coded) but will prevent program installation etc.
... somewhere in the 50 pages of monospaced all-caps text in a tiny window in a tiny font on a medium-dark gray background.
/reads/ EULAs - and it infuriates me just how well they're designed to make it an incredible chore.
Wonder if the next step will be to make the text intentionally blurry?
I'm one of those people that
*sigh*
They'll be detecting that they're running under 3dmark and changing how they render. The most interesting way to show that this was a benchmark cheat not a coding fuckup would be to
'grep -i 3dmark <nvdriverfiles>'
However they will have learned from ATI's q3 mistake and obfuscated any dodgy code if its there - oh well.
The point is that unless you run specific programs, the "optimisations" don't show up at all. If, in fact, they are benchmark cheats not driver bugs.
ExtremeTech never actually claimed that NVidia made specifc tweaks to the driver for 3dmark, and were in fact charitable enough to say they'd wait for a driver that "fixed" these problems before drawing any solid conclusions.
/need/ to say it...
Of course, they didn't really
Samba DOES support UNIX permissions. Use the "cifs" client module not the "smbfs" one, and enable UNIX extensions on smbd. Its not hard, and works well.
That said, I haven't tried it in real production yet. I do find it scary that a reverse-engineered MS protocol is now an option for UNIX<->UNIX network file access because NFS is so obsolete and crap that anything looks good in comparison.
... and the gaping wide security hole that is NFS.
... thanks ... my accounting data now!".
... about 15 years ago.
"Hello, I'm user ID 500 and I'd like my home directoy
NFS doesn't actually have security anymore, never has since IP-capable machines became physically portable but more importantly since the assumption that every box would have a trusted admin became invalid
KILL NFS, we need something that doesn't suck.
This is only really true of things that were widely used before being superceded. Something like an _experimental_ distributed FS could be dropped from the kernel because nobody could be stuffed updating it to track changes in 2.5 that had broken the code. Unless you had the skills to fix it and add it yourself, or could afford to pay somebody to do it, you'd be stuffed.
from my ISP. Looks like I'll be waiting a /long/ time, since the ISP appears to have no intention of implementing IPv6 until their customers absolutely demand it as a mass.
/sucks/, we need native ipv6 across the networks to make it really useful - and that doesn't look like happening anytime soon.
And so the cycle continues: "nobody is using it, so we won't bother yet." Ipv6 tunneling
At least windows and most linux and bsd variants now support ipv6 to a useful, if far from complete, extent.
NAT /sucks/. You have to bounce things through the firewall to talk to an internal host at all, requiring all sorts of nasty kludges. Lots of things don't work incoming through NAT, and lots more (H.323 VoIP, IPSec for example) don't work at all on NATed networks.
I'm waiting with great enthusiasm for my ISP to support ipv6.
*shudder*
You think OpenServer is bad... try OpenServer running Microsoft Xenix binaries using its compatability layer? Our Accts and bookings system doesn't read TERMCAP, and was written before Terminfo existed. Sheer hell.
Yeah....
... rm ... grumble.
:-)
I for one, do rather like the book. I vaguely remember hearing about it once before but had never even seen it 'till recently when you folk released it.
I admin a SCO OpenServer 5.0.5 box running Microsoft Xenix binaries in binary compatability mode. Need I say more?
Of course, I'm also running a RH9 terminal server, and only about 1/2 to 1/3 of the criticisms apply to it (no sendmail! yay!). Grumble
I happen to volunteer for my faviourite charity already, so I'm sure I can spare ten bucks
I must agree w.r.t to the issue with rm. Its a problem - not that you /can/ instantly and silently delete files, but that there's no option not to.
What's wrong with "srm" (safe-rm) or "trash" or "del" (pick a name) grafted on to *n*x, and a "purge" command to ditch files from ~/.trash/ or whatever?
It is, indeed, too bad that it'd break a lot of things to change the behaviour of unlink(1) for consistency with the shell etc.
Basically - its fixable, but it'll break backward compatability. I'm increasingly of the opinion that we need a UNIX2 - that is, STUFF backward compatability and fix the problems once and for all. There are lots of issues much like this one.
Unfortunately, a "compatable mode" would still be needed to allow for legacy apps - but it should be possible, even if slower. Just imagine if UNIX vendors could be made to agree to a standard for the changes...
Too bad it'll never happen.
The dual Xeon board of the machine I'm working on right now supports up to 8 gigs of RAM.
Of course, we only have 2 gig in it, but its room to grow for our nice linux terminal server.
I've since found out a bit more about the issue, courtesy of a bit of extra reading on both sides - and I can definitely see your position.
/they/ thought, and (c) why firebird - what's so special about the name?
There seems to be a very interesting attitude held by the moz folks. Not friendly at all - and some of the stuff that's been going on has been rather underhanded, especially the news reporting.
One question I have, for one, is - how did this name change get through in the first place? I, for one, was surprised to hear of it, as I'd heard of the Firebird DB before.
I seem to remember someone mentioning "newtzilla" - hell, at least that'd be a safe choice.
The things that puzzle me are (a) were NS's corporate lawyers the main decision makers here, as it sounds, (b) why didn't they get in touch with the firebird db folks and find out what
yeah ... I was familiar with the firebird project and thought the Phoenix project had made a rather poor name choice. My sympathy (for all it matters) has been very quickly eroded by the reactions of the Firebird folk. Not what you'd call appropriate behaviour.
... if you like giving up half a square kilometer of desk space and the output of a small nuclear powerplant, the Mitsubisihi 2070U is just amazing, with the 2060U a close second.
I work with a lot of quality 21" monitors (newspaper , so we have layout staff) and the new mitsubishi monitors take the cake. Sony might compare (haven't worked with them) but Philips can't even touch them.
Our new machine is a G4/2x866 with a Mitsubishi 2060U - nice, too bad its an (os9) mac. My housemate liked the quality so much he went and bought a 2070U and a Radeon 9700 card - the results are incredible. I've never seen colour like it, and its so crisp its incredible. Almost matches a mid-range LCD for sharpness, but is better than the colour quality of the very best. BLACK IS BLACK WHEN WHITE IS WHITE. Wow.
Sounds like a right pickle. I'm lucky in that I am all roles from IT manager to desktop support, and I set the schedule for deploymens as well as do the deployments. Aah, the IT department of one.
We're using Mozilla with great success, but only on a win9x network doing domain logons against an NT4 server. Since the users' always use the same machines and its a simple network, we have no problems.
Your post could be an important heads-up for others considering a roll-out though - so thanks for that. Its one of the important features of public forums - learn and share (and ignore the trolls and spoilt children).
I guess you know for next time: tell your senior types that "I won't be able to make this deadline X unless you guarantee me uninterrupted, dedicated time. If you can't, factor in wasted time and move the deadline". If you can get that though at the beginning, maybe it'll go better next time.
I'm presently deploying LTSP-based thin clients here at work, and have the pleasure of no deadline - "we go live when its finished and solidly tested." . That's how big stuff should be done, when possible. A development phase with no deadlines, and only once things are ready to roll should dates be set. So, it takes 6 months longer - it also works, every time.
Wow. Hot button material.
I have to agree in part, but also remember that unified interfaces mean less flexibility for tasks that aren't "what everybody does". For example, KDE/GNOME are all well and good but on a network booted single-task P100 thin client you probably don't want them when a customized version of IceWM does the same job infintely faster and more efficiently.
The "windows-like desktop" is not the only thing people need and want to use linux for. Be aware of that when you make your comments about WMs and desktop environments. I'm all for standardisation on one for normal desktop users, but that doesn't mean that other environments don't have uses for more specialized roles.
I admin a 10 use linux thin client network, so I'm speaking from experience here.
MS Word is a disaster zone like this. There are millions of obscure keyboard shortcuts, and I've run into a couple of problem that make word utterly unusable until its run with a command-line flag to reset the gui appearance to defaults. You know, full-screen on startup but not really fullscreen so you can't just go back to normal mode, etc.
I support 15 word users on a varied hetrogenous (mac, win9x and linux clients) and its a significant part of my job.
strace /path/to/binary 2>&1 | tee /tmp/tracelog
/tmp/tracelog and stdout.
/boot/vmlinuz*, ls, ld.so, /etc/ld.so.conf, etc.
will write a strace to
If for some reason MS were to come out with an actual WMP/lin release not just license the codecs to, say, Real, that'd be the way to make sure it wasn't modifying libc,
The key issue is that MS are considering licensing the WMP9 DRM platform for use on other platforms, including linux.
This does not mean that there will be a Windows Media Player release for linux. Hell, there might be - but why would we care, when it may be possible to get (eg) a Xine plugin?
Personally, I don't like it one bit. I quite definitely want to steer clear of DRM, its a slippery slope indeed. First, pay-per-view downloadable movies. Before you realise what's happened, its demanding a licence or auth number to play your CDs, etc.
I don't want to be someone that people like MS, Real can use in their stats to say "we have xx BlahDRM enabled users, so clearly DRM isn't actually excluding anybody or otherwise flawed - come use it".
since nt4 and cousins (eg XP tablet ed) are not network-boot friendly unless you can throw a REALLY large amount of money at Microsoft.
citrix connections can be encrypted, prevent the user from storing local data, are vaguely efficient, come up really fast, and have the facility for proper user authentication that you couldn't really do with network booting.
However, I presume you have a reason not to use citrix/termserv, most likely since you'll need "tablet pc" OS features or similar that won't work over citrix.
This being the case, I'd put a preliminary comment at "no can do" with the basic OS and tools. Microsoft might well have the "XP remote boot tools" hidden away somewhere - good luck finding them though. I'd be budgeting for tens/hundreds of thousands of dollars in customization by MS or a big sol'n provider with an inside channel to MS.
alas they'd also have to block legit SMB/CIFS access, yes?
/really/ needs it can set up an IPSec VPN to use it safely.
Of course, that's in many ways a good thing, and anybody who