Guessing something like BASIS is out of reach, guess OpenText has dropped the LiveLink brand and gone with using OpenText (customers had asked they the company didn't change its name to match its LiveLink branding.) OpenText was originally to commercially a natural language query parser, originally for the OOED, but the internet search provider to Yahoo before Google, and then I think it was AltaVista? ...
OpenText was losing too much money in trying to succeed at Internet Search, so it the company that produced what was its core LiveLink product....years later they acquired Information Dimensions, which Battelle, being a non-profit r&d think tank, had set up to commercialize its document management system, BASIS (Battelle Automated Search Information System). When I joined Information Dimensions, all of its servers still started with the letter 'b' for battelle...only one had an 'idi' prefix (idiusvr1 - IBM RS/6000 F50, it replaced a 390? that was named bibm6k) After OpenText, names were allowed to be more creative.....QA got to name their sun server, corona. They influenced the name of then new pair of HP servers, development server was named titanic, and the QA server was named iceberg. I was originally contract to hire as 'the' Unix developer for BASIS, at the time they were having issues introduced by the recent appearance of Unix servers with more than one CPU. (the software's roots were IBM mainframes....where it had implemented light-weight processes to improve file-i/o performance...but deep in the code, there was note that the mutual exclusion control had been simplified by assuming only one CPU.) They previous contract to hire was still there, but his proposed solution was replace mutual exclusion with code with assembler....at the time, meant code for POWER, SPARC, PA-RISC, ALPHA and MIPS. Since at the time Sun's (kernel) async-i/o was mainly for raw disks, and it generated excessive context switches to use its user space implementation which performed extremely poorly with non-page aligned (8k) i/o. At the time BASIS was still using a 2K page size. (only the Windows NT port had changed to a 4K page size.) So, I proposed a rewrite using Posix Threads. But, given the amount of change, it was decided to defer that fix to 8.3 (rather than the current pending release of 8.2.3...which needed to get done as it contained Y2K patches.) 8.3 also got large file support and move to 8K page size (and 8GB database files). Back then we had a 32-bit address pointer....a database could span 16 kernels, with 64 files under each kernel. There were four bits to indicate which sector of the 2K page, leaving our limit of 1048576 pages per file. (at the time you could store up to 2016GiB in a single database, one file per kernel as index....which switching to 8K pages allowed up to 16128GiB...) There were customers that were pushing the old limit. Solaris was why we went with 8K pages across all platforms. performance of non-page aligned async-i/o was acceptable before Solaris 8, but....our LFS version was also our shift from building on Solaris 2.6 for all, to building on Solaris 8.... we joked about going 8K for 8.... The build master was happy, it meant he went from a shared SparcStation 20 (bsvr4s) to a dev-only E250, QA had largely abandoned its SS10 (bsparc) and pushed onto the SS20, so dev's E250 (corona), with QA having their own E250 (phoenix?)...they later also got a V250 (phoebus), there was a 280R (jackeen) which was shared between dev/qa and helpdesk...as early OS releases had ABI bugs that prevented applications built on non-UltraSPARC III from working on it (so for early adopting customers, we had to build/qa/etc special builds on here. Helpdesk had named it...something to do with it being an Irish customer that had first brought us the issue. The 280R was our first rack mounted server, though had to use twist ties to secure it into rack...since our only rack was from Dell. (round peg into square hole problem....didn't know you could buy clip-in nuts until working in a datacenter with racks of all types (but now mostly all APC.....one with a contained hot aisle, and another without....in the latter we'd still burn out HPC nodes in one rack, requiring cardboard deflectors to keep the heat sneaking around ....eventually the HPC rack got evicted, yet it got shoved in our datacenter because administration didn't want to fund adding significant cooling to closets around campus. (ignoring that we needed more cooling in our datacenter, or space....and, there's already a walled off raised floor section due to not having enough cooling/power to use it.) After OpenText acquired Information Dimensions...they had a centralized IT structure, and in the field it was one IT person per 100 employees. So, the windows admin stayed (since the division had less than 100 people, he was also the IT person for some smaller 'condo' offices in Virgina and Maryland....as well as IT person for many of the tradeshows.), while the Unix Manager position moved to HQ (Waterloo, Ontario)....the person that had been in the position wasn't interested in uprooting his family for Canada. He opted to quit in protest and leave all the unix system administration to me. (I had tried to avoid having anything to do with it, because doing too much of it was the reason I was let go from my previous job....it cut into my billable hours...when I started I was doing 80+ hour weeks, of which only 35 hours could be billed. Years later they found they could use those unbilled hours for srtc's (science & research tax credits?)....which meant I finally had to pay me for my overtime when they let me go. After that half the company quit (and found jobs elsewhere, before I did.) The person that was supposed to be doing system administration wasn't a people person....made worse by being the only non-Engineer in an Engineering company (outside of office staff.) He was extremely pissed when years earlier he had written an integration function for graphing, and then had to prove that the it was integrating correctly. And, then I come along and write one for another application, and wasn't required to provide an extensive proof that it was correct. At the time, I was was still a member-in-training. I the time there was minimum 2 year MIT period before application to be come a Professional Engineer could be made. Due to NAFTA and alignment with the US, the minimum period was increased to 4 years. Since I was originally hired into a non-Engineering position, and there wasn't a Professional EE person in the company....I didn't get my Professional Engineer registration until I had been there 4.5 years. Had the Unix manager not quit in protest, I'm sure the development group would've found him a place (at least until we got laid off to meet profit shortfalls (before the crash....the previous CEO saw revenue grow by 20%, resulting in 20% profit growth...the new CEO promised 20% profit growth each year for the next 5. Problem was things started to slow and revenue only grew by 8%....leading to only an 8% profit growth. So, operations had to come up with the other 12% by laying off a 3rd of the company. (people got laid off to meet required profit growth, people got laid off to pay severance for, people got laid off so those that didn't got raises, and then more people got laid off to pay lease termination fees from all the empty office space. The BASIS development group got slashed down to 4.5 FTEs, the manager really wanted me to be one to stay, but got beat out by someone with a few months of seniority over me. Though I got a 5 month contract, to Xmas Eve, to do unix administration to wrap things up. One of the things I built during those last 5 months was a 'new' NIS slave server, on a near end of lease Dell Desktop and RedHat 7.3...... and its still running as of yesterday (10+ years)... As for BASIS, I was also responsible for its initial Linux port....lots of customers wanted it, though it's doesn't make it any cheaper, but I guess its the rest that counts. Sun used to be a big sponsor for Livelink user conferences, because they once said that for every dollar the customer spent on our software, Sun was getting $9. OTOH, at first job....we had originally developed our main app on Interactive Unix (SVR3.2) on PCs (386's and later 486's)....as they were the first to have Motif available, and marketing had decided that whatever we did, it must have a Motif interface. Later for price/performance reasons, we moved to HP snakes. Plus the president was anti-Sun. But, Intel hardware got faster and was much cheaper, so customers wanted a Linux version. The president was opposed to porting to a free OS, but customers still wanted something for PC hardware. Interactive had briefly come up with SVR4 for x86, but they got snatched up by Sun and its SVR4 product pulled from market....so that Solaris 2 for x86 could thrive. So, between being anti-Sun and anti-Free OS...he opted for Solaris 2 on x86 to satisfy the customers that wanted to use PC hardware. I got a Solaris 2.5.1 box...on a 200MHz Pentium Pro with a 1GB harddrive...forget how many megabytes of memory I got.... It was around the time we also got a real Internet connection (had gotten a dial-on-demand ISDN router that could bring up 1 or both B channels, though I had worked out that it wouldn't take much for a dedicated B to be cheaper than paying hourly rates....and eventually they went that way. One of the biggest things I got grief from the office staff was that the server with our accounting system (shared by the office manager and her assistant) didn't have Internet access. They had a second (older/slower) PC that did. Seemed reasonable... Since I was informed that I had been laid off by a strange phone call near the end of the day, and then escorted straight out of the building. I was to come back the next day after the rest of the employees were informed about my layoff, and they would have my stuff boxed up. Our business had primarily been defense contracts, had tried to diversify as defense spending was being cut....company eventually went out of business March 2001....had they lasted a little longer, I'm sure the story would be different.... But, while I was waiting for them to finish making boxes for my stuff (I accumulate a lot of books at work...though I think I'm going to abandon all of mine next time, especially since pretty much all of them are from O'Reilly & Associates, and a while back I had purchased eBook versions of everything. (wish I had thought of that with the stack of ORA books I won at LISA '12 (would've been one way to avoid the $100 overweight charge on my luggage, though the State did reimburse me for that charge.) Not sure where our office library has disappeared to though. But the office iPad 3 with LTE, that we got as an eBook reader is on my desk. I had asked about getting a Kindle (much cheaper) to use a eReader, but Kindles aren't something that are considered typical IT purchases.... Especially since it requires special approval to purchase anything from Amazon. (last June the power supply in my workstation died, the Dell 4 year support ended in March....since the replacement harddrives had been sourced from Amazon...original had a pair for Seagate ST2000L's ? vs the DM's....think that right, originally it had 7200RPM DM's which only had 1 year warranties, and when they failed we switched to 5400RPM DL's with 3 year warranties. The latter were sourced from Amazon. So, when the powersupply died, and my director/ciso said find something, and I'll approve it. I found something inexpensive but sufficient on Amazon....because it was Amazon it required approval higher than him. A few months later, I found a package from Amazon with the powersupply to get my workstation going again. (I wasn't totally stuck, since I still had my older Dell workstation...though it took some time to extend a number of my CFEngine policies to support Ubuntu to get that system to have what I needed. (somewhere during this time, the harddrive part of my 1TB fusion in my iMac, it took a few weeks to finally have internal repair come fix it....fortunantely, I was able to completely restore it from my time machine backup...before the 1TB drive I had been using died. (it had been making noises for months, to remind me that I had it stashed behind my drawers.) All the disks I've used for Time Machine at work have been personal disks. It is a 1.5TB drive that I had picked up from woot that has since replaced it. Wonder what I'll do when I get my Hackintosh built (tracking says the last, and important, part should arrive on Dec 26th.) UPS left the other parts on my doorstep today without knocking. Wondering again if it wouldn't have been easier to just buy a Mac Mini....but former manager had put me on to the idea of going with one, and originally, a CustoMac mATX....which if it didn't work out would make a nice FreeBSD server. or perhaps FreeNAS. Though my desktop is becoming less and less stable, and I suspect its the EOL nvidia card in it....since it does better when I'm not logged into it. (and it does plenty of things without me....like its my CFEngine policy server, BackupPC primary/only...since I haven't gotten to replacing power supply in the other box, and finally stole one too many drives from it to keep array on this system happy. Both arrays (zpools) had been done on 512 sector drives....so it'll be a pain when I migrate to 4K drives. On my other BackupPC server, I had tried replacing a mirror with a 4K one....I eventually had to create new array with new disk, copy over and then add old disk as its mirror....which lasted for a few months and then suffered an unrecoverable error.... and I let it go. Unlike the week I spent trying to recover a mirrored zpool on my this computer....before finally nuking it and restoring pretty much all of it from local BackupPC (since machine at other end of condo is down...) with help for CFEngine....restored my BackupPC setup. (since newer Perl's don't allow setuid scripts, to get BackupPC to run as backuppc, it runs out of public_html in the users's directory. But, the mirrored zpool that had gotten corrupted contained my home dirs, along with other things (like my entire Doctor Who collection...) That one was annoying/strange....while it had originally been a pair of 1.5TB drives, I had created the zpool as 4K...and when one drive died, I replaced it with a 3TB drive (WD Red) and then replaced the other....(the surviving 1.5TB disk is now part of the raidz pool of 1.5TB disks.) Strange how I keep reminiscing about my past lately.... perhaps because its nearing time for my annual self appraisal, and having had a meeting (that was imperative I have by the end of the week, a couple weeks ago) with HR and learning about FMLA and LTD....that maybe I will be going out into disability (SSDI) in the new year. Can't do everything I used to...and the plan is we'll definitely get rid of all Oracle hardware by July 1 (was supposed to have happened last July 1, CIO had said he had no idea where money for Oracle and Hitachi support was going to come from otherwise.... but so far not much had changed. (and things run fine when I'm not around for extended periods, thanks to CFEngine....now that I'm making it do more than copying files around. It had originally been setup to replace our old rdist based system....) And, have everything running in vBlock environment managed by Chef (they won't invite me into Chef so have no idea what's in there...but so far don't think there's anything actually running in vBlock.... But they've been churning out lots of dev VMs (some of which have lept into production ones...requiring them to be moved into a prod vlan so that the world can hit it. Which is what security planned (but new production F5 is far from complete to take anything....though its in production since we have one service on it....catfiles (using isilon's CIFS)...since our previous Windows file server was running 2003. catfiles went live the day before 2003 reached EOL (barely...since they had gotten he production F5 messed up and unusable while I had been away at the Mayo Clinic....though took me less than an hour to fix up, what others had spent a week on already.... Surprised it hasn't gotten messed up since.... Probably helps only other person that had access to the root/administrator accounts is gone (though one of these days, the other admins will get around to figuring out how generate and send me their GPG keys so they'll have access.... though they've been here more than 6 months now....and still haven't figured it out.... hmmm.... Not sure I like the next generation in system administration.... But, director had said clearly our attempts to hire traditional system administrators isn't work. (vraptor had been the last we had tried to hire...the closet we had come in doing searches in the last 3 or 4 years? Well, in the first attempt we did almost get a quarter of a level 1 admin, except we wanted at least a level 2 (replacing several level 3's that had quit)...but the quarter immediately turned to 0, and they eventually got disgusted with how we do things and went back to their old job (at another University.) He had quit a few times before, but I guess he finally couldn't be talked down....probably because one of the managers (the one that hired him, but didn't get any of his time) had quit. And others were out of the office, such as myself....had broken other foot (broke the right one a couple weeks before I was to fly out for Thanksgiving in 2010...) Technically it was just a left toe, since the break was of a bone on the other side of the joint compared to first time. Plus how I broke I was different....first time was from smashing foot into a first aid box, which I had walked past for years without incident. More recent I was maneuvering a large box down narrow hallway and bumped into a cart which had a heavy speaker on topple off and smash in to my left foot. While Birkenstocks don't provide much protection to prevent this, ER said to stay with them instead of issuing me a wooden shoe this time around. Oh, OpenText has its own document management product, since its the core of their product suite. When they acquired Information Dimensions, they had intended to absorb the thesaurus capability of our search parser. Except the divide between Oscript -- something like smalltalk or C#, at least that's how I looked at it when I picked up how to program and fix long standing bugs too quickly. I've always had a knack for fixing bugs, but not so good at new product development...which is the only thing that generates new revenue for companies.... And SLANG -- a Macro language built on FORTRAN 66 with system libraries originally written in assembler for IBM mainframes. It really bugged the the strict columns requirement for my brief encounter in write SLANG code. BASIS continues to be written in SLANG, because it sits on two levels of libraries....an interface layer and the system specific layer. interface layer handles the language interface. The SLANG 'compiler' used to be a tool that translated the Macro code into F66, which was then compiled with a FORTRAN compiler and linked with our libraries. VMS had a mix of C and assembler for its system specific layer, while Unix was purely C. Not sure about the NT version, since they had contracted out its port.... So at first we had Fortran and C compilers on Unix build servers. In later life....we switched to using f2c,...and then the source was adapted so that it became s2c. Hmm, since our old yahoo group of former IDIOTs (at the merger meeting, the CEO was writing out on the projector that we would be a merger of equals....IDI coming together with OT will turn us into IDIOT....which he then promptly scribbled out.....but the fact he made us into IDIOTs stuck.) has become a facebook group, I should ask if anybody knows what remains these days.... I actually thought the Yahoo group had died years ago, considering I own the group....Yahoo keeps locking me out of my account (though just realized that I'm using an anonymizing IM proxy, had set it up so work wouldn't know where I was when accessing work IM. Since that's how other's in my department opted to do it. (just as we have a 'private' (key hasn't been changed ever and the person that owns it doesn't work here anymore) FreeNode channel that we had set to communicate with each other....such as when we lost an entire row (and odd one) due to a single bad UPS...since the datacenter manager at the time got mad that we were making a mess cross connecting and only using less than 50% of his new UPSs. Especially useful, since campus had disappeared (in name only), because primary nameserver was in this odd row, but an admin had apparently decided that since secondary was on intel vs sparc....he'd put it in odd row as well (being backwards or something...though have since found it gone done backwards on SPARC as well. So, secondary was also down...so we vanished. Right now I know we are in different rows, and our main zone has off-campus secondaries. Though the on campus authoritative servers are all in the same datacenter. Probably won't change when its replaced with F5-GTM and (likely) Bluecat (the other contender is InfoBlox, while its an F5 partner, the committee seems to be leaning to Bluecat....) sometime between spring break and spring intersession. Though the primary driver is to replace/upgrade our current DHCP servers....which aside from configuration, hasn't been touched, patched, upgraded since 2006....its two Sun V240's running Solaris 9 and 3.x of ISC's DHCPd. Had offered to upgrade/replace them at times, have resisted us making any further changes after building it for them (networking, while its the unix group that manages DNS.) And, lives in a subnet that continues to be free from campus firewalls.... Most of the time when a Unix system is compromised, its one that is in this subnet. At the time there were only two new rows that we were migrating servers into. On 2015-12-20 18:08, john boris wrote: > Mike > Thanks. > > On Sunday, December 20, 2015, Michael Gurski <lo...@gurski.org> wrote: > > At work, we switched from Documentum to Reliance. It's browser based, has > access controls, versioning, can supposedly convert formats, and can require > quiz/acknowledgement that a user has read a specific document. With the sheer > number of SOPs, forms, plans, etc that we have (try being in a lab > environment without them, when a misstep could cost people's lives), it seems > OK. > > About the only legitimate complaint I have is that the interface implements > its own tabs, which breaks my personal workflow of opening every new link in > a tab when I need a lot of things open. > > No idea on cost, that's not my department. > On Dec 18, 2015 12:14, "john boris" <jbori...@gmail.com> wrote: > > Good Day to the list, > > Here at $WORK an old project was resurrected and dropped in my lap. The Execs > are looking for a Data Warehouse solution where they can store documents for > retrieval, provide access control and possibly live updating. As a part of > this system they are also looking for a Dashboard for certain functions. > > The Project leader understands that there isn't one product that is the end > all and figures this project will be resolved by using a few different > applications. > > So the goal of this email is to ask folks what product they are using in > their work environment (University, School District, Enterprise, Work Group) > as their document management system that has access controls. Even if you > rolled your own using Oracle, MySQL or some other database coupled with a Web > front end and some type of access control. > > I have looked at two products so far Knowledge Tree (which has since become a > sales product) and eFileCabinet (efilecabinet.com [1]) The latter is a cloud > based solution with a monthly charge per user. That isn't out of the question > even an Open Open Source product that we would roll our own and house > ourselves. > > In my searches on Google SharePoint (I think that is the Microsoft Product) > never popped up. Google Docs is out of the question and even just supplying a > share across the VPN is out as well as the access is not robust enough and > then you have the indexing and contents as well as naming of documents with > descriptions. > > People can reply directly to me or to the list. I will post some type of > summary in a few weeks to the list in case someone wants to use it for > themselves later on. > > Thanks in advance. > > -- > > John J. Boris, Sr. > > _______________________________________________ > Tech mailing list > Tech@lists.lopsa.org > https://lists.lopsa.org/cgi-bin/mailman/listinfo/tech > This list provided by the League of Professional System Administrators > http://lopsa.org/ -- John J. Boris, Sr. Head Freshmen Football Coach Camden Catholic High School _______________________________________________ Tech mailing list Tech@lists.lopsa.org https://lists.lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/ -- Who: Lawrence K. Chen, P.Eng. - W0LKC - Sr. Unix Systems Administrator with LOPSA Professional Recognition. For: Enterprise Server Technologies (EST) -- & SafeZone Ally Links: ------ [1] http://efilecabinet.com
_______________________________________________ Tech mailing list Tech@lists.lopsa.org https://lists.lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
