Two Solitudes: Desktop GNU/Linux and Server GNU/Linux

I find it amusing when the same human beings who rationally promote GNU/Linux on the server get hung up on small details when the subject changes to the Desktop. Sam Varghese points this out:

“Both Weber and Behlert were at pains to point out that the SUSE enterprise desktop would cost less, provide inter-operability, ease of use, desktop security and ease of management far in excess of other operating systems. But still, they said, it would be incorrect to position SLED as a full replacement for Windows.

Asked whether this was due to the licensing agreement which Novell had signed in 2006 with Microsoft, both replied in the negative.”

see But what happened to the desktop?.

The fact is the same characteristics that make GNU/Linux work on the server are available on the desktop and they are more than sufficient to choose GNU/Linux desktops. One does not “need” a particular M$-only application any more than the world needs only one brand of car. We need generic software for IT and we should shop around. Many can and do work without M$’s OS on the desktop. The performance, reliability and security are great as is the low cost. What’s not to love about IT based on GNU/Linux?

In fact, RedHat which is doing very well on the server has lots of clients using GNU/Linux and OEMs are shipping millions of units. They just are not as tied to M$’s apron strings as Suse.

RedHat:
“What do Vanderbilt University, the University of Stirling in Scotland, and Yuba K12 School District in Northern California have in common?

Along with thousands of other higher-education institutions and K12 school districts globally, they’ve chosen Red Hat® to enable their IT infrastructures and encourage students to learn in new and exciting ways.

Desktop or server. To teach. Learn. Manage networks. Or develop and conduct research. Red Hat can empower you with affordable, reliable open source technology.”

The quality of FLOSS is superior to non-FREE software in many ways but measures of cost and quality are undeniable:
“The democratization of innovation has also demonstrated a remarkable solution to the problem of The Mythical Man Month, thereby transcending the limits of conventional industrialization. For example, sourceforge.net is an open source development resource that hosts over 180,000 projects and has more than 1.9M registered users as of December 2008. Extrapolating from the extensive FLOSS (Free, Libre, and Open Source Software) survey of 2002 (and updated in 2005) there were over 490,000 sourceforge.net developers in 2006 [when the thesis of this section was first developed–tiemann] who spend more than 10 hours a week or more tending their open source projects —an aggregate effort of some 5 million person-hours per week. The three top reasons they list for their involvement is:
1. Because it’s fun
2. Because it improves their skills
3. Because it is good for society
Note that this does not include Linux developers (who use kernel.org, not sourceforge.net), nor Apache, nor the GNU project, nor many of the other larger and more heavily commercialized open source projects. To put these 5 million joy-filled person-hours per week into perspective (again, this does not include Linux, Apache, GNU, or many of the other “large” projects), let’s look at the productivity potential of the most successful proprietary software company, Microsoft, in two ways (using numbers that were contemporaneous with the FLOSS survey data, October 2006):
1. If all 61,000 employees wrote code, they would have to work over 80 hours/week
2. If Microsoft’s $6.6B/year R&D budget were spent on programmers averaging just $25/hour, they could pay for about 5 million person-hours of work per week
Thus, the sourceforge.net website has equaled or exceeded Microsoft’s productive potential using a social, not an industrial model. When we consider all the open source developers not included in the sourceforge.net numbers (numbers that are increasing exponentially), we see the clear emergence of a new software production capacity entirely outside the conventional limits of the industrial model. Moreover, we find precisely the kind of improvements that Deming would have predicted by taking a transformative approach: according to findings published by Coverity, typical proprietary software has a defect density of 20-30 defects per 1,000 lines of code (KLOC), a number relatively unchanged since the 1960s. When they measured the quality of the Linux kernel (and later, other open source software) they found the following results:
2004: 985 defects in 5.7 MLOC of Linux kernel source code, or 99.3% lower defect density than average (compare to 114,000 to 171,000 defects in same amount of code)
2005: While the Linux kernel grew 4.7% in overall code size, defect density decreased by 2.2%. Moreover, 100% of all “serious” defects identified were fixed within 6
months
2006: The survey was expanded to entire LAMP stack and an additional 32 OSS programs. No correlation found between size and defect density, implying OSS development methodology is not limited by scale (nor restricted to just Linux developers)

What the top industrialists could not achieve with proprietary software and financial capital, free software has demonstrated with community development and intellectual capital.

That’s not too different from my oft-repeated expression, “FLOSS is a cooperative project of the world.”. The world can and does produce its own software for server and desktop. Not using GNU/Linux without any reason is questionable. Dependence on particular applications is not an excuse for poor design of IT systems. The mistake was often choosing to be locked-in to the non-FREE model of software development long ago.

see Michael Tiemann – VP RedHat How Open Source Software Can Save the ICT Industry One Trillion Dollars per Year

Ignoring those kinds of savings on the desktop due to imaginary problems is foolish. RedHat and others have a hard time making money with GNU/Linux on the desktop because so many are locked in but that doesn’t mean the rest of us cannot save big money by using it. Canonical, Dell and many other OEMs do make money shipping PCs with GNU/Linux but we can install it ourselves customized as we wish at very little cost. I recommend Debian GNU/Linux on client or server because it works. GNU/Linux works for Munich, School Divsion 73 in BC, Canada, the city of Largo and Google on desktops and servers. Why not you?

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology and tagged , . Bookmark the permalink.

81 Responses to Two Solitudes: Desktop GNU/Linux and Server GNU/Linux

  1. “Yet you ban 3 GNU/Linux evangelists and a white supremacist for expressing their views.”

    No, I banned your sock puppet accounts 3 times. White supremacy is way off topic and the other two were mirror accounts posting spam. One signed up and continued posting after I banned the first for posting a link to a no-longer-available YouTube video several hundred times.

    Notice, Adam, that I haven’t banned your actual account at all though.

  2. oiaohm says:

    That Exploit Guy
    –Bytecode engines and sandboxed execution are two completely different things. Bytecode can be executed within a sandbox. Likewise, a sandbox is not limited to isolate bytecode execution. Why not try and read up on the subject before defending your failed argument?–
    I said bytecode engine. Not bytecode execution. There is a big difference. A bytecode engine and process the bytecode without really running it. Sees the code paths.
    http://blog.clamav.net/2011/11/bytecode-signatures-for-polymorphic.html
    –heuristic engine (emulator)– in Kespersky is in the same class as the clamav bytecode engine. This stuff Scary enough its like saying a lint or gcc runs the programs because it produces a flow map. Of course this is not true. The heuristic engine in kaspersky never really runs the program.

    Basically there is more than 1 emulator define.

    That Exploit Guy
    –Real-time protection of the file system against malicious code: intercept and analyze attempts to access files, disinfect and delete infected objects–
    Word there is file system you are miss reading. That is not run-time heuristics

    That is the kernel hook that does not monitor execution or start a sandbox.

    http://blog.clamav.net/2011/11/bytecode-signatures-for-polymorphic.html

    The emulated is basically a de-complier designed particular to detect possible hostile actions.

    That Exploit Guy the reality is that on Linux kaspersky does not check ram in either version you are pointing to at runtime. Kespersky anti-virus on Linux does not help you one bit against a in ram attack. You are depending on other sections of Linux to handle that.

    Problem here you don’t know the topic That Exploit Guy so when you read emulator you think qemu, sandboxes…… Anti-viruses have something else they call an emulator as well. That is basically an very advanced de-complier that can map out the flow of the program. This is how come the heuristics will trigger before you have really run the program in any form. Why because a new varent of a malware could include the same code as old except now have a new feature that can smash its way out of a sandbox of the anti-virus this is why the anti-virus requires this special form of emulator. If you really run it you can kill you system by by.

    You would put the anti-virus emulator in the same class as spice circuit testing software. Its never real.

    That Exploit Guy part of security is know what each product protects. Kespersky on Linux its filesystem accesses and integrates into samba that is it.

    The heuristic solution is identical between all of Kespersky current Linux products. The runtime monitoring is identical as well. The difference between the products are skin level.

    –Let me tell you two things that will definitely happen when you have a piece of malware that is on disk but not in memory: 1 Jack.2. Squat.–

    That is exactly what Kespersky is trying to make sure happens.

    Kespersky does not help at all with a in ram buffer overflow or equal attack against Linux. You other sandboxing solutions like selinux come into play there.

  3. oldman says:

    “Yet you ban 3 GNU/Linux evangelists and a white supremacist for expressing their views.

    The raceguy was banned for posting off topic – a fairly good reason IMHO plenty of other sites use.

    As far as the other “evangelists” please name them.

  4. Adam King wrote, “We are the 1%”.

    Don’t believe that. We are at least the 3% 🙂 I think we are more than 10% but under-counted.

  5. Adam King says:

    “Haha, yes, keep telling yourself that Microsoft invests millions of dollars into convincing 1% of the software market to switch to Windows. Meanwhile, it’s
    OSX eating Windows’ lunch right now, not any linux desktop variation.”

    We are the 1% and we will not give up until M$ sits beside Hitler in the panthion of villains.

  6. Adam King says:

    “You mean like how I’ve refused to ban you despite your repeated, obvious, drive-by tolling on TM Repository?”

    Yet you ban 3 GNU/Linux evangelists and a white supremacist for expressing their views.

  7. That Exploit Guy says:

    @Robert Pogson

    ‘I think you don’t understand malware. It gets into a system any way it can and does what it wants one way or another.’

    I hope you won’t too start talking about bytecode being sandboxes. That would just be irritating.

    Let me tell you two things that will definitely happen when you have a piece of malware that is on disk but not in memory:

    1. Jack.
    2. Squat.

    I hope I don’t need to elaborate on either of them given your obviously “40 years of observations” about IT and what not.

    Just to end this once and for all, this is the description of what Oiaohm claims doesn’t have:

    http://support.kaspersky.com/faq/?qid=208283750

    ‘Real-time protection of the file system against malicious code: intercept and analyze attempts to access files, disinfect and delete infected objects’

    Seriously, stop defending him.

  8. That exploit guy wrote, “the run-time heuristic analyser, however you slice it or dice it, is inherently useless for a file server?”

    I think you don’t understand malware. It gets into a system any way it can and does what it wants one way or another. It does not matter whether it is a file server or a workstation. There are just more/different ways to get into one or the other.
    “New heuristic technologies combined with traditional signature-based malware detection methods help to dramatically improve the efficiency of malicious object detection and ensure proactive protection against new malicious programs.”
    see http://www.kaspersky.ca/products-services/business-security/anti-virus-for-linux-file-server

    “Enhanced heuristic analyzer
    Traditional signature scanning is supplemented with the improved heuristic analyzer. This helps to detect currently unknown threats which are similar to those that have already been detected, and significantly strengthens the overall protection level.”

    see http://www.kaspersky.ca/products-services/business-security/endpoint-security-linux

  9. That Exploit Guy says:

    “I cringe”, not “I cringle”.

    @oiaohm

    ‘Key word emulates. This is not run-time analyser.

    ‘This is a bytecode processing engine. You don’t load a possible hostile and let the cpu at it before you scan it.’

    I don’t want to be blunt, but in all honesty the time and patience I have for reading this kind of day-dreams are quite limited. ‘Bytecode’? Do you even know what that word means?

    Bytecode engines and sandboxed execution are two completely different things. Bytecode can be executed within a sandbox. Likewise, a sandbox is not limited to isolate bytecode execution. Why not try and read up on the subject before defending your failed argument?

  10. That Exploit Guy says:

    ‘That Exploit Guy Can you explain getting the number wrong for loopback in winsock. There is more than 1 bug.’

    I don’t know exactly what “fg” and “wtb” stand for, but it appears that the latter refers to the so-called “WineTestBot” in the Winers’ (heh) circles. The former, on the other hand, seems to be the initials of one Francois Gouget who prefers other hardware virtualisation software to VMWare, for whatever reason:

    http://www.winehq.org/pipermail/wine-devel/2012-June/096043.html

    I cringle at the mere mentioning of QEMU.

    Also, unlike those by “fg”, all tests performed by “wtb” simply passed with flying colours.

    I wonder why.

  11. oiaohm says:

    That Exploit Guy every sandbox can have a hole. This is anti-virus software use emulation systems to assess if something is hostile. Bytecode processing engines that are not doing the real operations on the cpu are many times harder to break out of in some not known way.

  12. oiaohm says:

    That Exploit Guy

    –That’s essentially the stock-and-standard, sandboxed test runs almost every piece of anti-virus software use for evaluating potential threats. Again, however you slice it and dice it, it is a run-time analyser.–

    In fact no I am not. You don’t know anti-viruses and you just proved it.

    –‘When scanning a program the analyzer emulates its execution and logs all its “suspicious” actions, e.g. opening/closing files, intercepting interruptions, etc. On the basis of these logs, a program can be recognized as possibly infected.–

    Key word emulates. This is not run-time analyser.

    This is a bytecode processing engine. You don’t load a possible hostile and let the cpu at it before you scan it.

    Reality you cannot read in this case That Exploit Guy. You missed a complete word. analyzer never runs the program. It just Analyzers it by byte code processing emulation.

    There are two analysers. One is runtime one is emulates. Emulates never ever runs the program to scan it so can scan binaries not native to your platform. Runtime analysers only scan what is native to your platform. Also cpu flaws and other things could see system exploited before Runtime analysers can detect.

    All anti-viruses have some from of emulated.

  13. oiaohm says:

    That Exploit Guy even if you compare those two you notice as you scroll down MSI goes nuts on one and not on the other. Yes the lower rights are detected and only the correct tests are run.

  14. That Exploit Guy says:

    @oiaohm

    ‘Reality Endpoint on Linux does not monitor programs running.’

    Now you are just plain making things up.

    From Kaspersky:

    ‘When scanning a program the analyzer emulates its execution and logs all its “suspicious” actions, e.g. opening/closing files, intercepting interruptions, etc. On the basis of these logs, a program can be recognized as possibly infected.’

    That’s essentially the stock-and-standard, sandboxed test runs almost every piece of anti-virus software use for evaluating potential threats. Again, however you slice it and dice it, it is a run-time analyser.

  15. oiaohm says:

    That Exploit Guy Can you explain getting the number wrong for loopback in winsock. There is more than 1 bug.

    That Exploit Guy funny part is 4 test machines there are not running with elevated privileges. So it should have worked. Sorry invalid written test is not going to cut it.

  16. That Exploit Guy says:

    @oiaohm

    ‘No I did not mean direct. atl:registrar The MS toolkit path to edit registry.

    Microsoft nicely provides documentation to use it. And a few other places.’

    That’s part of the Active Template Library – not strictly speaking a constituant of the Windows API but more a development toolkit.

    I have serious doubts as to how to test such a thing proper.

    Also, the environments in which the error occured involves running the tests without elevated privileges. That simply adds to my suspicion that the tests are, in fact, poorly written.

  17. oiaohm says:

    That Exploit Guy

    –I still ain’t seeing anything ‘esoteric’ going on here or anything that involves local execution.–

    Reality Endpoint on Linux does not monitor programs running. This is what happens. Kernel reads file from disc anti-virus checks file. Kernel loads loader anti-virus checks loader. Loader runs program anti-virus does not see this bit. Running program is the Linux Security module problem.

    Really for someone who says they are an expliot guy you know bugger all about defeating anti-virus on Linux.

    The Kaspersky Anti-Virus under Linux does not watch local execution at all in any of the Kaspersky products. You are assigning a Windows feature of that product to Linux.

    That Exploit Guy reality here you don’t know the product.

  18. That Exploit Guy says:

    @oiaohm

    ‘reliably is not correct. Linux can by idiots have binfmt_misc set up to auto run windows .exe files with wine.’

    I am sorry but Wine is not what I consider reliable for the most part. Maybe you are religiously obliged to believe otherwise, but I ain’t.

    ‘Kaspersky Anti-Virus for Linux File Server works by hooking to the Linux kernel to catch disc io for HTTP, FTP, NFS and Novell NSS. It in fact does not integrate into any of those servers. It only integrates into SMB/Samba and the kernel.’

    Excuse me, but have you actually read the specs for SMB? You don’t appear to understand what it does or doesn’t, as far as I can tell.

    ‘The reality you have been tricked by the marketers.’

    No, but it certainly seems that you have next to no idea how those protocols are actually implemented in Linux.

    ‘There is one very simple answer. When you are not integrated into the server program itself and are only scanning where the kernel reads from disc.’

    So, open(), read(), write() and close() in terms of kernel API use.

    I still ain’t seeing anything ‘esoteric’ going on here or anything that involves local execution.

    None of the protocols in question involves local execution, by the way.

    Come back to me when you have a better story to tell.

  19. oiaohm says:

    That Exploit Guy
    –You mean adviapi32:registry? That’s the portion of the Windows API responsible for registry access and modification. If you mean anything else, then clearly the test suites should not be doing anything of such nature.–
    Funny so applications should not write to registry using it right?

    No I did not mean direct. atl:registrar The MS toolkit path to edit registry.
    http://msdn.microsoft.com/en-us/library/49tet6k7%28v=vs.80%29.aspx
    Microsoft nicely provides documentation to use it. And a few other places.

    Fun right it not stable. So you code a C++ or a .net program write to registry and have opps it ate my homework.

    faultrep:faultrep is funny. 2 machines did not log a fault report when requested by program.

    ws2_32:sock is down right warpped. Somehow it thinks that 0.1.0.0 is loopback instead of the standard 127.0.0.1 on 9 of the test machines.

    Now are you going to now try to tell me that the Windows ABI is not buggy it cannot even get what loopback for network is correct. I have never seen a default Linux install get that one wrong.

  20. That Exploit Guy says:

    @oiaohm

    ‘That Exploit Guy one of them includes entering data into registry does not process the same when it should.’

    You mean adviapi32:registry? That’s the portion of the Windows API responsible for registry access and modification. If you mean anything else, then clearly the test suites should not be doing anything of such nature.

    I don’t see no error for that particular part of the results, by the way.

  21. oiaohm says:

    That Exploit Guy same test testing same feature on different install of windows should return same result.

    Reality they don’t. Even for basic things like writing a value into registry.

    Claiming a test suite is defective in areas does not disprove at all tests showing miss alignment. Since a defective test should still fail uniformly.

    Quality of the test-suite does not come into it with a apples vs apple compare.

  22. oiaohm says:

    That Exploit Guy
    ==there is simply no valid reason to perform run-time analysis of any kind on exectuable files, especially ones that you cannot execute reliably in the local environment.==
    reliably is not correct. Linux can by idiots have binfmt_misc set up to auto run windows .exe files with wine.

    Kaspersky Anti-Virus for Linux File Server works by hooking to the Linux kernel to catch disc io for HTTP, FTP, NFS and Novell NSS. It in fact does not integrate into any of those servers. It only integrates into SMB/Samba and the kernel.

    http://support.kaspersky.com/linux_file80?level=2
    –Kaspersky Anti-Virus 8.0 for Linux File Servers provides antimalware protection for file servers and workstations running Linux and FreeBSD systems (both, 32 and 64 bit). The application combines functionality both of previous version of the Kaspersky Anti-Virus 5.7 for Linux File Server and of Kaspersky Anti-Virus 5.5 for SAMBA Servers, being able to intercept file access operations at two levels: a kernel level (kernel module) interceptor and a SAMBA interceptor;–

    Read the both descriptions.

    That Exploit Guy both uses the Kespersky kernel module and the SAMBA interceptor to integrate run-time analysis.

    The reality you have been tricked by the marketers.

    Yes its know to work with the following list of Linux servers. But any other server is also protected because all we do is monitor file-system operations.

    That Exploit Guy
    –there is simply no valid reason to perform run-time analysis of any kind on exectuable files, especially ones that you cannot execute reliably in the local environment.–
    There is one very simple answer. When you are not integrated into the server program itself and are only scanning where the kernel reads from disc.

    Remember Linux kernel does not know how to load dynamic binaries so uses a userspace program to link up .so files to a dynamic binary.

    Reality if you are reading a file to run it or reading a file to send across network it looks absolutely identical to the anti-virus.

    That Exploit Guy by the way Kaspersky on Linux does byte code processing on read/write not real background server running monitoring even with Endpoint. Linux kernel security does not like apps snooping on other apps memory that much.

  23. That Exploit Guy says:

    @oiaohm

    ‘That Exploit Guy Explain this one away.’

    Explain what away? I have already pointed out very clearly that these are results generated with not terribly reliable test suites devised by the same people that brought to you a compatibility shim that is not terribly good at what it does. What more do you need?

    It doesn’t matter if the test suites employed were the same or implemented differently by people from different planets. They are of the same poor quality – and that’s what matters.

  24. oiaohm says:

    That Exploit Guy Explain this one away.
    http://test.winehq.org/data/c0d2029560372974001f75acbbaebda54473a51a/index_Win7.html

    The testsuites are all the same. 14 installs of Windows 7.

    2 installs run the testsuite completely without a single issue.

    The remaining 12 are all over the shop. Like 9 perform locale identically 3 don’t.

    That Exploit Guy one of them includes entering data into registry does not process the same when it should.

    Even more fun some of these bugs are regional particular. So you make a app in the USA send it to Japan and it don’t work right even that it works perfect on your machine. Does not work because the machine in Japan has a few extra bugs.

    Linux Distributions don’t give you regional hell.

  25. That Exploit Guy says:

    @oiaohm

    ‘The first point is that wine testsuite is going to disregard a lot of errors since it only tests what wine has implemented not the complete windows ABI.’

    That’s great! We all know Wine is very good at doing what it’s meant to do, don’t we?

    Your argument has just taken a turn for the more ridiculous, I am afraid.

  26. That Exploit Guy says:

    ‘This is in fact wrong its not useless. Because if I am running some odd ball network protocol on the server I need exactly the same protection as running a desktop application.’

    Ugh… You just won’t back down from a failed argument, will ya?

    The product page for Kaspersky Anti-Virus for Linux File Server tells you right from the word “go” exactly what protocols it covers – namely, SMB, HTTP, FTP, NFS and Novell NSS (not to be confused with Network Security Service). There are no esoteric protocols, no local execution and thus no requirement for run-time heuristic analysis. Even if there were, that would imply running an application across two heterogenous operating systems via a network connection. No, we are talking about simple file servers, not terminal servers or application servers – there is simply no valid reason to perform run-time analysis of any kind on exectuable files, especially ones that you cannot execute reliably in the local environment.

  27. oiaohm says:

    That Exploit Guy you don’t need to test the complete windows ABI to see that the Stable ABI claim is false. That is not as stable as people try to claim.

    Yes Raymond Chen is truthful with windows that you need vendors to keep updating there programs as particular functions break or the programs will stop working. Same problem you run into with Linux.

  28. oiaohm says:

    That Exploit Guy
    –Most of the git commits were simply additional permutations of function parameters the should have been there right from the beginning. Some are rewrites that change the way the tests are done. Some are even bug fixes for the tests themselves. Can I honestly take the results from such tests with any kind of seriousness? I think not.–

    The first point is that wine testsuite is going to disregard a lot of errors since it only tests what wine has implemented not the complete windows ABI. This is included in when to write tests in the test production guide. Adding more permutations of functions line up to wine implementation expand.

    Testing what wine has implemented is very critical to wine to prevent regressions as the extras get implemented.

    Test suite for java and other systems get rewrite to sections all the time and have bug fixes applied this is normal for a testsuite. Just like programs test suites are not 100 percent bug free.

    Basically what you have just said is no test suite is worth a cracker. Since general operations of maintaining a test suite is extending and repairing it.

    That Exploit Guy
    –One must note that Raymond Chen also points out Windows does not always provide compatibility workarounds but instead advice users to acquire patches from the vendors responsible upon encounting certain applications.–
    Key point Mircosoft does not promise developers their programs will keep on running.

  29. That Exploit Guy says:

    @oiaohm

    ‘Particular programs require particular API bugs. So you are required to test to see if those bugs are their. If they are not particular programs will not run.’

    Certainly. One would recall the effort invested to make SimCity run on Windows 95, but, as Raymond Chen explains, that was for a time when then the Internet was an alien thing to most people (and for the fact that there was no easy way to work around a use-after-free bug). Today, API/ABI changes are mostly managed and made compatible with legacy applications via shims, if feasible. The bottom-line, however, remains that a bug is an unintentional behaviour of the interface, and you should not expect it to exist or persist even though it is or has been there. One must note that Raymond Chen also points out Windows does not always provide compatibility workarounds but instead advice users to acquire patches from the vendors responsible upon encounting certain applications. There is a point where emulating the buggy behaviours become infeasible, and when that point is reached, the buggy behaviours must be dropped regardless of legacy compability.

    Now, of course, none of that has anything to do with what I say about WineHQ’s test suites. The unit tests appear to be so ill-defined and poorly implemented that I even doubt if they can be used reliably to test for expected behaviours – let alone bugs. Most of the git commits were simply additional permutations of function parameters the should have been there right from the beginning. Some are rewrites that change the way the tests are done. Some are even bug fixes for the tests themselves. Can I honestly take the results from such tests with any kind of seriousness? I think not.

  30. oiaohm says:

    http://support2.kaspersky.com/2703 << Read this this is the Kaspersky Anti-Virus for Linux File Servers new features.
    –Enhanced Anti-Virus Protection:
    new Kaspersky Anti-Virus Engine;
    heuristic engine (emulator) included;
    support for archive contents curing;–
    Hell yes the it includes the heuristic engine. Same one.

    That Exploit Guy
    –Does that undermine the fact the run-time heuristic analyser, however you slice it or dice it, is inherently useless for a file server?–

    This is in fact wrong its not useless. Because if I am running some odd ball network protocol on the server I need exactly the same protection as running a desktop application.

    heuristic analyser allows you to detect what is not known to the signatures. Without it the anti-virus is pretty much useless.

    Also if I am running something like http://www.zentyal.org that includes a desktop and is running thin terminals you guess what you need it all.

    The on access scanners are identical and that is the only run-time scanners the products contains.

    –http://support.kaspersky.com/faq/?qid=208283750–
    Read
    –Kaspersky Endpoint Security 8 for Linux is designed for protecting workstations running Linux (32 / 64 bit) operating systems against malware.

    The application combines the functions of the last versions of two products – Kaspersky Anti-Virus 5.7 for Linux Workstation and Kaspersky Anti-Virus 5.5 for SAMBA Servers. It intercepts file operations on two levels: kernel level (as a kernel module) and SAMBA level. The application has a local interface for GNOME and KDE, and can be remotely managed via Kaspersky Administration Kit. —

    Notice something. I know the exact error you are making. That Exploit Guy you are slightly out of date. Before endpoint when it was Kaspersky Anti-Virus 5.7 for Linux Workstation and Kaspersky Anti-Virus 5.5 for SAMBA Servers. The split you are talking about existed. Has not existed since its been called endpoint.

    Since the merge that makes endpoint there is no difference between the two products scanning engines.

    For fileservers no graphical is option. SAMBA Servers edition also could be run without graphical. Endpoint version keeps that feature. Prior workstation did not contain that feature.

    The fact the workstation edition was designed to protect only Linux desktops.

    Endpoint is designed to protect servers and desktops as one.

    Yes send a query to a Kaspersky customer rep get competent one ask how to use Endpoint to replace Samba Server edition. Find out it can be run on servers without graphic no problems.

    Endpoint is almost everything. The for servers is cut back edition. Slightly cheaper.

  31. oldman says:

    “TM Repository its funny you want me to provide you with references supporting statement. Yet anyone can take out a TM or a FUD thing on your site and not have a single valid document backing their accusation.”

    This is not about TMR it is about your putting your money where your mouth is. And since you are normally not so shy about providing references to your points, one would think that doing so now would not be an issue.

    So cut the crap and provide the reference Hamster.

  32. That Exploit Guy says:

    @oiaohm

    ‘That Exploit Guy. The heuristic analyzer checks out windows PE and other windows binary formats as well as Linux ELF on Linux.’

    You misread. The cited page concerns Kaspersky Anti-Virus for Linux Server, not Kaspersky Endpoint Security for Linux.

    They are two different products.

    ‘That Exploit Guy the Kaspersky heuristic analyser is not in fact limited to runtime.’

    Does that undermine the fact the run-time heuristic analyser, however you slice it or dice it, is inherently useless for a file server?

    I don’t think so.

    Face it – Kaspersky Endpoint Security for Linux is a product designed to protect against security threats targeting Linux workstations. Would you like me to send a query to a Kaspersky customer rep and let him or her repeat this same very fact to you in writing?

    ‘Read and say sorry now The Exploit Guy.’

    How about no, my pompus fellow Internet user?

  33. oiaohm says:

    That Exploit Guy
    –One question that comes immediately to my mind is how they the measured “success” and “failure” in each and every of those unit tests. In other words, do they have a fixed set of criteria for each unit? Do they make sure that their unit tests don’t rely on API bugs to achieves certain results? And, most importantly, do they make sure their own tests are not, in fact, buggy and error-prone?–

    Particular programs require particular API bugs. So you are required to test to see if those bugs are their. If they are not particular programs will not run.

    http://www.winehq.org/docs/winedev-guide/testing
    That Exploit Guy there is a nice bit of formal documentation on the process. And the testbot you can go through and see the exact failure reason.

    Remember this is not some hick project. Code-weavers pays the maintainer a full time job. Everything has to go past him.

  34. oiaohm says:

    That Exploit Guy. The heuristic analyzer checks out windows PE and other windows binary formats as well as Linux ELF on Linux.

    http://support2.kaspersky.com/2703 read down notice Kaspersky Anti-Virus for Linux File Server also has the heuristic analyser.

    That Exploit Guy the Kespersky heuristic analyser is not in fact limited to runtime.

    http://usa.kaspersky.com/about-us/press-center/press-releases/kaspersky-lab-presents-kaspersky-endpoint-security-8-linux

    Read and say sorry now The Exploit Guy.

    Kaspersky Anti-Virus for Linux File Server difference one that it does not require a graphical in fact does not include it. You get Kaspersky Web Management and console with the for Linux File Server and with Kaspersky Endpoint Secuirty for Linux you get graphical and console. Otherwise its exactly the same anti-virus with the exactly the same functionality. If you are using something with a graphical it can be simpler to use the endpoint secuirty where you get console and graphical and no web management console. Lack of web management is not a major problem thinking if you install Kaspersky Administration Kit and push policy and pull stats back. Yes the one console to web rule them all.

    That Exploit Guy you could be a version out of date. Version 7.0 there was the split you were talking about. Version 8.0 and on the anti-virus is identical between the two Kaspersky products on Linux.

    There is no performance difference the difference is how you want to administrator it. Local GUI on the Linux box or remote web gui. If you say stuff GUI I use ssh and bash it makes no difference at all.

    Its always funny how many people don’t know crap about the anti-virus products they use and say two different products have different scanning engines when they are absolutely identical.

  35. That Exploit Guy says:

    ‘Along the top there NT4,2000,XP,2003,Vista,2008 and Win7. Those lines are not Wine running emulation mode.’

    http://wiki.winehq.org/ConformanceTests

    If I understand correctly, the results you have there were collected with the use of WineHQ’s own test suites. One question that comes immediately to my mind is how they the measured “success” and “failure” in each and every of those unit tests. In other words, do they have a fixed set of criteria for each unit? Do they make sure that their unit tests don’t rely on API bugs to achieves certain results? And, most importantly, do they make sure their own tests are not, in fact, buggy and error-prone? Based on some of the documented git commits, I believe the methodology employed by Wine developers is at best dubious.

  36. That Exploit Guy says:

    ‘iLia remember common usage of Linux is file storage.

    ‘Kaspersky Endpoint Security is Windows, Linux, OS X.’

    Excuse me, but I don’t think you understand what they mean by “Endpoint Security”. A file server, by definition, is not a network “endpoint” (i.e. a computer terminal).

    Also, Kaspersky offers two distinctively different products called “Kaspersky Anti-Virus for Linux File Server” and “Kaspersky Endpoint Security for Linux”. The latter also contains a run-time heuristic analyser that is inherently useless for a file server. Perhaps you should read the product manual before commenting on the functionalities?

    http://support.kaspersky.com/kes8linux/common?qid=208283109

  37. oiaohm says:

    TM Repository its funny you want me to provide you with references supporting statement. Yet anyone can take out a TM or a FUD thing on your site and not have a single valid document backing their accusation.

    This is why TM Repository is a pack of Trolls.

  38. oiaohm says:

    TM Repository
    http://www.tmrepository.com/trademarks/windowsisbuggy/
    while you at it delete this one as well person again has presented no facts proving me.
    http://test.winehq.org/data/c0d2029560372974001f75acbbaebda54473a51a/

    Along the top there NT4,2000,XP,2003,Vista,2008 and Win7. Those lines are not Wine running emulation mode. That is the wine test suite being run on Windows of that version. So wine has bug for bug compatibility. Wine cannot code by the MSDN because its not correct in many places.

    This is what makes some third party programs unstable on windows because they followed the MSDN and there is some Conner case bug that causes there program to fail. Issue is other Windows coders find these bugs and exploit them so there program runs faster. Result bug stuck for evermore.

    I never said that Linux world could not copy it. Process of coping it you find the bugs.

    TM Repository Dr Loser is the one lieing here. Not me. Your site is filled with lies who say things I don’t say.

    Wine first goal was to make a Windows ABI that was bug free. This had to be given up on and copy Windows bugs so applications could run. This is recorded in Wine Official history by the way if you look it up TM Repository.

    TM Repository somehow I am not lie teller you claim me to be. People on your site do deformation against me state things I have never said.

    This is why I beat you into the ground. I am not the lie teller you think me to be. The people at TM Repository get to place a TM without having to source references or do proper research if the claim is valid or not.

    I could keep on going with TM’s targeted at me that are simply bogus. Really start doing a audit TM Repository of your FUD and TM’s. Lot are bogus. Or were correct and are now bogus because time moves on.

  39. oiaohm says:

    iLia remember common usage of Linux is file storage.

    Kaspersky Endpoint Security is Windows, Linux, OS X. A Linux installed server point can be looking after Windows and OS X clients as well. So you deploy a Linux small business server distribution you have Kespersky on the windows clients you install the Kaspersky management server on the Linux box and everything works.

    Linux does in fact support real-time scanning. The majority that Kaspersky is looking for on Linux is Windows and OS X programs not items that infect Linux.

    iLia clamav is not how to detect infection on Linux. By the time anti-virus normally works its way too late. You have been infected for months.

    If you are worried about infection you use a HIDS/NIDS solutions and rootkit hunters.

  40. That Exploit Guy says:

    @oiaohm

    ‘Lets take you I have caught you claiming Windows has a Stable ABI. What is complete bogus and is provable.’

    Perhaps I have missed your explanation somewhere. Care to repeat it?

    ‘This is the problem lot of your TM’s at me are bogus. The reality is poorly written business software fails to run on compatibility mode on IE 9.’

    Excuse me, but I don’t recall Javascript or HTML being part of any ABI in Windows. Again, care to elaborate on that?

  41. oiaohm says:

    TM Repository
    –You realize, the TM’s are all based on things the FOSS community has said, right? You realize several of the TMs are based on things you have said specifically. You’ve been caught several times in your own lies by people posting links to those TM’s you inspired.–
    Really lot of those TM’s are in fact bogus. I am not perfect I have had a few errors.

    Lets take you I have caught you claiming Windows has a Stable ABI. What is complete bogus and is provable. Yet you have not taken a TM against yourself or any of your others for the mistake.

    TM Repository really that is the funny one.
    Like this so called lie?? http://tmrepository.com/trademarks/ie6willkillbill/

    This is the problem lot of your TM’s at me are bogus. The reality is poorly written business software fails to run on compatibility mode on IE 9.

    Yes I have made a few mistakes. But most of the TM’s about me are in fact bogus and provable so if you do some research. TM Repository most of the support is that I said it. So they are attacking the person not the information with links and information.

  42. That Exploit Guy says:

    “or use ‘&#42;’ instead of ‘*'”

  43. That Exploit Guy says:

    @Robert Pogson

    ‘* = Linux, 1 hit’

    Either you must have a very twisted sense of humour, or the ‘Google’ you are talking about is just a bit different to mine.

    ‘Ubuntu won’t hibernate’ – first result.

    ‘Ubuntu 9.10 won’t enter suspend/hibernate’ – third result.

    ‘… my machine will suspend but won’t hibernate’ – fourth result.

    ‘… Why won’t my computer hibernate and shut down properly’ – ninth result.

    ‘mint 12 won’t suspend or hibernate’ – twelfth result.

    ‘Hibernate doesn’t function’ – thirteenth result.

    Really, how on earth did you manage to get 1 hit, of all possibilities?

    Also, you are using list item tags (<li>) without encapsulating them in a list (<ul> or <ol>). Since WordPress also uses <li …> to display comments, you are in effect messing up the part of the comment section following ‘* = that other OS, 7K hits’. Get rid of the <li> tags if this is what you have done manually in your comment, or use ‘*’ instead of ‘*’ for asterisks if your WordPress install happens to treat everything starting with a ‘*’ as a list item.

  44. iLia says:

    Download antivirus and update….

    Ok, on Windows no problem with it, but when I tried to install Clam Antivirus on Ubuntu it failed because some files cannot be downloaded from Ubuntu repository, so now I even don’t know if my Ubuntu is infected and spreading spam. So now I have to compile it from sources.

  45. iLia says:

    The performance, reliability and security are great as is the low cost.</b.

    reliability

    Ok, let’s see, on my Linux (Ubuntu)

    — sound is broken — SoundBlaster sound card don’t work

    — Flash is simply mute, no sound here

    — Huawey modem does not work properly, regularly loose connection and I have to:

    sudo killall modem-manager
    sudo killall NetworkManager

    (no restarting services does not work)

    security

    If Linux so secure and virus- and malwarefree why there are so many solutions to protect against these non existent threats, for example:
    http://www.kaspersky.com/anti-virus-linux-file-server and Kaspersky Endpoint Security for Linux

    If there is supply — there should be some demand for it.

  46. Chris Weig says:

    Re: hibernation.

    Like that’s a show stopper. I use Linux Mint 9 on the desktop and Mint 11 on my Dell 1525 Inspirion at the office, both hibernation features work just fine, perhaps you did not know what you were doing?

    Dougman, it is a show stopper. It is so much of a show stopper that Canonical deactivated it in Ubuntu 12.04 by default. Precisely because it doesn’t work reliably.

    Your pitiful insinuation claiming that esoteric knowledge is required to get hibernation to work in the first place (“[P]erhaps you did not know what you were doing?”) just shows how much the Cult of FLOSS is disconnected from reality.

  47. “One of the people on your blog seemed to think my comment about package compatibility between distros was incorrect”

    Feel free to sign up and respond. We welcome discussion and encourage differing opinion; That’s why I built the site after all. Hell, there’s plenty of disagreements between the most active members but things stay calm, rational, and group-think is avoided.

    Pog and oiaohm would disagree, but all they did was take a cursory glance at the About page, assumed it was an anti-linux site even though the page clearly states it isn’t, and dismiss it. Naturally, they both noticed the site was made using open source technologies (and hosted on a linux server) which they thought was ironic because, as usual, they didn’t read the About page.

  48. ssorbom says:

    Whoops, sorry, I meant to continue
    Mr. Pogson,
    I am sorry if I got too off topic, but it seemed relevant to some of the things that have been said so far.

  49. ssorbom says:

    TM repository:
    One of the people on your blog seemed to think my comment about package compatibility between distros was incorrect…
    I present an excerpt of a wiki post WRITTEN BY MARK SHUTTLEWORTH HIMSELF on the subject:

    “Some people might say “but I installed a Linspire package on Ubuntu, and it worked, so they must be compatible”. And yes, in many cases a binary package from Linspire or Debian will Just Work (TM) on Ubuntu. But this is “accidental compatibility”, not “certified binary compatibility”. Your Mileage May Vary (YMMV) is not the sort of certainty most people would accept, and can hardly be called “certified compatibility”. Many packages have very simple dependencies, and don’t really require specific versions of system libraries, and they may well Just Work. But if you look below the hood, at some level or other, you will find binary incompatibility in every significant derivative distribution, from Knoppix through Linspire and the DCC, with Ubuntu being no different.”

    https://wiki.ubuntu.com/MarkShuttleworth
    I apologize for not finding a suitable source sooner.
    Shuttleworth goes on to say this about non compatibility with Debian:
    “That’s what makes open source development, focused on the SOURCE CODE and collaboration around the code itself, more productive than proprietary development.”

    Not all are happy about this (including, I think, Ian Murdock: http://ianmurdock.com/debian/ubuntu-vs-debian-reprise/)

    I just thought you should know…

    Mr. Pogson,

  50. “Why do you treat this like its some kind of battle and why do you feel like you have to “win” so badly?”

    Get this through your head; I don’t. I don’t want anybody to “win”. I don’t want to see Linux die, I want to see it improve!

    That can’t happen when fanatically devoted folks like yourself keep defending flaws and being yes man, towing the company line no matter what. You hate Microsoft’s corporate culture, yet you wholly subscribe to the exact same thing when it comes to FOSS.

    THAT is what I have a problem with. A grown man so emotionally invested in a piece of software that he sees everyone as opponents rather than people being vocal about the condition of something they use. I’m interested in the social and psychological aspects because if a man like you were poorer or suffered some REAL hardships (instead of perceived software ones) then there’s a good chance you would become even more unstable and violent.

    (Plus, you make the rest of us Canadians look like right wing fundamentalists with your inability to yield)

  51. “In fact some of those who create TM’s on TM Repository and don’t post else where are Microsoft. How do you pick it the message on TM Repository matches Microsoft Anti-Linux documents released latter exactly. Either MS is plagiarism or you are infested.”

    That’s not true at all. Please present these documents! Please, show me the similarities.

    You realize, the TM’s are all based on things the FOSS community has said, right? You realize several of the TMs are based on things you have said specifically. You’ve been caught several times in your own lies by people posting links to those TM’s you inspired. That was the whole purpose of the site, to disseminate misinformation that LIARS such as yourself try to peddle.

    So once again, PLEASE PRESENT THESE DOCUMENTS YOU CLAIM MY SITE IS SOMEHOW INVOLVED WITH!

  52. TM wrote, “Why do you treat this like its some kind of battle and why do you feel like you have to “win” so badly?”

    Why do you treat this like its some kind of battle and why do you feel like you have to “win” so badly?

    In fact I don’t see this as much of a battle. */Linux is growing by leaps and bounds and M$ is stuck serving its slaves or vice-versa.

  53. TM wrote of that other OS hibernating, “Then why does it work on Windows?”

    Hmmm… Google search for “* won’t hibernate”:

  54. * = that other OS, 7K hits
  55. * = Linux, 1 hit
  56. It doesn’t seem that hibernation is much of an issue for anyone, so why is it an issue for TM? Perhaps TM should change his tune.