Present:

• Chair: Martin MacLeod, Engineering Computing
• Erick Engelke, Engineering Computing
• Hon Tam, Engineering Computing
• Dave Walsh, Engineering Computing
• Mahindra Sangara, Systems Design
• Richard Scott, Civil Engineering
• Dennis Herman, Chemical Engineering
• Steve Hitchman, Mechanical Engineering
• Ray White, IST
• Hector Budman, Associate Dean for Computing, Faculty of Engineering
• Mike Herz, Civil Engineering

Regrets

• Paul Martin, Management Sciences
• Bernie Roehl, Electrical and Computer Engineering
• Eric Praetzel, Electrical and Computer Engineering

Discussion

1. Netapp

• EC has bought more disk space and is increasing all departmental quotas by 50% - no objections raised
• Dennis raised issue of NetApps disks costing significantly more, Erick mentioned free replacement and reliability
• online stats available at  http://monmon.uwaterloo.ca/mrtg-filer-1.0.0/engfile.uwaterloo.ca/

• Several departments have a Davis Centre presense
• Erick has allocated passwords for departmental reps on Extreme switches and is currently in the process of getting keys for the various rooms

• Mike suggested the EC homepage have links to useful resources like the DNS change page, etc.
• Page has been set up:  http://www.eng/~eng_comp/system/

• APC has approved the name: Nexus
• Hon explained the empty root for schema, DNS, GC for better reliability and flexibility
• test domain will be in different forest to test schema changes
• no plans for another peer domain, but it is possible
• Hector said APC aware we will move many machines to W2K, but timing is undetermined
• Bruce suggested conversion to be completed by September 2002, but this assumes hardware upgrades can be budgetted and implemented
• Mike asked if there will be any problems if, say, Civil doesn't join Nexus for staff and faculty.  EC doesn't envision any major problems, but did mention several reasons.  Mike was not swayed by these suggestions.

• $ 175 per drop estimate includes parts and labour
• better keep very good records if costs are more than average - eg. asbestos, long runs
• cannot claim $175 for easy runs and more for hard runs, whichever is higher
• must keep records of the time spent
• $135k is left for Engineering for the 3rd year
• Martin will get an updated spreadsheet
• switch priceses have fallen - uncertain how much money this will free up for the 3rd year
• Dennis wonderred if 2nd year work can be filed still, Hector will look into this case.  First estimate approximately $10k, but then Steve points out that Chemical budgetted only $6k, and Martin observed that Chemical was billing $175 + parts.  Dennis will re-visit these costs.

• Mike: predoc is unnecessary and wasteful of time.  In his construction experience, detailed pre-built wiring drawings for these types of wires are never done.  Usually only textual wire data and sometimes schematics.
• Mike hesitant to do predoc when installing departmental room wiring, but understands when doing public hallways
• Dave: if you want CFI you basically have to predoc
• Hector: this is not only a CFI issue, it is a requirement for all wiring
• Several people pointed out that plant ops and IST would prefer if we didn't wire at all, we had to struggle to be allowed to wire, predoc is one of the concessions we have to make to retain this right
• Mike: this creates 3 layers of beauocracy, it's too much
• Dennis agreed
• Mike: EC knows all the parts we buy
• Dennis: we buy all hardware from same source
• Bruce: points to section 1.5, #4 in  the guidelines , we are bound to follow it
• Hector: it's not approval, it is a sharing of information
• Dennis: 90% of time we don't need special shared information
• Bruce: who convinces plant ops that we need to change rules
• Mike: can't we make some sort of compromise.  We already provide details during IP requests, why not add a few more details.  Also, IST's requirement for predocs is void of any details, this means anything we provide could be considerred adequate.
• Bruce drew a simple diagram and showed the bare essentials of what is required
• Mike: can't we just draw two boxes and a line connecting them
• Martin gaven an example of why a straight line is insufficient - wiring must follow the contours of the wall
• Mike: don't always know everything in advance, like heating ducts, until ceiling is open, should draw the diagram afterwards, happy to do it then
• Mike wants a detailed checklist of all the steps to be taken before and after laying a wire, current document is too vague
• Martin: current document allows room for flexibility and user judgement
• Erick: need predoc, otherwise everything is already laid and thus grandfatherred