I’m not sure that I am going to like it. I won’t know unless I try. What will it fix? What will it break. Can I roll back if it doesn’t work for me. Ah, yes, the real question: CAN I ROLL BACK IF IT DOESN’T WORK FOR ME?
Recently in web development Category
Coming soon: orcmid @ Apache
It is natural for me to want to use my existing deployment model for the new site.
Not So Fast, Sparky
There are two important differences.
First, the new site serves up as a folder of http://people.apache.org/ so I don’t know how much site substructure there is under the public_html/ directory of the ~orcmid account (and I am not that certain of the URL, for that matter).
Secondly, I have to deploy from my new desktop system, Astraendo. That is because my developer access to ~orcmid and ~orcmid/public_html/ are via Secure Shell (SSH). I will use PuTTY and its companion SFTP utility to deploy to the account and from there to the Internet. This works better if I deploy from a working folder on Astraendo rather than on the (to be migrated someday soon) development server that my other sites are deployed from.
The first step is to produce a single default page and have it visible over the web. I will do that this way:
- Setup http://compagno/orcmidApache. This is a new sub-web on my local development server. It is where I will make my pages using FrontPage or other web-authoring tools.
- Create Visual SourceSafe project $/orcmidApache and have it bound to the compagno/orcmidApache development sub-web as its source control system. I can then make a couple of starter pages and also create an images subfolder.
- Create Visual SourceSafe project $/apacheOrcmid/web. This is a mirror (by sharing) of content in $/orcmidApache. This is where a current set of authored pages are found for publishing.
- Create computer folder C:\publicca\ApacheOrcmid\web. This is where a current version of the web content is staged. It is the assigned working folder (on Astraendo) for $/apacheOrcmid/web. It is refreshed by Get Latest Version from $/apacheOrcmid/web only when I want to refresh the public site. Staging here is decoupled from subsequent development until another Get Latest Version is pulled over.
- Newer pages are published by SFTP transfer from C:\publicca\ApacheOrcmid\web to the ~orcmid/public_html folder at people.apache.org.
This chain seems lengthy only the first time. After that, everything just flows. This structure also allows for backup from the site to the development source-control system, although I don’t expect to have need for that.
Wholesale upgrading to Microsoft Windows 7 on the Centrale SOHO LAN breaks the existing nfoCentrale web deployment model. Until I can resolve that problem, I must keep a Windows XP PC system operating well enough to sustain the web-deployment model.
Pages can be checked-out for editing in FrontPage-aware client software, but they can’t be checked back in from the same Windows 7 client PC. The development server reports that the page is checked-out to a different user (e.g., compagno/vicki instead of scritto/vicki) even though the check-out was done using the same Windows 7 client (and scritto/vicki successfully logged into the development server as compagno/vicki).
I do not know what is different between Windows 7 client PCs and Windows XP client PCs that is at the root of this difficulty. I also don’t know to what degree the use of a Windows XP Professional PC as the development server might contribute to the problem.
I do have some ideas about what I might be able to adjust and where I can adjust it. But while I explore that, I must not do anything that puts the procedures that do work out of business.
Because Scampo, my Windows XP developer PC, had to be retired quickly with an emergency upgrading to a Windows 7 Ultimate developer PC, I adjusted my plans for the preservation of web deployment into three main stages:
- Urgent Mitigation: I will always have a Windows XP client for web development and deployment until I can remedy the situation with Windows 7 client PCs. I needed to ensure that while Scampo was being retired and before the replacement machine was in place. I needed to ensure that my Tablet PC, Quadro, my last-remaining Windows XP PC, was up and running with the necessary client software before Scampo failed completely or was decommissioned.
- Client-Side Stability: I shall preserve the Windows XP client even even after the web-site development configuration is moved from a Windows XP PC to the Windows Home Server (based on Windows Server 2003 and a later version of IIS). This is necessary to reduce the number of changes happening at once and have a known successful case as a backstop while trouble-shooting and confirming the move.
- Preservation of the Model with New Tools: I will solve the Windows 7 FrontPage Client difficulties using the new location of the development IIS, FrontPage extensions, and VSS. This will include migration to Microsoft Expression Studio if possible. When that has succeeded, whatever it takes to do that, I can retire use of Microsoft FrontPage knowing that I can accomplish all essential provision of my deployment model with Expression Studio on Windows 7 and beyond.
I am not prepared to give up the current deployment model. Under that model, all of the nfoCentrale sites and hosted blogs depend on the integration of a Microsoft Internet Information Services (IIS) development site with Visual Source Safe (VSS) for version-control and backup of development site content. FrontPage Extensions and FrontPage-aware clients are then used for orchestrating the authoring of static pages. The public, hosted-web sites are populated from the IIS development site via File Transfer Protocol (FTP) for synchronization between a local extract of a development-site release and the hosted web.
Certain design-time capabilities of FrontPage and Visual SourceSafe are essential to the model and I will find a way to preserve those essential characteristics by any means possible.
Minimum First Step
As the result of beta-test usage of my Tablet PC, I no longer have a copy of Microsoft FrontPage on that machine. I also have managed to discard my install disc for Microsoft Office FrontPage 2003. So I shall use the copy of FrontPage 2002 Upgrade that had been on Vicki’s machine until we updated to a new Windows 7 Machine. That secured step 1.
It now occurs to me that I can also use Microsoft Office FrontPage 2002 in Virtual PC XP Mode on the new developer desktop. This may be an opportunity to troubleshoot the check-out and check-in breakdown as well.
To be continued …
- 2011-01-27 Quadro: Standing in a Smaller Window
- 2011-01-20 Scampo Deathwatch – My Desktop Computer in Its Decline
- 2010-12-21 Astraendo: In Production
- 2010-10-31 From Scampo to the Stars and Beyond
- 2010-09-23 nfoCentrale Web Deployment Pillars
- 2010-09-01 20 Active nfoCentrale Domain Names
- 2010-05-21 nfoCentrale: Anchoring All of It
Because it is not an Alienware Aurora, I thought of changing the name from Astraendo to Eo (or E0 for you old mainframe hackers). But, I figure, a good category is silly to waste. In addition, I realize that the desire to return my attention to abstractions and foundations of computing inspires me. I’ll keep the name.
I must resist premature retirement of Scampo. I can’t switch to a Windows 7 machine until I have figured out how to support my web site development using IIS, FrontPage Extensions, and Visual SourceSafe. The problem is having Expression Web and Windows 7 be able to do check-outs and check-ins properly. Until I have that resolved, I need a copy of Windows XP running Microsoft Office FrontPage 2003. Beside rearranging the SOHO Layout and cabling, that is the critical-path prerequisite for the development-system upgrade.
There should be FrontPage Server extensions on WHS, my Windows Homer Server operating on a Windows Server 2003 base. Assuming that there is a standard configuration of IIS 6.0 (which is already used for web functions of WHS on the HP MediaSmart Server), it is straightforward to install and enable FrontPage 2002 Server Extensions (and WebDAV for that matter), and I have done that.
If I need to revert to an IIS 7.0 Server on Windows 7 or Server 2008 at some point, I’ve just gone by the use-by date for IIS 7.0 FrontPage Server Extensions. However, there is an independent source.
Once I have the extensions, I need to create a site in addition to the ones already there. I then need to see if I can have the site, with FrontPage Server Extensions, integrate with Visual Source Safe. After confirming that I can get check-in and check-out working properly from FrontPage 2003 on Windows XP, I need to see how to get it working from Expression Web from Windows 7. At that point, serious migration of functions can proceed.
I remind myself of that here because I must stifle myself and not go overboard installing and cutting over to Astraendo until this is accomplished.
I must declare bankruptcy for the timeline and work items of “Got SMRs?” and my 2011 nfoProjects Launch Sprint. I have completely fallen off the cliff into the rapids and that timeline and its work items are DOA. I will update the current spreadsheet timeline documents to reflect that sorry state.
All of the activities are relevant, but I have been completely distracted from them. Furthermore, it is clear that I don’t have the capacity to forward them in parallel. This may have to do with having too much work in progress as well as lacking a sustainable pace (a major discrepancy between my vision and my perfornance).
I shall tear down my planned structure and create a fresh one with an honest and workable critical path. That means finding what there is that is essential and indeed critical. One problem is that there is only one resource (me), so moving something off the critical path means putting it beyond the essential immediate. The dependencies must work for that. I suspect that means creating technical debt and I will have to risk-manage that too.
In particular, I am still shifting gears in 2011, and the support for that includes straightening out the blogs, web sites, and the development procedure for them. Now to identify and work on the critical path for that, with the critical end point being whatever that much takes.
Whenever I write “SMRs” I think of smores, the Graham cracker, Hershey bar, and toasted-marshmallow confection that my family learned on my sister Carol’s return from a Girl Scout Brownie outing. I think our mother already knew of them, but my recollection is that’s when they entered my experience.
That is, perhaps, my playing with the notion of Specific Measurable Results, the outcomes in reality of defined, actionable tasks. SMRs as tasty morsels. I did not realize how much the term has been taken over by business-coaching lingo. I confess to first hearing the expression in 1993, from Landmark Forum Leader Linda Zraik, and it was already a catch-phrase in that setting.
The combination came up for me today when I realized that my nfoProjects Launch Sprint is too airy and lacks SMRs for me to chew on. Consequently, it is easy to let days go buy without measurable results. That was happening here:
I noticed this because there were no simple actionable work items that led to SMRs, even basic, produce-this-today ones.
Looking more closely, the labels on the stages that the boxes represent are conceptual. It made sense for envisioning and looking at opportunities in the available time, but that isn’t concrete enough.
Today, I have been stepping back to create SMRs for the stages themselves, and then for those individual results that I will produce on the path to the SMR of the stage.
Although it would be wise to start at the end point of the Sprint and work backwards, I will fudge and start by looking closely at the stages that I am managing over the current fortnight. I won’t go long without filling the gap from the objective back to what’s next right now, but I am forcing in some small immediate activities to ground my effort in the meantime.
The tracking spreadsheet is not going to include all of that detail. I will use my Personal Kanban board to manage them in a fluid way and also ensure that I provide supplemental envisioning, planning, and reflection on a weekly basis.
I’ve been looking at all of the projects that I have and the number of burners there are for my efforts.
In the past week, I realized that I needed a consolidated plan with defensible, risk-managed timelines for getting to where I want to be on New Years Day, 2011: Launching a family of interoperability-centered tutorial, test-fixture, and reference-implementation projects around document formats and their processing. (Hmm, short enough for an elavator speech but not something I would be able to recite at gunpoint. Needs work.)
I have gathered my envisioning into a spreadsheet that keeps my commitment and my management of it in a structure that holds my vision in existence. Here is the finish line and, at the bottom, where I am standing right now:
This is a high-level perspective. I blocked out the calendar weeks and arranged more-or-less week-sized chunks in which to make headway along four tracks. Since I am a solo developer, this is enough for me to follow and use as an instrument for keeping my eye on critical setbacks and interdependencies that I must keep my attention on. There’s an opportunity for some Personal Kanban and Personal Software Process.
The OpenDocument Format continuation work is related to my current involvement on the OASIS ODF and OIC Technical Committees, as well as the ISO/IEC JTC1 SC34 WG6 working-group for ODF maintenance. As ODF 1.2 moves toward ratification as an OASIS Standard in early 2011, there is a hiatus that allows me to devote more attention to Interoperability and Conformance and the alignment of the ISO/IEC and OASIS efforts.
The blog, web, and development system (Centrale) efforts have been identified here in a piece-meal fashion. This three-track view provides my first projection of a coherent feasible structure on those activities. There will be many opportunities for adjustment and course-correction as progress is made and details of further steps are reviewed.
In simply arranging these activities, I found how to avoid a situation in which I was retiring Compagno and in a sudden-death gap on being able to move everything I needed to the Windows Home Server. This was definitely sky-diving without a parachute. What I had not seen until I placed the activities in this form, is that Quadro, my Tablet PC, can mirror Compagno perfectly and also assist in verifying how migration from FrontPage to Expression Web can be accomplished while preserving practices that serve me too well for them to be abandoned without a fight.
I have also noticed, already, how having my eye on New Years Day gets me out of bed in the morning, excited to make further headway on these activities.
I will be reporting results and updates here on the week-by-week progress. And, for those playing along from home, here’s the tracking document on-line:
I have a serious itch to move from Scampo to a fire-breathing multi-core Windows 7 development platform.
Although I have my eye on some contenders (basically a hot Dell XPS 9100 or its comparable Alienware Aurora configuration), I must resist making the jump. Although I will likely upgrade before the end of 2010, there are some things I need to keep my eye on.
Core Scampo Functionality
Before I cut loose from Scampo, I must remember that there is on-going use of this Gateway 831GM Media Center PC running Windows (Media Center) XP SP3. As much as I want to be on Windows 7 and desire to be able to upgrade to tools that are no longer supported on Windows XP, there is some baggage.
- An E-MU 1820M that I need to move to Windows 7 Ultimate 64-bit (and there are new drivers available, and Cakewalk Sonar will support the 64 bit operation). I must make sure that the upgrade has slots and connectors for the 1820M PCI boards.
- My Hewlett-Packard ScanJet 7400C, although a little long in the tooth, is still my primary scanner and it will need to run on the upgrade (name yet to be chosen).
- Battery Backup may need upgrading to handle the 550-850 watt power-supply, depending on configuration. Apparently, some of the higher-performance machines will see spikes on backup kick-over and shut down. That is an additional problem, as well as making my office setup provides enough power through the outlet cluster my computing equipment is dedicated on.
- Gigabit Ethernet? A router/hub upgrade to provide Gigabit Ethernet connectivity to the Windows Home Server might be wise. The greater capacity of the new development engine, with multiple virtual machines for interoperability testing and confirmation of software porting on multiple platforms may stress the WHS capacity. There may need to be some upgraded drives and perhaps even an upgraded WHS machine.
- Although not part of the initial upgrade, the desire is to eventually upgrade from 24” to a larger monitor (at least 27” but 30” preferably) once they are more affordable and Energy Star compliant. The upgrade will have the graphics power necessary for that and for multiple monitors, even if the completion of upgrading is later.
Although I want to deal with fundamental system configuration and upgrade issues, there are also problems of preserving in-progress project and dealing with my current dependence on software that will have to be changed-out as part of platform upgrading.
- Microsoft Money is now on life support. I still use it and have not migrated to Quicken. I doubt that the version I have will run very well on Windows 7 (although it might on the Windows XP VM that will be provided). I need to be prepared to do a financial software migration along with or prior to pulling the plug on Scampo.
- Microsoft FrontPage and web-development Issues. It appears that Microsoft FrontPage and FrontPage extensions don’t operate properly between a Windows 7 system and a Windows XP SP3 machine running IIS and the FrontPage extensions. This may also be related to issues with regard to Visual SourceSafe integration. There is a long-overdue retirement of Compagno that moves the IIS development web to WHS, moves VHS support, and then moves from FrontPage use to Windows Expression for continuing web development. I can’t stop web development using Windows XP SP3 until I accomplish that transition. It will be an eventful November-December 2010 because of this issue alone.
- Microsoft Office 2007/2010 Upgrading. Because I run two copies of Outlook, one on desktop, one on laptop, I must have the same version of Microsoft Office, with Outlook, on both machines so I can move my .pst files back and forth. I am on Office 2007 now, but my Windows XP SP3 Tablet PC does not handle Office 2010 well enough for me to do a synchronized move to Office 2010. It is also a pain to run multiple versions of Office on the same platform, another reason why I want to have a VM-capable development machine.
I need to be running 2010 and beyond for interoperability testing purposes, but I will have to stay on Office 2007 as my working installation simply because of the need for Outlook parity. At some point I would like to have a single PST that is on the traveling machine, but I can use it from the desktop when I am in the office. If I keep the PST on the traveler, using it in the office by remote-desktop is an option but it may not go quite so well with the degree that I use copy-and-past between Outlook and other applications. A weird but promising solution would be to run Exchange server on a future traveling replacement for my current Tablet PC, and then I would have full Outlook access to it locally and from my office desktop when I’m at the home office.
There are probably more considerations and setbacks I have to deal with. But this is enough top-level to serve as a reminder that I can’t just throw a switch and have a shiny new high-performance development system up and running. I must address the staging issues. And I shall.
[Cross-posted from Spanner Wingnut]
For those wondering if the now-dormant blogs will every see the light of day anew, here are the dreary steps that will be confirmed before going farther:
- The procedure for configuring and maintaining images uploaded via Windows Live Writer will be fully documented on nfoCentrale Status. The accuracy and completeness of that procedure will be verified by adding the support to Spanner Wingnut and also maintaining it there..
- The next visible changes will involve learning to adjust the sidebars on the blogs, including customization of plugins, templates, and styles for individualization of the different blogs operated under the single MovableType engine. This will also be seen on both nfoCentrale Status and Spanner Wingnut.
- There will be some invisible work to support how a blogs features are adjusted over time but a record of changes and the progression can be maintained. This will be accompanied by infrastructure adjustment as part of the deployment model for the various blogs, starting with nfoCentrale Status and Spanner Wingnut, with the big test being restoration of an existing, now-dormant blog. The prime candidates for first restoration are Orcmid’s Lair, Pursuing Harmony, and Kiln Sitter’s Diary.
- There is also more work to complete and stabilize the nfoCentrale site as the repository of all of the procedures and template-development work that is promulgated across all of the blogs and sites.
My current thinking is to have this completed enough by the end of the year that additional blogs start waking up and there are useful developments on those blogs and their related sites.
When I had to become much more web-server savvy as part of having a private installation of Movable Type on my web-hosting account, I realized that my web site deployment and maintenance model had exceeded my mental grasp. I spent several days making diagrams using borders around groups of annotated spreadsheet cells. Although I could have used Visio, I found that Excel was much simpler for this kind of diagramming.
A full image of the diagram is not great for viewing on the web. But you can get a general sense of the way the pillars are connected, starting with my development web folders on the left and my Internet domains on the right. This image is better for keeping inside my notebooks where I can remind myself how to stage something I am working on.
This diagram is going to see heavy duty in defining and documenting the more-detailed procedures that make this work. A view of the worksheet itself is a little better for taking a closer look:
This is draft 0.12. I will be refining the diagram as I work through further details. Here is the Excel worksheet that I used for this version: