hi tomasz86
you and bristols are doing some
totally awesome mighty work!
huge kuddos and much much appreciated as well
a quick thought (and I have no idea if it might be possible) but I don't imagine too many people are running 2K with IE5, script 51, mdac253, dx7, mp6, etc. (although I prefer classic 6 with 9 codecs) ... if you're worried about serving them, why not do a simple R2preinstall (or whatever) to update their core components to a sensible system (?IE6sp1; runtimes (script 57, mdac 28sp1, VB6sp6, VC++05sp1, ... ); mp9, dx9c(std), codecs9) then ... maintain ONE patch list.
that alone will be hard enough work ... I have huge respect for that ... otherwise you burn out when you open up list management to hairball territory
back when, just to prove how thick I was, I did a manual frogwalk through "the Vast Hairball" i.e. MS patch domain (endless rabbit holes everywhere ... and then some ... and nowhere a mushroom to rest my head on ... alas!)
what I found was that no two
expert system builders and listkeepers, building nearly identical systems, will come up with the same patch list. It's the very definition of impossible hairball. I compared them all including Microsoft itself, and Shivlik, and and and
it was amazing ... here's my 2009 comparison, posted on this forum
Windows 2000 update lists compared. it's was such a hoot ... that it took months to frogwalk (I'm on dialup) and twice as long for my brain to fully recover! harhar!! (I haven't figured out smilies yet, but I'm sure there's a good one for that ... I'm still looking). In fact, I gave up on building a 2k system because of MS exhaustion, and went back to my old standard ... 98se ... it's still the greatest!! but there is some newer software ... sooo here I am building a 2k system ... I'm going to use terabyte's Image as my antivirus (in fact, I'd like to automate shutdown ... prehaps someone could help me ... on shutdown execute a script that gives me the option to defrag and shutdown, or shutdown and reboot to cd, whereupon terabyte linux would restore boot partition and then shutdown. terabyte will restore off of a batch, so the restore code would need to have the usual warnings ... is this the image file you want to restore? the hdd partition 1 to be restored? or do a test to see if a 40-char random filename for the image file matches a system file in root of the partition, with the same 40-char random filename, to uniquely match the image sourcefile with the target partition, etc.) and then shutdown.)
R2coreup would bring us user-consumers up to a set of core components that you think is the most manageable, from your perspective, listwise and patchbuildingwise, as regards your available energy, time and focus. The rest of us will follow your lead. The last thing we want is for you or bristols to try to satisfy the world (an empty set that), and then burn out. The two of you (and lots of others on msfn team) are listkeeping and patchbuilding a path through the MS security hariball that is beyond the talents and abilities of most of us. Focus will make the project stronger.
you and bristols have pretty much already started towards that end anyway ... on bristols list, who's going to HFSLIP a hairball of patches for script 51 when there's a link to 57 sitting right there? .. you with your rollups for ie6 and dx9 (MS made a mistake when they merged dx9 x86 with x64). Pack a rollup (easy huh? haha) for a few runtimes, script 57, vb6, c++, mdac, and you've already got R2coreup put to bed.
then you and bristols, and the rest of us, have just one list to look after. That will be hard enough. As shown by my 09 work above!! That was 11 listkeepers! Most pros, and I mean, pro as in earning their living that way ... all hugely different list results.
after R2coreup gets put to bed, its cousin, R2patchup, gets put to bed ... R2patchup would update the standardized core system, as defined by you and bristols, using MS-only patches ... (you are right to separate out .net and xml even ... daht net's a whole 'nuther hairball dar boiz).
once R2coreup and R2patchup are patted down, they become a standard 2K platform upon which to build the R2patchplus rollups. In your mindsets, you and bristols, and the rest of us, are already working towards that anyway.
For stability, and reputation, you could then add an R2hbrup. In fact, you've already done that! so that's put to bed already. Just a few pieces to go.
I think this is the important part ... don't commingle your or wildbill's work, et al, with MS ... when a stability issue pops up, if MS is separated, the issue can be quickly isolated ... was it an unofficial by wildbill, or you or bristols, or whoever, ... or was MS again, by patch or HFR, which can be isolated in a separate rollup? you and others on the msfn team don't need to totally commingle your reputations with MS. in fact, msfn contributors are preserving MS's reputation.
the R2patchplus rollups could then plugin the hugely helpful unofficials ... if a stability issue should crop up, it can be quickly isolated as MS or HBR ... or as unofficial and quickly fixed (IMHO the former is more likely).
It's just a thought ... I have such HUGE respect and admiration for the hard work that you and bristols (and everyone else on msfn) are doing.
Major kudos to you all ... just one molecule here ... harhar
edit add
I may be able to help with testing but I need more experience with hfslip (files in right dirs) ... I have an ASUS CUV4X (P3socket 370, 512M) with Matrox g400 AGP, AHA2940AU (cd's scanner), VIA642 controller card (for 48-bit HDD bios), a NIC, a USRobotics 5610, and a USB20 card ... I cold swap HDDs and 98se gives me no problem ... (my first builds of 2k are puking when I swap my data HDDs ... that'll have to be another post ... I'm just not quite there yet.) and I have a MSI P43neo-f3, core2 duo 7200, P43 northbridge, ICH10 southbridge (sata), with Intel drivers for 2k (and XP), MSI video 7300g-256meg, Realtec LAN and Audio (with Intel 2k drivers (and XP drivers), 2 gig ram, aha2940, pci modem, cold swappable data HDDs, ... blah blah
This post has been edited by Molecule: 04 July 2011 - 09:06 AM