dpkg has it's own technical advantages too, such as suggests and recommends
I know that. You know I know that. But you asked about Red Hat advantages so that's what I gave you. I'm well aware dpkg has advantages over RPM in many ways and I've indicated such in my discussion of the topic. I've never alleged RPM wass a better all round solution, in fact, I've done the opposite - state that there's more to the argument than technical considerations.
Ignoring for the moment that IMAP4, ical, vcard and MAPI are all very different things,
MAPI or IMAP can be used as tranbsports for a variety information stored in a variety of formats, such as mbox, ical, vcard, or the Exchange message store.
I was aware that someone might misunderstand that analogy, so I added another -/sbin/init.d,/etc/rc.d/init.d,/usr/doc and other nonstandard file formats that aren't wrong (it may be more logical from acertain point of view to put intiscripts in sbin). You haven't responded to this.
As I said, the LSB does indeed accept compromises - there's no need to tell me this. As I've also said, that doesn't mean those compromises are OK.
To me, alien sounds like a fine way of installing LSB-conformant RPM packages.
They're not packages once alien is finished with them. `Package Management',as I see it, is based on relationships between packages which are destroyed once alien has been used on them.
The LSB have realised that trying to force all distributions to switch to the RPM format internally would be a fatal mistake
The LSB couldn't force anyone to do anything - it's an optional standard. Gentoo can exist, so can Slackware, so can LFS - they see advantages in not being LSB compliant and choose not to participate. More power to them. There's no practical way could `destroy the usefulness of innovative new distributions like Gentoo' by specifying a complete rpm implementation anymore than they stopped Slackware from existing by specifying SysV init.
However, for those who seek to use Linux in a larger scale environment or for something they consider important, some kind of structure is desired. Currently that structure is provided by a particular Linux distribution. People write and certify (closed source) apps for distros. People learn skills only applicable to distros. People write apps only applicable to distros. People compile and package applicable to distros. Use a less popular distro, and you'll either have to do a lot more work or miss out entirely.
The LSB is about writing apps for Linux. There's going to be
the toll on the Debian project would be enormous
The toll on packagers from the current situation is already enormous. Enhancing RPM to have a superset of current Linuxpakcage management features would and migrating the this RPM implementation would involve alot of short term work for a lot of long term gain when maintaining packages in future. Hell, if every (common) Linux distribution standardized on a package format you might even see people writing more apps that package nicely rather than using custom build jobs you have to play with in sed to install to the correct location.
As I've said, its no skin off my nose. If you've never been pissed off that a particular company hasn't packaged their proprietary app for your distro, then I suppose my point has no basis.
If you mean technical advantages, there's a couple, notably ghost files, triggers, and a more verification system which provides more information about exactly how a file has changed.
Practically, both can have policy, its the matter of the distribution to produce and stick to good policy. Some areas are handled better by some distros (eg, all the packages are signed, or they have standard naming conventions) but that's a higher level issue and neither RPM nor Dpkg based distros excel in that area.
and what advantages does RPM have over the.deb format, apart from the LSB's rubber stamp?
I'll be evil and answer with a question, thanks Alex.
What advantages does/etc/init.d and/usr/share/doc have over/sbin/init.d,/etc/rc.d/init.d and/usr/doc apart from the LSBs rubber stamp?
What advantages do IMAP4, ical, and vcard have over MAPI, apart from the IETF's rubber stamp?
The lack of standardization amongst Linux distributions and compromises that result (like labelling alien an aceptable implementation or RPM) have a greater negative effect on Linux adoption than a few minor technicaladvantages that various formats have over the other. More Linux adoption means hardware works, upstream providers can support your operating system, and a bunch of other stuff.
Its the Linux packaging method, created by Red Hat. See http://www.rpm.org.
Now if only Debian would start using it (and improving it in any area they see lacking) instead of supporting the LSB by turning packages into dumb archives with alien.;)
Yes, but if your day requires you to run Outlook 2000 throughout your day, then its not practical to shut Wine off (the Ximian Connector still doesn't do everything Outlook does with regards to Exchange).
One mitigating factor: codeweavers do built in a protection against executable attachments in their winex product.
Run Office setup fro myour menu (thats ~/cxoffice/bin/officesetup)
Click configuration
Hit the advanced button
Notice the Outlook security tab, which is turned on by default. "prevent MS Outlook fro mrunning files with these extensions: vbs;wsf;vbe;wsh;hta;bat;pif;exe;scr;lnk"
Wait for StarOffice to get anough market share to have its own real viruses.
My home machine is an Athlon 900 with 640MB of RAM. My work machine is a Duron 800 with 384MB.
I've yet to see a version of Mozilla, up to an including 1.1 and Phoeniz 0.2, whose UI performs adequately. Gecko is fine, Galeon is great, XUL is not. Why is Mozilla so slow, and if I can't stand having mymenus lagbehind my mouse clicks, why should I expect my users to?
Don Becker does indeed ssom to have done some great work, especially on network drivers and clustering. I can't seem to find any information on his work regarding `making Linux usable' that's mentioned in the byline of this story. Am I missing something, or is that part of the intro a little bit confused (maybe Slashdot has a different definition of usability from the rest of the computing world)?
he: Hey babe, wanna watch a movie ?? she: sure he: wait till i boot the player she: ?????? he: here we go... she: is it me, or is it getting hotter in here?? he: so take off all your clothes! she: I am getting so hot, I'm going to take my clothes off!
I find your post quite typical of/. myself: loud, rude, lacking supporting arguments, and illustrating a fundamental misunderstanding of your topic.
As a system administrator (among other things), I find that using RPMs are the sure path to unmaintainability and broken systems.
Why not provide some supporting arguments? I'll go first: standardized install, uninstall, querying, verification, GPG signing, and repeatability, LSB compliance.
Considering that it is *easier* to build from source as well as less problems
Er, if you think that building from source is somehow seperate from RPM than you have very little understanding of the packaging system you're dismissing.
Debians apt system at least keeps track of dependencies and test the stuff they release...
Huh? Comparing apt to RPM makes no sense and again shows very little knowledge of packaging systems and their function. rpm and dpkg are packaging systems. up2date, apt, and urpmi are frontends (which all work on top of RPM - one works on top of dpkg too) to index the packages and then resolve dependencies using these indexes. Again, comparing a front end to a packaging system makes little sense. Again, there are many other tools that to automatically resolve dependencies using Linux Standard Base (RPM) packages and these tools have existed for years (in up2date's case, since Red Hat 6).
According to its makers, it stops being Open Source if you exercise your rights under the AFPL and produce a LSB standard installation of WineX (also known as an RPM). Which is kinda sad - either its Open Source or it isn't.
I don't mind a good closed source app (I'm a WineX subsciber), its just that I think WineX isn't a half bad product but you'd be surprised by the amount of people (users new to Linux) who jump on IRC, try an incredibly broken CVS install and then think that WineX is an unstable pile of shit because CVS often is. I think this odd licensing is bad for Transgaming too.
As a system administrator, I also find the idea of installing software in any other method than RPM very bad practice.
This isn't the first time I've seen a segment of the Open Source/Free Software community turn on itself. What is it that causes these kinds of conflicts and mistrust? Are inflated egos allowed to remain because of their coding ability where in the business world they would have been let go?
DEP doesn't work on KDE, and I think that it would be reasonable to expect him to be paid fopr his Linux and Main work. What causes these kinds of attacks? Ego, yes - Dennis E Powell's was burnt because many KDE contributors disagreed with aspects of Israel's involvement in the middle east, and then took offence to his article in Linux and Main on April 7 labelling them antisemites for their views.
I have a fairly comprehensive article detailing what's been changed, with possible motivations for doing so, my own impressions, screenshots, and bug reports.
Judging by the constant up and down moderation of that comment as either underrated or offtopic, perhaps some context is in order: that article, which alleges that many KDE project members are somehow antisemitic for not supporting Israel, is why Andreas Pour won't speak to Linux and Main an specifically Denis E Powell - he believes DEP has a clear agenda against the project simply because some of its members don't agree with his political views.
I often hear someone defend Mozillas memory usage and speed (which I still find incredibly sluggish on an Athlon 900 / 640MB, a Duron 800 / 256MB machine, and a few others with noticable delays with any on screen widgets) by saying that Windows loads many of the components to support IE into their base OS.
This ignores the obvious argument that this only addresses launch times and rendering ignores the still noticably sluggish widgets. I wonder why somebody didn't just integrate gecko with these components? Create an ActiveX gecko to use instead of MSHTML.DLL, use standard Windows toolbars (because we all know how sucessful Xul was) and add in the other good features of Mozilla, like pop up blocking and security.
They don't want to be the authority for the kernel. They want to know what new features to add to the interface and the features.
I'm going to suggest they update the standard packaging system to the current version of RPM - the spec as it stands only recommends RPM version 3 as documented circa 1997, which is getting fairly long in the tooth now. The current edition is much more improved in terms of both security and reliability, and very few Linux distributiosn actually use the 3.0 edition.
In future, I'd also like to see the inclusion of one tool to to automatically fetch packages from a configurable list of sources with the dependencies matched in the LSB. I think its a reasonable thing to expect of most modern Linux distributions, which inevitabley include such a tool, but up2date, urpmi and apt-get all deal with Linux pakcages is slightly different ways, and standardization on these higher level tools would be useful.
Long term, I'd like to set a target date, at some point in the future (2004), where the LSB will plan to redefine RPM support as a genuine bona-fide implementation of RPM, not something which unpackages the installs into useless tarballs - yes, I mean alien. In the meantime, the few distributons that currently use such hacks to qualify as LSB compliant can update RPM with whatever improvements they're looking for.
I don't support th killing of innocent people by either terrorists from Israel or Palestine (look up a dictionary definition of the word and tell me there aren't groups on both sides that qualify). But it annoys the shit out of me that everyone who questions Israel is labelled as a racist. Locally, in Australia, the `Anti Defamation League' labelled the Wesley Mission as anti semites because they expressed sympathy with the Palestinians during a prayer around the beginning of the recent incursions. I'm sure I'd be labelled the same by saying any nation that publically supports torture disgusts me.
Slashdot Mirror
Your grammar is atrocious!
;)
you're grammar, friend.
CDE
:)
Cheers
dpkg has it's own technical advantages too, such as suggests and recommends
/sbin/init.d, /etc/rc.d/init.d, /usr/doc and other nonstandard file formats that aren't wrong (it may be more logical from acertain point of view to put intiscripts in sbin). You haven't responded to this.
I know that. You know I know that. But you asked about Red Hat advantages so that's what I gave you. I'm well aware dpkg has advantages over RPM in many ways and I've indicated such in my discussion of the topic. I've never alleged RPM wass a better all round solution, in fact, I've done the opposite - state that there's more to the argument than technical considerations.
Ignoring for the moment that IMAP4, ical, vcard and MAPI are all very different things,
MAPI or IMAP can be used as tranbsports for a variety information stored in a variety of formats, such as mbox, ical, vcard, or the Exchange message store.
I was aware that someone might misunderstand that analogy, so I added another -
As I said, the LSB does indeed accept compromises - there's no need to tell me this. As I've also said, that doesn't mean those compromises are OK.
To me, alien sounds like a fine way of installing LSB-conformant RPM packages.
They're not packages once alien is finished with them. `Package Management',as I see it, is based on relationships between packages which are destroyed once alien has been used on them.
The LSB have realised that trying to force all distributions to switch to the RPM format internally would be a fatal mistake
The LSB couldn't force anyone to do anything - it's an optional standard. Gentoo can exist, so can Slackware, so can LFS - they see advantages in not being LSB compliant and choose not to participate. More power to them. There's no practical way could `destroy the usefulness of innovative new distributions like Gentoo' by specifying a complete rpm implementation anymore than they stopped Slackware from existing by specifying SysV init.
However, for those who seek to use Linux in a larger scale environment or for something they consider important, some kind of structure is desired. Currently that structure is provided by a particular Linux distribution. People write and certify (closed source) apps for distros. People learn skills only applicable to distros. People write apps only applicable to distros. People compile and package applicable to distros. Use a less popular distro, and you'll either have to do a lot more work or miss out entirely.
The LSB is about writing apps for Linux. There's going to be
the toll on the Debian project would be enormous
The toll on packagers from the current situation is already enormous. Enhancing RPM to have a superset of current Linuxpakcage management features would and migrating the this RPM implementation would involve alot of short term work for a lot of long term gain when maintaining packages in future. Hell, if every (common) Linux distribution standardized on a package format you might even see people writing more apps that package nicely rather than using custom build jobs you have to play with in sed to install to the correct location.
As I've said, its no skin off my nose. If you've never been pissed off that a particular company hasn't packaged their proprietary app for your distro, then I suppose my point has no basis.
But chances are you have been, and it does.
If you mean technical advantages, there's a couple, notably ghost files, triggers, and a more verification system which provides more information about exactly how a file has changed.
.deb format, apart from the LSB's rubber stamp?
/etc/init.d and /usr/share/doc have over /sbin/init.d, /etc/rc.d/init.d and /usr/doc apart from the LSBs rubber stamp?
Practically, both can have policy, its the matter of the distribution to produce and stick to good policy. Some areas are handled better by some distros (eg, all the packages are signed, or they have standard naming conventions) but that's a higher level issue and neither RPM nor Dpkg based distros excel in that area.
and what advantages does RPM have over the
I'll be evil and answer with a question, thanks Alex.
What advantages does
What advantages do IMAP4, ical, and vcard have over MAPI, apart from the IETF's rubber stamp?
The lack of standardization amongst Linux distributions and compromises that result (like labelling alien an aceptable implementation or RPM) have a greater negative effect on Linux adoption than a few minor technicaladvantages that various formats have over the other. More Linux adoption means hardware works, upstream providers can support your operating system, and a bunch of other stuff.
Its the Linux packaging method, created by Red Hat. See http://www.rpm.org.
;)
Now if only Debian would start using it (and improving it in any area they see lacking) instead of supporting the LSB by turning packages into dumb archives with alien.
1. Install a distro that actually cares about fonts. Eg, Red Hat 8.
Congratulations, you're finished.
That really depends on your definition of XML and human readable.
That's a very good point. If your definition of human readable is illogical and bizarre, that could indeed be XML. Or a pair of pants.
One mitigating factor: codeweavers do built in a protection against executable attachments in their winex product.
My home machine is an Athlon 900 with 640MB of RAM. My work machine is a Duron 800 with 384MB.
I've yet to see a version of Mozilla, up to an including 1.1 and Phoeniz 0.2, whose UI performs adequately. Gecko is fine, Galeon is great, XUL is not. Why is Mozilla so slow, and if I can't stand having mymenus lagbehind my mouse clicks, why should I expect my users to?
Yes, that's a serious question.
but is wired deep into the bowels of the OS, using interfaces not available for other apps
Otherwise, this would exist.
Oh wait. It does.
Don Becker does indeed ssom to have done some great work, especially on network drivers and clustering.
I can't seem to find any information on his work regarding `making Linux usable' that's mentioned in the byline of this story. Am I missing something, or is that part of the intro a little bit confused (maybe Slashdot has a different definition of usability from the rest of the computing world)?
It either:
a) lives on in its music
b) Works at the local 7/11 with Elvis and Osama
he: Hey babe, wanna watch a movie ??
she: sure
he: wait till i boot the player
she: ??????
he: here we go...
she: is it me, or is it getting hotter in here??
he: so take off all your clothes!
she: I am getting so hot, I'm going to take my clothes off!
I find your post quite typical of /. myself: loud, rude, lacking supporting arguments, and illustrating a fundamental misunderstanding of your topic.
As a system administrator (among other things), I find that using RPMs are the sure path to unmaintainability and broken systems.
Why not provide some supporting arguments? I'll go first: standardized install, uninstall, querying, verification, GPG signing, and repeatability, LSB compliance.
Considering that it is *easier* to build from source as well as less problems
Er, if you think that building from source is somehow seperate from RPM than you have very little understanding of the packaging system you're dismissing.
Debians apt system at least keeps track of dependencies and test the stuff they release...
Huh? Comparing apt to RPM makes no sense and again shows very little knowledge of packaging systems and their function. rpm and dpkg are packaging systems. up2date, apt, and urpmi are frontends (which all work on top of RPM - one works on top of dpkg too) to index the packages and then resolve dependencies using these indexes. Again, comparing a front end to a packaging system makes little sense. Again, there are many other tools that to automatically resolve dependencies using Linux Standard Base (RPM) packages and these tools have existed for years (in up2date's case, since Red Hat 6).
According to its makers, it stops being Open Source if you exercise your rights under the AFPL and produce a LSB standard installation of WineX (also known as an RPM). Which is kinda sad - either its Open Source or it isn't.
I don't mind a good closed source app (I'm a WineX subsciber), its just that I think WineX isn't a half bad product but you'd be surprised by the amount of people (users new to Linux) who jump on IRC, try an incredibly broken CVS install and then think that WineX is an unstable pile of shit because CVS often is. I think this odd licensing is bad for Transgaming too.
As a system administrator, I also find the idea of installing software in any other method than RPM very bad practice.
This isn't the first time I've seen a segment of the Open Source/Free Software community turn on itself. What is it that causes these kinds of conflicts and mistrust? Are inflated egos allowed to remain because of their coding ability where in the business world they would have been let go?
DEP doesn't work on KDE, and I think that it would be reasonable to expect him to be paid fopr his Linux and Main work. What causes these kinds of attacks? Ego, yes - Dennis E Powell's was burnt because many KDE contributors disagreed with aspects of Israel's involvement in the middle east, and then took offence to his article in Linux and Main on April 7 labelling them antisemites for their views.
(Then again, I probably do as much preaching about Linux as he does about God - maybe we should get it declared a religion and get tax-free status...)
;)
Hey, it worked for Scientology
Actually, the ActiveX plugin someone mentioned above seem to do the trick. KMeleon is currently stagnant and doesn't work with recent gecko releases.
It's high time people stopped complaining about what people do in their spare time and give away for free.
Then its also high time those developers stopped telling new users that they provide usable desktops built for real people.
I have a fairly comprehensive article detailing what's been changed, with possible motivations for doing so, my own impressions, screenshots, and bug reports.
Read it here
Judging by the constant up and down moderation of that comment as either underrated or offtopic, perhaps some context is in order: that article, which alleges that many KDE project members are somehow antisemitic for not supporting Israel, is why Andreas Pour won't speak to Linux and Main an specifically Denis E Powell - he believes DEP has a clear agenda against the project simply because some of its members don't agree with his political views.
Can be seen here.
I often hear someone defend Mozillas memory usage and speed (which I still find incredibly sluggish on an Athlon 900 / 640MB, a Duron 800 / 256MB machine, and a few others with noticable delays with any on screen widgets) by saying that Windows loads many of the components to support IE into their base OS.
This ignores the obvious argument that this only addresses launch times and rendering ignores the still noticably sluggish widgets. I wonder why somebody didn't just integrate gecko with these components? Create an ActiveX gecko to use instead of MSHTML.DLL, use standard Windows toolbars (because we all know how sucessful Xul was) and add in the other good features of Mozilla, like pop up blocking and security.
They don't want to be the authority for the kernel. They want to know what new features to add to the interface and the features.
I'm going to suggest they update the standard packaging system to the current version of RPM - the spec as it stands only recommends RPM version 3 as documented circa 1997, which is getting fairly long in the tooth now. The current edition is much more improved in terms of both security and reliability, and very few Linux distributiosn actually use the 3.0 edition.
In future, I'd also like to see the inclusion of one tool to to automatically fetch packages from a configurable list of sources with the dependencies matched in the LSB. I think its a reasonable thing to expect of most modern Linux distributions, which inevitabley include such a tool, but up2date, urpmi and apt-get all deal with Linux pakcages is slightly different ways, and standardization on these higher level tools would be useful.
Long term, I'd like to set a target date, at some point in the future (2004), where the LSB will plan to redefine RPM support as a genuine bona-fide implementation of RPM, not something which unpackages the installs into useless tarballs - yes, I mean alien. In the meantime, the few distributons that currently use such hacks to qualify as LSB compliant can update RPM with whatever improvements they're looking for.
The discussion.
I don't support th killing of innocent people by either terrorists from Israel or Palestine (look up a dictionary definition of the word and tell me there aren't groups on both sides that qualify). But it annoys the shit out of me that everyone who questions Israel is labelled as a racist. Locally, in Australia, the `Anti Defamation League' labelled the Wesley Mission as anti semites because they expressed sympathy with the Palestinians during a prayer around the beginning of the recent incursions. I'm sure I'd be labelled the same by saying any nation that publically supports torture disgusts me.